Previous fix for VC++/strcasecmp() was wrong-- use StrCaseCmp().

[xboard.git] / ChangeLog

ChangeLog for XBoard/WinBoard 
-----------------------------

(latest entries created by git log --no-merges --pretty="* %ai: %s%n %b" vnew vold)

* 2009-09-05 09:44:21 -0700: updated AUTHORS and README files for release
 
* 2009-09-05 08:52:09 -0700: fixed some typos
 
* 2009-09-04 22:39:45 -0700: set castling rights after setting up a position
 I was going through old WB forum posts, and I encountered a
bug report for a bug that I never fixed: The castling rights should
be properly set (actually: cleared) after setting up a position
in Edit Position mode (as opposed to pasting a FEN).

* 2009-09-03 23:11:18 -0700: removed install files from master
 they now live in their own branch

* 2009-09-03 22:30:18 -0700: updated polyglot protocol
 I added a new option type "-reset" to the protocol for the new option feature.
4.4.0 does not implement it yet, but Michel wants it for future Polyglots,
and I think it is best to update the protocol in one big swoop. So I want to already
include it in the specs now. (engine-intf.html)

I made xboard's response to an unknown option type compliant with the new protocol specs.

* 2009-09-03 22:28:50 -0700: fixed some typos
 
* 2009-09-03 20:47:13 -0600: Updated project files for DevCpp and Visual Studio to add wsettings.c
 
* 2009-09-03 20:46:33 -0600: Added newline at end of file to silence warning.
 
* 2009-08-31 20:29:49 -0700: workaround for error message for missing pieces used in variants and initstring bugfix
 error messages showed for non-existing pieces showed up for old user supplied pixmaps. Loading a pixmap of the king for
 the missing pieces now.

Another bad bug surfaced: the -initString argument does not work in any
XBoard version! (Including 4.2.7b) The escape sequences for linefeed in
the string are not understood: -initString "new\nrandom\n" actually sends
the \ and n to the engine (which then does not recognize the line as a
valid command)!

So I also added a patch to expand escape sequences in the InitString and
ComputerString command-line options.

* 2009-08-31 20:27:24 -0700: updated installe script
 I re-organized the doc files in the install, so they now all go in the
WinBoard\doc folder
(and none in the root). I so included the COPYING and COPYRIGHT files there.
This required some updates of the links in some html files, and of the menu
link
to README.html in the install script. I also fixed an error with the
current directory
in the menu short-cuts of some engines (which started in the fonts folder, and
could not find the opening book there...).

* 2009-08-31 20:25:17 -0700: updated documentation
 
* 2009-08-31 20:20:55 -0700: updated installer script
 
* 2009-08-31 19:51:48 -0700: new help files for windows
 
* 2009-08-25 20:41:29 -0700: updated installer
 
* 2009-08-25 20:36:53 -0700: undid some previous changes
 some files got mixed up

* 2009-08-25 20:32:22 -0700: fixed small bug in wsetting
 
* 2009-08-25 20:29:06 -0700: trigger a focus switch to ICS input field when typing
 old behavior was that "." or "," triggered this, now every printable character  except 1-9 will trigger the focus switc
h

* 2009-08-25 20:26:20 -0700: bug in backend.c when receiving lines from ICS
 A space was appended to every character received from an ICS in stead of just at line breaks.

* 2009-08-24 23:47:48 -0700: fixed warning messages from compiler
 
* 2009-08-23 14:24:38 -0700: fixed build on OS X
 this seems to fix it. Bug reported by Louis Zulli who also did the testing.

* 2009-08-23 11:54:28 -0700: OK, this is the beefed up version of the Polyglot info-string kludge, for WB and XB.
 
* 2009-08-23 10:22:48 -0700: added missing X_LIBS references to Makefile.am
 
* 2009-08-22 23:33:44 -0700: made the install script and executable
 
* 2009-08-22 23:19:07 -0700: changed beta1 to beta2
 
* 2009-08-22 23:14:11 -0700: fixed enable zippy in configure script and added some checks for OS X
 make tried to compile zippy even when --disable-zippy was given. Also fixed a bug in xboard.c where #ifdef instead of #
if was used to test for ZIPPY (it's either 0 or 1, so we need #if).

Also added some special checks for OS X, not sure if these will do it, but it's a start;)

* 2009-08-22 19:09:54 -0700: added configure summary to configure output, also cleaned up configure.ac a bit
 used AS_HELP_STRING where possible

* 2009-08-22 18:10:57 -0700: added check for malloc.h
 this should fix a build error on OS X

* 2009-08-22 17:47:44 -0700: A small cosmetic change in the size of buttons inside a groupbox of the Engine-Setup dialog
.
 
* 2009-08-22 17:32:14 -0700: make xboard usable with UCI engines
 I made a small patch in xoptions.c (where the corresponding Engine Settings
dialog was for XBoard), to make it at least usable for UCI engines: I now
divide the options over upto 4 columns, if there are too many to fit into
a single column. This seems to work. Still looks very ugly, but it is
functional.

* 2009-08-22 17:30:05 -0700: fix premove for winboard
 This might fix the illusive premove problem. In 4.2.7 the code was not properly
indented, and this led to an unbalanced braces error that I then fixed the
wrong way.
Click-click premoves did only work after clicking the toSquare twice. It
might be
that they actually did work after clicking it once, but that the highlights
were not
displayed (although set) for lack of a display refresh.

* 2009-08-22 17:28:11 -0700: fixed bug for double saved games
 This patch should solve the problem with doubly saved games when XBoard is
closed.
I calculate a cecksum of each saved game now (including move comments),
and suppress saving if the checksum is not changed.

* 2009-08-22 17:25:58 -0700: fixed xengineoutput.c, to keep the duplicated code identical to that of wengineo.c.
 
* 2009-08-22 17:24:08 -0700: updated xboard.texi: added warning that pixmaps are not build for all variants
 
* 2009-08-22 17:19:11 -0700: added a dialog for engine-specific option settings
 
* 2009-08-22 17:09:58 -0700: updated some documentation
 
* 2009-08-22 17:07:10 -0700: fixed several bugs in the backend and added some more polyglot support
 * I added recognition for the new UCI3 option types that are recognized in the WB
option features, to make sure this version is compatible with future Polyglots.
Also fixed a bug that failed to strip an asterisk prefix from the first
combo-box item.

* I fixed the ICS-time-in-PGN bug.

* I fixed a bug which caused a space to be missing when long lines broken up by
an ICS were joined.

* 2009-08-22 17:02:58 -0700: better polyglot support for engine ouput
 I fixed the algorithm for clearing the Engine Output window to work better with
the latest Polyglot, whight outputs UCI "info strings" amongst the thinking
output.

* 2009-08-22 16:23:32 -0600: Beginnings of installer for 4.4.0.  HGM asked me to forward the nsi file, but it seemed bes
t to just add the installer tree to git even though it's not finished.  You must install NSIS then FontName-0.7.exe (inc
luded in installer folder) then run build.bat in the same directory.
 
* 2009-08-05 22:18:29 -0700: patch backend.c to make the -noGUI option also work in ICS mode.
 
* 2009-08-01 13:09:49 -0700: temporary fix for pre-select
 As a temporary kludge I replaced looking into the holdings for the piece type by figuring out which piece would belong 
on the clicked holdings square. This is a non-trivial process, as it depends on which pieces participate in a certain va
riant.

* 2009-07-30 00:19:37 -0700: changed click-click moves
 I got complaints about the way XBoard 4.4.0 handles click-click moves, in particular changing the selected piece after 
selecting King (for which I suppressed interference with FRC castling in a rather brutal way). People tell me it is very
 annoying that clicks sometimes don't work to select a piece. WinBoard uses refined logic in the back-end to handle this
, but the XBoard front-end is not interfaced with that yet. (This would require a major rewrite of the mouse handler.)

This temporary patch suppresses this behavior in all variants that do not support Fischer castling, and even there only 
exhibit it for Rook-after-King selection (which will be interpreted as an attempt to enter a castling move as K x own R)
.

* 2009-07-26 08:43:35 -0700: option -adjudicateDrawMoves is not included in the XBoard list.
 added it to the option list

* 2009-07-21 20:23:23 -0700: converted files from dos to unix format
 run dos2unix over a bunch of files

* 2009-07-16 07:43:38 -0600: Stripped out unused VersionInfo
 
* 2009-07-15 23:16:19 -0700: changed version to 4.4.0.beta1
 
* 2009-07-15 23:12:37 -0700: updated INSTALL file to add ./autogen.sh
 
* 2009-07-15 22:44:05 -0700: higher number of engine options possible now
 I upped the number of options per engine from
50 to 100, and the average number of characters per option from 10 to 20.

* 2009-07-15 12:30:37 -0600: Fix of bad bug
 
* 2009-07-14 23:41:36 -0600: Added version information to resources.
 
* 2009-07-14 23:40:50 -0600: Added Jaws targets to IDE projects.
 
* 2009-07-14 23:39:12 -0600: Final makefile adjustments and compiling instructions.
 
* 2009-07-13 22:35:17 -0600: Cleaned up garbage left behind from my makefile update.
 
* 2009-07-13 22:26:56 -0600: Updated cmdline makefiles to be able to produce JAWS binaries.
 This breaks support for MSVC 4.x.  It may still be possible to
use that compiler if you obtain the latest SDK from Microsoft.
MSVC 6.0 still works, but I had to change the target windows
version to 0x500 which causes warnings about building beta
applications.  This is all due to the use of MENUBARINFO data
type in the jaws binary.  You can reduce the WINVER to 0x0400
safely if you are not building a JAWS binary, and MSVC 4.x
will work fine.

* 2009-07-13 22:26:10 -0600: Removed dependency on jfwapi.h
 Allows anyone to compile a jaws compatible binary.

* 2009-07-13 22:23:31 -0600: Fixed duplicate resource in winboard.rc
 PLEASE remember to pull before commiting.  This is the 2nd
time I've fixed this.  It's a waste of time to keep fixing
the same problems.

* 2009-07-13 19:36:58 -0700: resolved conflict between Xt and Xaw libraries
 seems like we don't need to link against Xt.

* 2009-07-13 19:33:57 -0700: help-menu was not un-grayed in the JAWS version, when it played with engines
 
* 2009-07-13 19:29:53 -0700: integrated jaws into winboard
 I integrated the JAWS patch now into WinBoard in such a way that the
JAWS and standard version can use the same resource file. This new
winboard.c contains the improved dialog layot of the JAWS version,
but none of the menus and JAWS-specific acclerator keys. The latter
are now generated at run time by code in jaw.s. As a result jaws.h
could be cancelled, as there the JAWS-specific defines are no
longer needed outside jaws.c (now the .rc does not use them), and
is incorporated in it.

* 2009-07-12 15:37:26 -0700: fixed problem in autoconf using Xaw3d
 rewrote the section for Xaw3d in the autoconf file

* 2009-07-12 15:29:28 -0600: Added Jaws support to GCC makefile
 
* 2009-07-12 15:28:20 -0600: added help support: MSVC IDE and DevCpp IDE project files
 
* 2009-07-12 15:26:49 -0600: Removed obsolete makefiles and support files.

* 2009-07-12 12:33:26 -0700: updated AUTHORS with the help of git
 use

git log --pretty=oneline --name-only --author=AUTHORNAME v4.2.7..HEAD  |sort |uniq

a bit of hand editing and a quick emacs macro later and you have a nice comma seperated list of all files changes by one
 author between two commits.

* 2009-07-12 12:31:47 -0700: updated Changelog using git log --no-merges --pretty="* %ai: %s%n %b" HEAD v4.2.7
        
* 2009-07-12 11:34:28 -0700: cleaned up some more copyright notices
 
* 2009-07-12 11:22:30 -0700: updated copyright to reflect A. Scotte as copyright holder
 need to mention him, since the paperwork won't be finished before the next release

* 2009-07-12 10:56:29 -0700: updated patchlevel to "k"
 
* 2009-07-12 10:53:50 -0700: fixed autoconf problem with VERSION/PATCHLEVEL variables
 xboard used it's own version of PRODUCT, VERSION and PATCHLEVEL variables. Switched to PACKAGE_STRING provided by autoconf

* 2009-07-12 10:29:46 -0700: cleaned up old CVS left overs
 removed $Id tags from all files

* 2009-07-12 10:10:17 -0700: converted xhistory via dos2unix
 
* 2009-07-12 10:09:24 -0700: fixed crash in history window
 someone discovered that opening the move-history window in XBoard leads to
an immediate crash. Turns out that renaming the menu item broke it! (The
name of the item was apprently used elsewhere in the code to access it,
for putting a check mark there.)

* 2009-07-12 10:04:30 -0700: fix for Animation of castling moves in Chess960
 Animation of castling moves in Chess960 turned out to sometimes make the
king disappear from the display. For the time being I patched this in
xboard.c to suppress animation of FRC castling moves altogether. As FRC
castlings internally are represented as KxR, the to-square does not make
sense anyway, which made the animation look very weird even when it led
not to a king disappearance act. And the Rook move of a castling is never
animated.

* 2009-07-11 22:44:53 -0600: Update GCC makefile to remove .obj targets.
 Dunno how they got in there, but they should be .o

* 2009-07-11 22:42:41 -0600: Update makefile dependencies on help.c and help.h.
 
* 2009-07-11 22:36:59 -0600: Allow compiling with MSVC6.0 and 4.1
 Move prototypes into help.h, after making sure that DWORD_PTR is
defined.  Include help.h from winboard.c and help.c so both files
benefit from the definition.

* 2009-07-11 18:26:16 -0700:  lists these accelerator keys in the menus for JAWS
 
* 2009-07-11 18:24:19 -0700: added shortcuts
 This patch allows the use of Ctrl+C and Ctrl+V for Copy Game and Paste even in I C S mode when the board has focus.

* 2009-07-11 18:22:17 -0700: better handling of pasting a game without PGN tags
 patch in wclipbrd.h to make it smarter to decide between pasting a game or a position so it can handle games without PGN tags. If it doesn't parse as a FEN it is assumed to be a game.

* 2009-07-11 18:20:56 -0700: fixed parser error for knight move
 The parser had a strange quirk: it recognized bd2 as a valid Knight move
(Nb1-d2) while this SAN format should be reserved for Pawn moves only.

* 2009-07-11 18:19:30 -0700:  fixed a few warnings in connection with the help patch (now warning free)
 
* 2009-07-11 18:13:50 -0700: some major re-assignment of accelerator keys in the JAWS
 
* 2009-07-11 16:54:53 -0600: several small fixes
 uptracked winboard.ini

Cleanup: moved "Some definitions required..." from winboard.c to config.h,
and combined some definitions from borland w/ MS.

Changed internal HtmlHelp return type and 4th argument to facilitate various
compilers.  These changes, while not strictly adhering to HtmlHelp() conventions,
won't impact anything since the function modified is internal only.

Modified HtmlHelp return values to integer to silence warnings.

Added help.c to the project in makefile.gcc & makefile.ms.

Fixed duplicate resource in winboard.rc.

* 2009-07-09 22:08:15 -0700: added JAWS support; help files for Vista; bugfixes for winboard.c
 * JAWS support
* help files should now work under Vista (the user has to have either a program to view old windows help files or to view html-files installed)
* I fixed the tabbing between all windows that were added according to Tim's sugestions.
* I added typing in of a move number in the move type-in dialog as a shortcut to going to that position by use of the < and > buttons. Mainly intended to ease life for the blind, but I think it can be generally useful. (Although the move-history dialog provides that functionality too, this is quicker if you don't happen to have it open.)
* I added typing of a FEN in the move type-in dialog in Edit Position mode. (Which could not be used without mouse at all, and is defective in crazyhouse even with mouse, as there was no way to populate holdings.)

* 2009-07-09 22:03:00 -0700: some bugfixes: ICS error handling
 * play a different sound for takebacks in ICS mode
* pop up error messages for illegal moves for typed-in moves

* 2009-07-09 21:50:41 -0700: removed no longer needed debug message
 
* 2009-07-09 21:48:23 -0700: fixed compile error from a previous change
 
* 2009-07-09 21:44:54 -0700: user selectable font in game list
 This makes the game list use the same (user selectable) font as the the move-history window.

* 2009-06-30 21:27:34 -0700: fixed typos in xboard.texi documentation.
 
* 2009-06-30 21:16:59 -0700: updated the FAQ
 
* 2009-06-27 17:06:26 -0700: changed font in engine output window
 The engine-output window now uses the same font as the move-history window
(which was selectable in the "Options  -> Fonts..." menu, while the other
wasn't).

* 2009-06-27 17:03:19 -0700: fixed some characters in xboard.texi
 It seems that copying manual sections from the RTF to th texi file did
somehow mess up the double-quote characters, and changed them into some
strange (unicode?) symbols.

* 2009-06-27 16:56:36 -0700: two shogi related bugfixes
 I discovered a bug in the move generator for the Shogi Dragon King. I
guess it went unnoticed so far because this is a promoted version of the
Rook, which hardly occurs in games. (moves.c)

There was also a problem with the piecesfix I sent you before; Shogi
chosen as initial variant did use wrong piece symbols. (I had tested this
on gothic, but there it happened to work for a different reason, unique to
gothic.) This affects xboard.c.

* 2009-06-27 16:54:04 -0700: small cleanup
 
* 2009-06-27 16:51:02 -0700: adds bitmaps for shogi promotions, and selects the bitmaps or pixmaps for the more rarely used piece types on a per-variant basis.
 
* 2009-06-27 16:44:36 -0700: adds new piece types in the promotion dialog and edit-position piece menu
 
* 2009-06-27 16:13:15 -0700: updated bitmaps files
 files for fairy pieces where missing, others had names that didn't work with autoloading.

* 2009-06-27 16:06:35 -0700: fixed a bug concerning bitmaps
 
* 2009-06-27 15:53:45 -0700: give a default directory for bitmaps files (tiny change)
 this is also one of the debian patches, should be rewritten to use the correct install directory. No Author was given.

* 2009-06-27 15:47:52 -0700: verbose_error_message_about_recognized_options (tiny change from Debian patches)
 
* 2009-06-27 15:45:14 -0700: add a visual low time warning (tiny change from Debian patches)
 
* 2009-06-24 23:31:52 -0700: dos 2 unix conversion of backend.c
 
* 2009-06-24 23:30:07 -0700: exchanged some sprintf with snprintf
 found in a patch for the debian system, extended it a bit to other sprintf's
original patch included only 11 lines by Florian Ernst <florian@debian.org>

* 2009-06-24 23:07:23 -0700: enabled a patch for default settings in winboard
 
* 2009-06-24 22:53:11 -0700: updated copyright notice
 
* 2009-06-24 22:52:20 -0700: patches a bug in chosing the initial board size on a virgin startup I introduced recently
 
* 2009-06-24 22:49:33 -0700: I put in testing for the winning condition in 3checks
 on a check it just goes through the entire game to subject every position on a check test, to see if there were two more.

* 2009-06-24 22:47:01 -0700: updated copyright notice in docs
 
* 2009-06-24 22:41:40 -0700: added small patch from debian (tiny change)
 according to the debian patch tracker, this patch is by Florian Ernst <florian@debian.org>.
It fixes whitespace issues in filename called from cmail.

* 2009-06-24 22:33:13 -0700: converted dos2unix format for xboard.texi
 
* 2009-06-24 22:31:57 -0700: deleted two more files that don't need to be in git
 these files can be generated by autogen.sh

* 2009-06-21 12:10:31 -0700: major overhaul of the -stickyWindows feature
 I did a major overhaul of the -stickyWindows feature. The game-list and ICS-interaction windows are now also sticky. I also added a form of stickiness when the main window is resized; edges of auxiliary windows sticking to the right or bottom side of the main window stay attached when the window resizes (due to nr of board squares or their size). The way the window coordinates are saved in the .ini file is now relative to the main window, and some dummy options are added to be able to recognize right- and bottom-edge sticking, and reconstruct that when WB starts up with another board size. Also added are volatile options to ensure there will be space for auxiliary windows above and left of the main window.

* 2009-06-21 12:07:01 -0700: fixed long-algebraic form of drops
 The WB parser turned out not to understand the long-algebraic form of drops send by an ICS. (like B/@@-e4), so even with legality testing off I now use the SAN move for drops.

* 2009-06-18 18:35:31 -0700: Refined stalemate adjudication in Suicide, some cleanup
 Refined stalemate adjudication in Suicide, which seems to be not always won
for the stalemated side, but for the side with the smallest number of pieces.
Cleaned up the adjudication code in the process. (backend.c moves.c, moves.h)
Also includes a tiny change in the -autoKibitz format, as requested on
Talkchess.

* 2009-06-18 18:32:57 -0700: Enhancement of the way -autoLogo decides which logos to display
 
* 2009-06-16 00:06:59 -0700: bugfix and enhancement in autoKibitz mode
 changed the format of the autoKibitz: score/depth in stead of depth/score, as in PGN. I also supress the ICS time for the move now in the PGN if extended info is available.

When enhancing the functionality of the autoKibitz command (using the
opponent engine kibitz for the saved PGN) I discovered that the xengineoutput.c
in our tree was obsolete, and contained a routine of which the arguments
were no
longer compatble with the call from backend.c, and caused segfaulting. I
copied in
the  corresponding routine in wengineo.c, which was compatible. That the
compiler
did not flag this is a sign of badly organized code, I suppose...

The enhancement itself affected backend.c.

* 2009-06-16 00:04:13 -0700: modified clock layout
 I altered the display of the clocks in the presence of engine logos; I used to
stack them, but some inquiry on Talkchess sowed that people disliked the
stacking. So now I display them next to each other again, but as 2-line-high
fields to make room for te logos.

* 2009-06-15 23:59:32 -0700: new zippy option zippyShortGame
 On request of Mike Adams I added a new zippy option, -zippyShortGame, to make life more difficult for abusers of abort and disconnect for cheating.

* 2009-06-15 23:56:40 -0700: fixed some bugs in the animation of atomic captures
 
* 2009-06-13 15:27:26 -0700: some small bugfixes
 I had overlooked one place from which the FEN writer (of whicf I changed the argument types), which made WB crash if you copied a FEN (or PGN that cotained FEN) to the clipboard. I must have overlooked a warning during compile, or some depedency is not correct in the makefile. Anyway, wclipboard.c had to be changed.

I also moved adjudication of some losing conditions to before tetsing for stalemate, (e.g. in atomic, if your king get destroyd you lose, even if you have no moves after that because it was your only piece.). this affects backend.c.

I had used the same bits in the frule-modifier flags for indicating mandatory capture as was already used for indicating FRC-style castling. This led to frequent illegal move calls in FRC... (moves.h)

* 2009-06-13 10:45:17 -0700: cleaned up HTML in FAQ.html -- still need to work on content
 
* 2009-06-13 09:36:26 -0700: added option for work-around for some FRC engines in regards to FRCFENs
 It suddenly occurred to me that I had promised on CCC to include an option to provide a work-around for the problem that some FRC engines do not understand FRC FENs (and thus do not play at all under WinBoard, as FRC always has to begin with sending the initial position).

So I added the options -firstNeedsNoncompliantFEN, -secondNeedsNoncompliantFEN. This affects winboard.c and xboard.c, (for defining the option), common.h, backend.h, backend.c, (for implementing it), xboard.texi and winboard.rtf (decribing it). This option provides users with a means to always force the FEN castling field to KQkq (but can be used for many other work-arounds as well).

* 2009-06-13 09:30:31 -0700: fixed bug with random calls
 There also seemed to be a bug with the shuffling of opening positions; I had changed the declaration of the position number from long long int to u64 to comply with MSVC, but this suppressed shuffling completely. I now changed it to int. I recently noted a problem with true randomizing anyway: this was based on srand(time()), but time() apparently did not change fast enough, so that starting several games in rapid succession (because the book line contained a 3-fold repetition) produced identical games. So I now seed the random once and for all during startup, based on the msec of the time mark, and removed the other calls to srand().

* 2009-06-13 09:26:14 -0700: changed name for fairymax program in xboard.h
 
* 2009-06-13 09:25:07 -0700: removed my name from copyright message in help->about menu
 
* 2009-06-13 09:22:02 -0700: fixed -stickyWindows option
 I figured out why -stickyWindows was not working in XP and Vista
(while it always worked under Win2K), and I think I found a fix for
it that works under all Windows versions. (I tried XP and Win2K.)

* 2009-06-13 09:20:38 -0700: Updated docs for Engine-Settings menu
 
* 2009-06-13 09:19:05 -0700: Awareness of winning conditions for atomic and giveaway variants; two bugfixes
 fixed bug in claim verification
Bugfix on initial position of Great Shatranj

* 2009-06-13 09:16:47 -0700: Legality testing (mandatory capture) in giveaway variants
 
* 2009-06-13 09:16:17 -0700: don't ignore winboard/parser.c
 
* 2009-06-13 09:12:15 -0700: Added new variant Twilight Chess
 
* 2009-06-13 09:08:57 -0700: some renaming and restructuring of WB menu items, to conform more to XBoard
 
* 2009-06-13 09:07:07 -0700: Fixed bug in atomic patch that interfered with non-atomic variants
 
* 2009-06-10 21:46:11 -0700: fixed different use of engine option compared to Polyglot
 it turns out the new Polyglot uses a sightly different version of the GUI->engine option command than the one I proposed (and implemented in XBoard; Michel sneaked in an '=' sign between the NAME and VALUE of the option, to allow for more unambiguous parsing.

I guess in principle this is a good idea, so I adapted XBoard to Polyglot rather than the other way around. The result is that there are canges in xboard.c, and the documentation.

I noticed that I unjustly had internationalized the option strings to be send to the engine in xoptions.c; this was of course nonsense, as this was WB protocol. So I undid that too.

* 2009-06-10 21:37:47 -0700: add the already supported variants 3check and great (=Great Shatranj) to the New Variant... menu.
 
* 2009-06-10 21:32:22 -0700: animate explosions in variant atomic (if animate moves is on).
 
* 2009-06-10 21:30:19 -0700: fixed castling bug in variant twokings.
 
* 2009-06-10 21:22:36 -0700: fixed forteitary wins for special game variations
 feature that corrects forfeitary wins (e.g. on time) of a bare King to draws of course breaks proper operation in losers, suicide and giveaway

* 2009-06-09 21:02:51 -0700: automatically test ini file extension for files from commandline option @filename
 The winboard.c is a patch that automatically tries indirection files on the
command line (@filename type arguments) also with an extension .ini if the name itself
did not match any existing file and contained no period. This saves the user typing the
.ini all the time (which I started to find pretty annoying while working on the installer
package).

* 2009-06-08 15:23:58 -0700: added gamelisttag option to xboard
 also updated documentation.

* 2009-06-08 15:20:00 -0700: updated RTF file for winboard
 
* 2009-06-06 09:12:48 -0700: fixed syntax error in texi file
 
* 2009-06-06 08:57:52 -0700: updated to patchlevel "j"
 
* 2009-06-06 09:00:25 -0700: fixed adjudicating probelm in Shantranj
 Refrain from adjudicating a draw based on insufficient mating material
(except bare King vs bare King) in Shatranj (where in this case you can
still win by baring the opponent's King).

* 2009-06-06 08:55:17 -0700: updated documentation and FAQ
 
* 2009-06-04 22:52:31 -0700: source clean up. Fix compiler warning, removed unused variables, etc.
 
* 2009-06-04 20:43:29 -0700: adding support for different windows compiler
 Eric kindly provided us with some makefiles for different windows compiler and also fixed some issues that came up while using those compilers.

* 2009-06-04 18:27:26 -0700: cleanup: got rid of a lot of files that are not needed for a new release
 lots of documentation was available in several places, old unused source files and not used graphics file were floating around in the tree. This should clean up most of it. No need to keep these files around, especially since we are using version control.

* 2009-05-29 18:56:06 -0700: added the Action-menu items that allow the user to adjudicate ongoing games in Two-Machines mode.
 This was an already existing backend function from Winboard_x; I only had to create a menu item for it to be called. Due to an oversight I had forgotten to do that so far. (Logically, I should have done it in the 4.3.5 release, where I updated the XBoard menus to conform to the WinBoard menus.)

I tested them, and they seem to work. (As expected.)

* 2009-05-28 22:23:00 -0700: updated patchlevel to "i"
 
* 2009-05-28 22:19:08 -0700: small change to mousewheel support; two changes in window behaviour
 * Visible effects during resizing suppressed, bug in chosing new size fixed.
* No longer reserves left and top margins next to main window on first popup.
* built in some hysteresis into mousewheel support, ignorng the first event in any new
direction to suppress effects from accidential mouse movement.

* 2009-05-28 22:11:11 -0700: corrected 'animate dragging'
 I finally figured out how the 'animate dragging' works in XBoard,
so that I could add code to do the unstacking of multiple pieces
in the holdings, rather than always leaving an empty square while
you drag them. (And then suddenly re-growing the remaining pieces
as you make the drop.)  It does not look as fancy as in WinBoard
yet, because the counts still only decrease when you release the
piece, rather than when you lift it. But this is not nearly as
disturbing as leaving empty holdings (if only temporarily) when
you still have pieces there.

* 2009-05-28 22:06:15 -0700: Added comand-line options -debugFile (as synonym for -nameOfDebugFile, as WinBoard also has these synonyms), and -engineDebugOutput (wich already existed in the backend).
 
* 2009-05-28 21:57:53 -0700: updated documentation for new release
 
* 2009-05-28 21:55:51 -0700: added Xiangqi perpetual-chase detection
 
* 2009-05-28 00:06:53 -0700: Updated all files to GPL version 3.
 
* 2009-05-28 00:03:32 -0700: converted two files from dos to unix format
 
* 2009-05-27 23:12:24 -0700: updated INSTALL, NEWS and other files for new release
 cleaned up a bit and updated some information in the text files.

* 2009-05-07 21:48:23 -0700: winboard/config.h shouldn't be in .gitignore, only ./config.h
 
* 2009-05-07 21:47:37 -0700: bumped patchlevel to 0g
 
* 2009-05-07 21:46:02 -0700: corrected path in makefile
 
* 2009-05-07 21:45:48 -0700: updated rtf
 
* 2009-05-07 21:45:25 -0700: added firstOptions, secondOptions, noGUI, firstXBook, secondXBook
 
* 2009-05-07 21:44:03 -0700: handle promotion case for Shogi
 
* 2009-05-07 21:43:34 -0700: add note about RESULT command for GUI programmers
 
* 2009-05-07 21:42:37 -0700: small changes to book.c
  added code for different board sizes
 added some debuginfo
 handle promotion of pieces

* 2009-05-05 23:07:18 -0700: added a .gitignore file
 
* 2009-05-03 12:41:06 -0700: wrong line ending for Unix... ran dos2unix on both files
 
* 2009-05-02 13:51:50 -0700: fixed texinfo files, added windows help file, removed config.h from git
 texinfo.in is not needed anymore, since it only provided a version string, but this can be created automatically using autotools

deleted config.h from version control, since it is automatically created.
added winboard.hlp file (not sure if needed)

* 2009-04-26 23:34:34 -0700: updated autoconf init file
 * renamed autoconf init file from configure.in to configure.ac
 * added bug-report email address to ac_init
 * replaced ac_output(...) with ac_config_files, ac_config_commands and ac_output
 * 2009-04-26 14:49:03 -0700: parser.c is generated by lex, so it shouldn't be tracked and also parcer.c shouldn't be in Makefile.am 
 * 2009-04-26 14:44:51 -0700: added book.c to Makefile.am 
 * 2009-04-26 13:49:15 -0700: use autoheader to create config.h.in
 no need to track this in git anymore

* 2009-04-26 11:13:26 -0700: fixed autoheader warings, remove depreciated acconfig.h
 
* 2009-04-26 10:30:23 -0700: added a autogen.sh
 
* 2009-05-02 10:55:39 -0700: changes from H.G. Muller; version 4.3.16
 
* 2009-05-02 10:50:10 -0700: changes from H.G. Muller; version 4.3.15
 
* 2009-05-02 10:38:39 -0700: changes from H.G. Muller; version 4.3.14
 
* 2009-05-02 10:13:41 -0700: HGM fixed cygwin compile for winboard
 
* 2009-04-24 16:11:57 -0700: cross compilier working
 
* 2009-04-19 10:00:52 -0700: changes from H.G. Muller; version 4.3.13
 
* 2009-04-19 09:52:47 -0700: changes from H.G. Muller; version 4.3.12
 
* 2009-04-19 09:48:59 -0700: changes from H.G. Muller; version 4.3.8
 
* 2009-04-19 09:44:15 -0700: changes from H.G. Muller; version 4.3.7
 
* 2009-04-19 09:39:19 -0700: changes by H.G. Muller; version 4.3.4
 
* 2009-04-19 09:34:31 -0700: changes from H.G. Muller; version 4.3.2
 
* 2009-04-19 10:26:24 -0700: changes from H.G. Muller
 
* 2009-04-16 23:06:13 -0700: cleaned up the directory a bit. Removed winboard-dm-beta, merged the two Changelogs together
 
* 2009-04-16 14:12:48 -0700: changes from Alessandro Scotti from 20060129
 
* 2009-04-16 14:09:33 -0700: changes from Alessandro Scotti from 20060112
 
* 2009-04-16 14:04:02 -0700: changes from Alessandro Scotti from 20051231
 
* 2009-04-16 13:59:12 -0700: changes from Alessandro Scotti from 20051129
 
* 2009-04-16 13:53:41 -0700: changes from Alessandro Scotti from 20050322
 
* 2009-04-10 04:18:59 +0000: translated a handfull of strings and set the codeset to UTF-8 for the translations.
 
* 2009-03-18 03:20:50 +0000: Translation: added German po file (roughly 50% translated)
 
* 2009-03-01 04:23:54 +0000: fixed bug #22853: changed all .Xdefaults to .Xresources in xboard.texi
 
* 2009-03-01 02:43:43 +0000: fixed bug #25672, patch #6182 by adding DESTDIR support to Makefile.in
 
* 2009-03-01 02:33:26 +0000: fixed a warning message in the configure script
 
* 2009-02-28 22:58:57 +0000: fixed parser.l. Should fix bug #22812
 
* 2009-02-28 22:58:14 +0000: fixed parser.l. Should fix bug #22812
 
* 2009-02-28 22:55:30 +0000: added myself to the AUTHOR file
 
* 2007-06-15 04:26:57 +0000: Initial check-in of NSIS Winboard installer files.
 
* 06/12/2005: bugfix some more possible buffer overflows. 
 
* 06/27/2004: added simple icsEngineAnalyze for xboard

* 06/26/2004: added icsEngineAnalyze for winboard: Analyze observe games 
with a chessprogram on a Internet Chess Server (ICS) 

* 02/12/2004: Bugfix ICS interaction console crashed while chess 
engine starting and user type in some stuff. We wait now to finish 
the engine boot.

* 02/11/2004: Winboard: Added color lines for the ics interaction 
console if zippy used 
 
* 02/09/2004: Added mousewheel support for winboard. Play moves 
backward and forward with the mousewheel. 
 
xboard/WinBoard 4.2.7 -- Fri Nov 28 13:36:00 PST 2003 -- Tim Mann

* 11/26/2003: In WinBoard, setting the -debug flag now logs ICS output
to WinBoard.debug too.  (This isn't needed in xboard because ICS
output already goes to stdout, so it's easy to capture along with the
debug output on stderr.)  This was issue #280 in the ToDo file.

* 11/23/2003: Cleaned up some minor problems in cygwin.mak.  Thanks to
Hans Werner Strube for the problem report.

* 11/23/2003: Updated the default list of chess servers in
winboard/defaults.h.  Deleted two that appear to be dead and added a
new one that is active.

* 11/19/2003: Swapped icon_white.bm and icon_black.bm in XBoard.
Hopefully this will get them the right way around by default on more X
window managers.  I wish I understood this better.

* 11/18/2003: Implemented a simple 32-bit random number generator for
WinBoard, replacing the Windows rand(), which ranges only up to
0x7fff.  Formerly on Windows, only zippy.lines sayings that started in
the first 32767 bytes of the file could ever be used.  This was issue
#505 from the ToDo file.

* 11/16/2003: The xboard man page and info files are now built from a
common set of texinfo source files, and the xboard man page now covers
cmail as well.  Formerly, ever since the texinfo file was first
contributed to the project, it was separate from the man page and the
two files had to be updated in parallel, making for a maintenance
nightmare.  To make combining the files possible, I took the Perl
script "texi2man" from the GNU Units project and added support for a
larger subset of the texinfo markup language.  I then carefully
compared the old man page and old texinfo file to make sure the latest
and clearest words survived into the new combined texinfo file.  It
would be great to merge in WinBoard's help file too, but that's a
larger project.  It would have to be heavily conditionalized to
deal with differences in features between XBoard and WinBoard.

* 11/15/2003: Cleaned up and applied a small patch from Daniel
Mehrmann, to stop overly long PVs in thinking output from causing a
buffer overflow crash in backend.c.
        
* 11/5/2003: When a chess engine sends a "telluser" command (including
the cases where "tellall" or "tellopponent" acts like telluser), the
information now goes into a nonmodel popup that is automatically
dismissed when the user clicks anywhere on the board.  This fix is
more significant on WinBoard, where formerly you got a fully modal
dialog that had to be dismissed by clicking on OK before anything else
could happen.  The bug was issue #406 in the ToDo file.

* 11/1/2003: Removed email addresses from this file to reduce spam
load, as it gets linked to from the Web.

* 11/1/2003: Updated READ_ME and winboard/READ_ME.txt.

* Modified xboard and winboard makefiles to be able to build the
project in the new combined directory structure.

* Applied a small security fix to pxboard, from Martin Maeok.

* Fixed a bug in the game list dialog.  The change in 4.2.6 to opening
games in text mode (meant to avoid getting extra \r's into comments)
caused a new bug in determining seek offsets when parsing a game file
to form the game list.  Also, the change was incomplete; on some code
paths games were still being opened in binary mode.  Thanks to Lenik
Terenin for reporting the offset bug.  I've now gone back to always
using binary mode when reading game files, and I've fixed the comment
bug by adding code to remove \r's from parsed comments explicitly.  I
still use text mode for writing game files so that games written by
WinBoard will have Windows-style line endings.

* Fix for minor bug in WinBoard installer.  If you chose a non-default
destination directory, the default directory was still used for a
couple of unimportant things, namely setting the App Paths registry
keys (unused by WinBoard) and trying to copy the existing WinBoard.ini
file to WinBoard.old.  I inherited this bug from the InstallShield 5
sample template.  Thanks to "l.d." for noticing the incorrect keys.

* Patch from Chris Priest: when two engines are being run through rsh,
avoid reusing the same stderr port for both.  I'm not sure why this
should be needed, but it's harmless, at least.

* Bugfix: The kludge to deal with old engines that give an error
message because they don't know the "st" command was too sloppy; it
could hit on an "st" substring anywhere in an error message.  GNU
Chess 4 is the only engine I know of where the kludge was needed, so I
changed the kludge to match a longer, more GNU-specific string.

* Changed the WinBoard self-extracting installer to choose a unique
temporary subdirectory name within the Windows temp directory instead
of always putting wb-setup directly in temp.  Besides being better
practice, this avoids the need to deltree wb-setup first, which
occasionally alarms someone who thinks that deltree is only used by
trojans to delete all your files.  (Yes, I'm serious.)

* Fixed a bug where FICS-style "wild/0" strings were not parsed as the
proper wild type.  The "/" was not being skipped.

* Fixed a WinBoard bug where temporary files created by game
copy/paste were being created in the root directory of the current
drive instead of the TMP directory and were not always deleted on exit.

* Fixed a small bug in winboard.c's ErrorDialog() function.  Thanks to
"Ron" (no last name given).  The bug might have caused a problem when
pressing OK or Cancel in an error dialog when more than one error
dialog was being displayed, but I'm not really sure.

* Removed a hack that worked around a bug in very old versions of
Crafty, where it would sometimes reply "illegal move" to a "."
command.  The hack kept us from recognizing real illegal move messages
in analyze mode from engines that don't respond to the "." command.
Thanks to Fabien Letouzey for the report.

* Fixed a bug in the xboard version of GetDelayedEvent.  It would
return the most recently scheduled event even if it had already fired.
This caused at least one visible error: in -ics -xreuse mode, starting
the engine again and seeing another "feature done=1" would re-run the
initialization code in InitBackEnd3, including the code that opens the
connection to ICS.  The WinBoard version didn't have this bug.  Thanks
to Bob Hyatt for the bug report.

* The promotion popup for ICC wild 26 (giveaway) now includes King.
Thanks to Fredrik Josefsson for the bug report.

* The first game in a file (or being pasted from the clipboard) can
now start with a bare move -- no PGN tags, no move number "1", etc.
Suggested by Shane Harrelson.  One could imagine taking this farther
and accepting something like "e4 e5 1/2-1/2 d4 d5" as two games, but
I haven't done that.

* Bug fix: xboard didn't really accept a paste of multiple games,
contrary to what I wrote when adding the feature to WinBoard in
4.2.4.  Now it works in both.

* Fixed generic start/end of game messages to work on chess.net, where
they put ratings in parens after the player names.

* Updated config.sub and config.guess to newer versions from
automake-1.4p5-2.  This makes configure work with MacOS X + X11.

* Bugfix: when examining a game fragment on FICS where black plays
first, don't show two "(0:00.000)" times on the initial "none" move.

* If TestLegality is on and one of the engines makes an illegal move
in TwoMachines mode, it forfeits immediately.  Formerly the game would
get stuck at that point and the engine that made the illegal move
would eventually lose on time.

xboard/WinBoard 4.2.6 -- Fri Feb  1 22:26:31 PST 2002
Tim Mann

Another small bugfix release.

* Put in a small change from Alexander Mai to allow xboard to build in
an OS/2 EMX environment.

* Fix WinBoard-only bug where editing a PGN file with comments would
insert more \r characters before the \n after every Load/Save cycle.
For some unknown reason we were reading the files in "rb" mode but
writing them in "a" mode, so we kept the \r characters on reading and
added an extra one on writing.  Fix: change the "rb" to "r".  Thanks
to Joel (last name not given) for the bug report.

* Put the arguments to rsh/remsh in the right order for better
portability.  Thanks to Michael Kalisz.

* FreezeUI is now implemented in WinBoard; it was previously only
implemented in xboard.  This function prevents the user from entering
moves or other commands while we are waiting for an engine to
initialize itself.  Implementing it fixes a bug where the user could
start a game (etc.) during feature negotiation.  One of the symptoms
of this bug was that we could send "new" one or move moves *after* the
first game started!

* Fixed some handle leaks (WinBoard only), including a nasty one that
leaked one handle per command sent to ICS.  Thanks to several folks
who helped characterize the bug and tried to look for leaks.

* We now install and look for the info file in ${prefix}/share/info
instead of ${prefix}/info.  The latter is still the default in
autoconf, but the former is currently used in Red Hat Linux and
probably elsewhere.

xboard/WinBoard 4.2.5 -- Sat Dec 15 11:42:51 PST 2001
Tim Mann

* Changed the bundled GNU Chess 5.02 in the WinBoard package to
include a much smaller book and to use much less hash table space by
default (8 MB transposition table, 0.5 MB pawn evaluation table).  The
large book included with 4.2.4 made the WinBoard download too big for
some people, and the default hash tables were larger than the
available physical memory on some people's computers. I built the
small book by running Crafty's "small.zip" book line collection
through a Perl script that makes it look enough like real PGN games
that the GNU Chess book builder will accept it.  The script and some
minor patches to GNU Chess 5 are available in the WinBoard source
distribution and at http://www.tim-mann.org/gnuchess.html.

* Fixed a bug in the code that deals with engines that complain that
"time" and "otim" are illegal moves.  This was a new bug in 4.2.4.

* Revamped the implementation of ToStart in analysis mode.  Formerly
there was some ill-conceived code that would try to get back to the
start of the game by doing a "new" command (and all the followup
commands that requires) while remaining in analysis mode.  This is
problematic for engine authors to support, though it works with
Crafty.  This code was trying to solve the problem of ToStart being
slow because the engine would try to start an analysis after every
"undo" command all the way back to the start.  But ToEnd already had a
much better solution to this problem, and now ToStart does the same
thing: we exit from analysis mode and enter force mode instead,
make/unmake moves to get to the end/start of the game, and then
reenter analysis mode.

* Formerly, Zippy could try to start playing a game before the engine
was initialized; that is, during the initial timeout for protover 1
engines, or before "feature done=1" for protover 2 engines.  We now
wait until the engine is initialized before connecting to ICS, which
fixes this in the common case.  If you turn off engine reuse for
multiple games (that is, if you give the -xreuse option), there could
still be problems; see item 503 in the ToDo file.  Thanks to Dieter
Buerssner for the bug report.

* Bugfix: All WinBoard versions since the merge with WinBoard Plus
(but not xboard) have had a bug in changing the PeriodicUpdates,
PonderNextMove, and ShowThinking options from the menu.  The bug
mostly just caused changing these options during a game to not take
effect until after the next reset.  For TwoMachines games, though,
changing PonderNextMove just before the game would take effect
immediately for the second engine but not until the next game for the
first engine.  Thanks to Koundinya Veluri for the bug report.

* Bugfix: In 4.2.4, we sometimes got confused when trying to leave and
reenter analyze mode.  One symptom was that Analyze File did not work.
Thanks to Igor Syry for the bug report.

xboard/WinBoard 4.2.4 -- Sun Dec  9 14:56:30 PST 2001
Tim Mann
        
* WinBoard now includes GNU Chess 5.02 in place of GNU Chess 5.00.

* WinBoard's PasteGameFromClipboard will now handle a paste with
multiple games in it, popping up the game list dialog.  xboard
already did this.  Suggested by Robert Gerstman.

* We now use the "clearboard" command on ICC where appropriate.

* Bugfix: don't generate a bogus "variant normal" command to an engine
for the loadable ICC wild types that we internally flag as
VariantLoadable.  Also clarified the message that Zippy sends when
declining such wild games -- it can't deal with the possibility of
loading an arbitrary position into the game.  Thanks to Dieter
Buerssner for reporting the problem.  It would be nice to be able to
actually play these wild types some day; presumably there would have
to be an option to either send "loadgame" with a specified position,
or to deal with an opponent sending it.

* If we get a holdings message from ICS when we thought we were
playing normal chess, we now try asking for a move list so that we can
find out from the header what's really happening.  Formerly we guessed
bughouse in this case.  Response to a comment from Gian-Carlo
Pascutto.

* Fixed two longstanding bugs in the clock code.  These affected only
play between a user and a local engine, or between two local engines,
not ICS play.  (1) The last fractional second used by a player before
moving and virtually pushing his clock was being charged to his
opponent instead of to him.  (2) The time and otim commands were being
sent to an engine before the (buggy) fractional second update was
done.  The second bug was pretty harmless in itself, but the first one
was serious.  Many thanks to Peter Rosendahl for carefully diagnosing
and reporting these bugs!

* Changed WinBoard timestamp key.  Nothing is changed other than the
key and a recompilation.  ICC may phase out the old key soon; if they
do, you'll need to use the timestamp.exe from WinBoard 4.2.4 or later
to connect to chessclub.com.

* Fixed an xboard-only bug where most dialogs would not take keyboard
input unless the mouse was actually over the dialog, even if the
dialog window had focus.  Thanks to Jason Varsoke for the bug report.

* Bugfix; Remember to send time and otim commands to engine before
sending playother.  Thanks to Bob Hyatt for the bug report.  Also
fixed a bug where time and otim were sometimes sent even with feature
time=0.

* Added small patches from Wilkin Ng that are meant to fix a crash in
CopyGameToClipboard and a "can't unlock clipboard memory" error in
PasteTextFromClipboard, under Windows 2000.

* Changed cygwin.mak to not use -mno-cygwin.  In other words, it now
uses cygwin1.dll instead of the mingw libraries.  This works around
bugs in fileno() and stat() that I encountered in mingw.  (fileno
seems to always return 0.  stat fails because the library routine uses
a different definition of struct stat than the header files provide.)
Thanks to Robert Gerstman for reporting the symptoms these bugs caused
when WinBoard was compiled with Cygwin.

* Added -showButtonBar option.  This lets you delete the << < P > >>
buttons and thus widen the message widget a bit.

* Added code to address a very obscure bug.  If an engine dies and
needs to be restarted while in one of the analyze modes, it was not
getting put back into analyze mode.  This bug should not really ever
have gotten tickled unless reuse was turned off and the engine sends
game end commands (such as 1-0) when a game ends by rule while in
analyze mode.  The next version of the protocol spec will say that
engines should not do that, but version 2 and earlier didn't address
the issue.

* We now use the FICS "iset ms 1" feature to get times in ms instead
of seconds.  Thanks to DAV for email telling me about this feature.

* Changed clock display to show tenths of seconds for 9.9 seconds and
less, instead of 0.9 seconds and less as before.

* It has been discovered that Zippy can play simuls on ICC (but not on
FICS).  If you arrange for Zippy to send the ICC command "simulize" in
the -zippyGameStart string, it will accept additional games while
playing.  Zippy will use the same engine for every game, so whenever
it switches opponents, the engine's state will be reset with the "new"
command.  This will of course weaken its play, so don't enable simuls
if you want your engine to have the highest possible rating.  Zippy
was never designed to work with simuls; it just works by accident, and
it hasn't been tested much.  So please report any bugs you notice, but
don't expect them to be fixed rapidly.  Thanks to Paul McGuire for
noticing that this works.
        
* Bugfix: Suppress the direct command to engine popup if there is no
engine.  In WinBoard this case used to cause a crash.  Thanks to
"Dargon" for the bug report.

* Bugfix: formerly we did part of the engine initialization for the
very first game too soon, before feature negotiation.  One result of
this is that engines would never get the "ics" command for the first
game.

* Lengthened timeout to detect protocol version 1 chess engines to 10
seconds.  This should reduce problems with protocol version 2 engines
that are slow to initialize missing the timeout and not getting a
chance to send their feature commands before the first game starts.

* Klaus Friedel says that adding a short sleep to WinBoard after
starting a new chess engine solves a problem that occurs under Windows
2000, in which engines sometimes don't see the initial command(s) from
WinBoard and hang.  I don't understand how that can happen, but the
sleep is harmless, so I've put it in.  Others have also reported what
may be the same problem, so hopefully this will fix it for them too.

* Bugfix: editing the Result field in EditTags will no longer cause a
later crash.  Thanks to DAV for a clear bug report that let me
reproduce the problem.

* ICC wild 28 is now recognized as shatranj, but it is not supported.
You might be able to play it by turning off Test Legality.

* Formerly we would always turn off "feature time" if the engine
printed an error message with the string "time" or "otim" in it; now
we do that only if the message comes before the engine makes its first
move.  This change is useful because some engines spew a lot of bogus
error messages about commands they don't fully parse, so something
like "result 0-1 {White lost on time}" could generate such a message.
Engines shouldn't do that, but if they do, mysteriously turning off
"feature time" is not a good way to react.

* Fixed an infrequently occurring Zippy bug: formerly, resuming from
adjournment by position (that is, with GetMoveList turned off) or
starting/resuming a wild game (nonstandard starting position) would
try to set up the position without putting the engine into force mode
first.  That was probably tolerated by most engines, but if the
position was black-to-play and the engine hadn't set feature
setboard=1, then the fake "a2a3" move (which we use as a kludge to get
black to be on move before sending the edit command) would appear to
be a real move, and the engine might reply to it!  This bug could also
be seen when trying to use Zippy to play a simul -- something that
Zippy was never designed to support, but which seems to mostly work
anyway.  Thanks to Paul McGuire for sending a WinBoard.debug file that
showed the bug occurring.  Some other cases of resuming adjournments
or starting games from nonstandard positions may have been broken too;
I think I've straightened out the problems.

* Agreeing to a draw in the human vs. local engine case formerly did
not work if the human offered first; now it does.  A human's
unsolicited offer is considered valid until he makes another move.
Thanks to Bruce Moreland for the bug report.

* Bugfix: We used to look for just "fr" in Event tags or strings from
ICSes to recognize Fischer Random games.  This caused a lot of false
matches.  Now we look for "wild/fr" instead, which is what FICS uses.

* Bugfix: Clicking on a move in the MoveList window only updated the
displayed position; if a chess engine was active, its state was not
updated.  Thanks to Alejandro Dubrovsky for the bug report.
        
* Bugfix: The "resign" command from engine to xboard formerly did not
work in ICS (Zippy) mode.

* Bugfix: Formerly we would register a garbage premove if the user
clicked on a piece, then clicked on an edge or outside the board.  Now
such an errant click is ignored.

* Bugfix: GameEnds would send "exit" to an engine in analyze mode even
if we did not actually want to leave analyze mode.

* Bugfix: "feature pause" was being rejected.  We don't currently use
it, but that is no reason to reject it.  Thanks to Gian-Carlo Pascutto
for the bug report.

* Bugfix: when "partner" needed to be sent both to ICS and the engine,
Zippy was erroneously sending the ICS prefix character to the engine
too.  Thanks to Gian-Carlo Pascutto for the bug report.

* Zippy now declines challenges when the engine is not yet ready to
play again, instead of ignoring them.  This is needed so that the
server will let the opponent repeat the challenge.  It might be better
to remember such challenges and accept them when the engine is ready,
but that would be a bit more complicated to implement.

* Bugfix: Zippy's emotes didn't work on FICS because there, "i" is a
built-in alias for "it", not a real command.  So sending "$i foo" gave
an error message instead of doing "it foo".  We now leave out the
alias-suppressing prefix for emotes.

xboard/WinBoard 4.2.3 -- Mon Feb 19 19:55:05 PST 2001
Tim Mann

* Bugfix: The error message "Variant X not supported by gnuchessx"
formerly could pop up even in -ncp mode.

* Fixed an xboard bug in premove highlighting.  If you had
HighlightLastMove turned on, premoves got highlighted in
highlightSquareColor (yellow) instead of premoveHighlightColor (red).

* Modified premove again.  Now we are back to displaying the move
locally immediately after sending it to the server, as in versions
prior to 4.2.1.  This is good because it lets you register your next
premove sooner.  As a better fix to the problem that the change in
4.2.1 had been trying to address, we now suppress animating the
opponent's move if you have a premove reply pending.  However, it's
probably still a good idea for you to turn off AnimateMoves entirely
if you are trying to play extremely fast games.
        
* Removed an erroneous patch that left zombie chess engines around
when both sigterm and reuse were turned off.  (Bug was in xboard only,
not winboard.)

* Bugfix: loading a game with autostepping did not work for games with
PGN result "*" (unfinished).  This bug was introduced in 4.1.0, caused
by some outdated code that was supposed to leave you in EditGame mode
after loading an unfinished game.  I fixed this, and also changed it
to leave you in EditGame mode after loading *any* game, but without
changing the tags to say it's an edited game.  It's not really clear
whether that is the best thing is to do here -- ideally, perhaps, the
tags should change to say "edited game" if you actually make any
changes, but that is harder to make work.

xboard/WinBoard 4.2.2 -- Tue Feb  6 20:00:00 PST 2001
Tim Mann

* Fixed a problem introduced in 4.2.1.  In -zippyPlay mode, 4.2.1
started sending the initString immediately after the old game ended,
but then when the next game started, it would still send "force" and
another initString.  Another problem was that the first initString put
the engine out of sync with the displayed board position, though that
might not usually have been noticeable.  This is now changed to do a
full Reset at the end of the previous game, and avoid doing a Reset at
the start of a new game if we are still in BeginningOfGame mode.

* Fixed some limitations in -zippyPlay mode.  The engine was not being
fed the moves or history of games that were displayed but not being
played, so it would get confused if you tried to do something like
examine a game, then enter EditGame mode from somewhere in the middle
of the game.

* Made a small change so that cygwin can compile xboard for Windows
(an X server is required to run it).  Thanks to Volker Zell.  Most
people would probably prefer to run WinBoard on Windows, however; note
that cygwin could already compile WinBoard.

* Added some missing documentation to engine-intf.html: It is okay to
send "feature done=0" even before you receive the xboard and protover
commands, if this is needed to give your engine enough time to
initialize.  See engine-intf.html for more discussion.

* Bugfix: "feature done=0" did not work for the second engine; now it does.

xboard/WinBoard 4.2.1 -- Sat Feb  3 19:52:26 PST 2001
Tim Mann

* Bugfix: in Zippy mode with feature san=1, the elapsed time was being
sent to the engine after the SAN move; for example, "e4 (0:01)" was
sent instead of just "e4".

* It seems that scroll wheel mice with incompletely installed software
(under Windows 95, at least) can send repeated middle button up events
when the wheel is turned.  Moved QuickPaste from middle button up to
middle button down to avoid problems with this.

* A premove is now sent *before* the opponent's move that it replies
to is animated.  The premove itself is no longer displayed (or
animated) when sent; instead, it is displayed when the chess server
echoes it back as a board update.

* Bug fix: in -xreuse mode, xboard was only sending command line
arguments to the engine the first time it was started.  (WinBoard
didn't have this bug.)

* In -zippyPlay mode, we now initialize the chess engine for a new
game immediately after the old game ends.  (In -xreuse mode, this
means starting a new engine process; in the default -reuse mode, it
just means sending "new".)  If the engine supports ping, we don't
accept challenges until the "new" has finished.  This should help
avoid losing time (re)initializing the engine after the game starts.

* We now handle the new FICS field in style 12 that says whether the
clocks are ticking.

* whiteFlag and blackFlag are now reset whenever an ICS board image
arrives, giving autoflag a fresh chance to work.

* Suppressed "geometry error" debug messages unless -debug flag is given.

* Bugfix: avoid core dump on Solaris with -debug flag, caused by
passing NULL to a %s format in GameEnds.

* Added -firstProtocolVersion and -secondProtocolVersion.  This will
allow use of extremely broken engines that hang or die when given the
"protover 2" command.  It should be rarely if ever needed.

* Zippy now strips highlights from player names in the Creating
message; previously the ratings would not be extracted from this
message properly if you had set the highlight variable.

* Bug fix: Zippy with -zp but not -zt used to be fooled by false
partner tells inside channel tells; e.g.: 
Garf(24): Garf (your partner) tells you: sit

* Bug fix: when an engine reported its name with the new "feature
myname=" command, we had been putting just that name into the window
title even when in TwoMachines mode.

* By default, WinBoard no longer uses the new feature introduced in
4.2.0beta of keeping the ICS menu, ICS server list, and lists of chess
engine command lines in separate files; instead, the lists are back in
the winboard.ini file.  This gets rid of some problems where the
separate files could not be opened because the working directory was
not set to WinBoard's installation directory.

* WinBoard command lines and ini files now have another optional
syntax for quoting an option value.  All characters within { } curly
braces are interpreted literally except for '}' itself.  SaveSettings
now uses this syntax for string values that contain a backslash or
newline and do not contain a '}'.  This should address the problem
that the separate .ini files had been meant to fix; engine command
line lists can now be written without \-escapes.
        
* Fixed three bugs in relaying moves between engines in TwoMachines
mode: an engine would be sent SAN if the *other* engine had set
feature san=1, a relayed SAN move did not end with a newline, and
feature usermove was not implemented for relayed moves.

* Fixed several bugs in exiting: File/Exit would sometimes hang;
timestamp/timeseal would not get killed off; an engine crash would
cause error messages to be printed recursively until stack overflow.

* Fixed a bug in the code to detect which ICS is in use and adapt to
it; the chess.net case was broken.

* Removed the leading alias-suppression character when sending a move
to ICS, because it seemed to break accuclock on chess.net.

* Fixed a bug where a FICS prompt like "10:01_fics%" would be matched
as finger note number 10, causing a problem detecting when the user is
logged in.

* Added tellicsnoalias command to protocol.

* Changed st command in protocol back to old behavior.

* Added done=0 feature to protocol, to lengthen initial timeout.

* Added missing -colorSeek command line option to xboard.  (Was
present in WinBoard.)

* Bug fix: setting feature reuse=0 was not working.

xboard/WinBoard 4.2.0beta -- Sat Dec 16 16:34:56 PST 2000
Tim Mann

* Setting -searchTime no longer turns off -clockMode.  Note that when
-searchTime is set, some engines may search for the lesser of
searchTime and the amount of time their normal clock management would
dictate, while others will disable their normal clock management and
always search for exactly searchTime.

* Fixed WinBoard bug where on Windows 2000 (and maybe Windows 98) the
ICS Interaction window would scroll back to the top whenever it filled
to capacity and WinBoard trimmed some text off the top.

* Changed xboard dragging so that the center of the dragged piece is
forced to be over the mouse cursor.  (WinBoard dragging already worked
that way.)  This should avoid confusing cases when you start a drag
with the mouse near the edge of a square -- formerly most or all of
the piece could be over one square but the mouse cursor over another.
Suggested by DAV.

* Added move list window for xboard.  Code contributed by Manuel
Hoelss and re-hacked a bit by me.  Performance of this window is poor,
because the implementation regenerates the entire window contents
whenever anything changes.  It's hard to avoid this when using the
Athena List widget, though we could be a little smarter and notice
when the current move number has changed but the move list has not.
There are also several drawing problems caused by bugs in the Athena
Form and List widgets.  The window should not really allow a
horizontal scroll bar, but the bugs are much worse if I turn it off.
In both Xaw and Xaw3d, the List widget tries to resize itself when
XawListChange is called, even if you tell it not to; and what's worse,
if the widget cannot resize itself to be large enough to avoid
clipping any of the list item, it does not change the list!  Arrgh.
On top of that, the Xaw Form widget is very buggy in XFree86 4.0.1
(X11R6.4.3).

* The OK button in the WinBoard startup dialog is now simply disabled
if none of the radio buttons is selected, instead of bringing up an
error dialog.  Suggested by Ted Milbaugh.
        
* Used ping to address the worst of the race conditions in the
protocol.  Now if an engine supports ping, we ping it after sending
the "new" command and after trying to place it in force mode at the
end of a game.  When a ping has been sent and the corresponding pong
has not yet come back, we (1) ignore or undo any moves the engine
sends, (2) ignore game end messages (such as "offer draw", etc.) from
the engine, (3) ignore thinking output from the engine, (4) delay
starting the next game if in -matchGames mode, (5) in -zippyPlay mode,
respond to challenges with a polite message to try again soon instead
of accepting them.  There are a few more things that should be done
(and item 5 might not be quite what we want), but it's hard to fit
them all into the existing code.
        
* Patched the version of GNU Chess 5.00 included with WinBoard to
support protover, feature, setboard, and ping, and to fix some minor
bugs.  See gnu500+.patch.

* Changed the GNU Chess 4 kludges for missing "st" and "sd" commands
to be invoked only if the engine gives an error message for "st" or
"sd", and to always use protocol commands instead of engine
command-line options.  See "Idioms" in engine-intf.html.

* Added new protocol command "protover 2" to tell the engine what
version of the protocol is in use.  Version 2 is still compatible with
old engines; see engine-intf.html.

* Add new protocol command "feature ..."  The engine can send this in
response to the protover command command, to say what extended
protocol features it would like to enable, etc.  xboard responds
"accepted F" or "rejected F" for each feature F that is set.  See
engine-intf.html.

* Added new protocol command "setboard <fen>".  This is an alternative
to "edit", used only if the engine says "feature setboard=1".

* Added new protocol command "ping <msg>".  The engine is supposed to
respond "pong <msg>".  This gives us a tool to fix some race
conditions in the protocol, but it is not used to the fullest yet.

* Added new protocol command "tellopponent".  This will do a "say" if
you're on a chess server in Zippy mode, or pop up an information
dialog otherwise.

* Added new protocol command "tellall".  This will do a "kibitz" if
you're on a chess server in Zippy mode, or pop up an information
dialog otherwise.

* Added new protocol command "tellothers".  This will do a "whisper" if
you're on a chess server in Zippy mode, or do nothing otherwise.

* Changed the WinBoard /icsNames, /icsMenu, /firstChessProgramNames,
and /secondChessProgramNames options to accept a filename preceded by
"@", meaning to find the value in the file.  Changed the defaults to
use this feature, with filenames icsnames.ini, icsmenu.ini, fcp.ini,
and scp.ini.  This should make it easier for users to edit the values,
since one less level of quoting is needed, and each value is in its
own file, not mixed into winboard.ini.  Those who upgrade from an
older winboard version and already have a winboard.ini file will not
automatically have these values moved to a file; that has to be done
by hand if desired.

* We now strip .exe from engine names for "name" command, tags, etc.

* Allowed WinBoard input widget to hold more than one line.  You can get
a newline into the widget with copy/paste or with Ctrl+Enter.

* Fixed an xboard bug where the [P] button would appear in the
opposite highlight state from what it should be after being pressed.

* We now use the "/" or "$" alias suppression feature on the chess
servers, to avoid problems if the user inadvertently aliases a
needed command to something else.

* Changed the method used to deiconize xboard in AutoRaise and cmail.
This may fix bugs under some window managers where after
deiconization, the board window was mapped but nothing was drawn
inside.

* Preliminary, partial support for playing several chess variants
against a local engine or editing variant games.  The -variant option
must be given on the command line or WinBoard startup dialog; there is
no menu.  The current variant is not displayed (except in the
Tags/EditTags window).  Many variants will not work quite right even
if the engine supports them; see item 326 in the ToDo file for
details.

* Added a workaround to detect the variant type when examining a
scratch wild game on an ICS.

* Fixed two cases where "variant" command should be sent to the engine
when necessary but was not: loading a game from a file and restarting
a dead engine.  Thanks to Gian-Carlo Pascutto.

* Added missing support for FICS games that start from a bsetup position
with Black moving first.

* Fixed a bug in loading a file from the game list window while in
Analysis or AnalyzeFile mode.  This now leaves us in AnalyzeFile
mode, analyzing the new file.

* Fixed a bug in loading games with a comment before the first move.

* Fixed a bug in handling LoadGame (or -lgf) with -td 0.

* Added a command line option to select an initial minor mode from the
Mode menu.  This should satisfy requests from a couple of users with
special applications.

* Fixed a problem where some things in WinBoard's General Options
dialog box were grayed out at the wrong time.

* We look for a few messages from the chess engine pipe that are
intended to catch problems starting an engine via rsh, such as "No
such file".  Moved code so that these strings won't match until other
protocol messages have been looked for, and documented the messages in
engine-intf.html.

* Fixed the WinBoard bug that sometimes made the console input box
white-on-white even when not in password-entry mode.

* Small changes to WinBoard to be compilable with the latest Cygwin
net release (as of 10-16-2000).  Thanks to Mark Schoenberg.

* Recompiled GNU Chess 5.0 for WinBoard with the latest Cygwin net
release (as of 10-16-2000).  Also rebuilt the book just to be sure,
since I had one bug report about it.  It came out identical and seems
to work fine.

* Added messages for some common InstallShield errors instead of just
printing InstallShield's stupid error code numbers.

* Added -firstComputerString and -secondComputerString, allowing you
to suppress or change the command that is sent to a chess engine when
its opponent is another computer.

* Added "configure --with-Xaw3d" to allow use of Xaw3d widgets instead
of Xaw widgets in xboard.  Thanks to Johnny C. Lam of the NetBSD
project.  I've left Xaw as the default because Xaw3d is too ugly.
        
* The version of the Xaw Form widget released in XFree86 4.0.1
(X11R6.4.3) has a bug that causes xboard to display the button bar
on top of the message widget.  I've reported the bug and have
installed a kludge workaround into xboard.

xboard/WinBoard 4.1.0 -- Sun Sep 17 17:30:14 PDT 2000
Tim Mann

* Merged Winboard Plus 4.0.8 from Mark Williams into xboard
  source pool.  Status of Winboard Plus features in xboard:

  - White pieces now have borders, but the implementation is
    completely different from WinBoard's; see below.

  - Options menu not changed, probably won't be.

  - Copy&paste support working, using Mark's changes to backend.c, Ben
    Nye's changes to xboard.c for FEN positions, and some code of my
    own in xboard.c for PGN games.

  - Premove hooked up and working.  No pre-first-move, though.

  - ICS Alarm hooked up and working.

  - Auto-flip working.

  - PGN Training mode hooked up and working.

  - Improved menu item enable/disable working.

  - Improved >> button working.

  - Auto activate board working.  I changed this to be optional; you
    might not always want it, especially in Zippy mode.

  - Blindfold working.

  - Documentation updated.

* Fixed more colorization bugs.  Commands like "message foo tells you:
xxx" and "tell 33 tells you: xxx" will no longer generate bogus
colorization.

* Added low-tech sound support to xboard, by invoking an external
program on a filename for each sound to be played.
        
* ICC wild 27 is now recognized as VariantAtomic, and it sets the
"ignore check" flag, but is not otherwise supported.

* Converted the piece bitmaps from .bm to .xpm for use with xboard's
xpm support.  (An .xim version was not created.)  Wrote a shell script
using ImageMagick to do the conversion and to bucket fill from
coordinate 0,0 with the background color.  Used xpaint by hand to
clean up a few pieces where background color "leaked" into the inside
or didn't flow all the way around the outside, then a sed script to
fix up color names and bitmap names in the xpaint output.  If any more
sizes are added later, they should be straightforward to convert with
reference to the existing scripts.  With this conversion, not only do
the white pieces now have dark borders, but (unlike with WinBoard) the
light details on the black pieces are opaque.

* If the XPM library is found when building xboard, the new xpm pieces
are compiled in as the default.  External bitmap (or xpm) pieces can
still be used by giving the -bitmapDirectory (or -xpmDirectory)
option.  If the XPM library is not found, the old bitmap pieces are
the default.  In this case external bitmap (or xim) pieces can be used
by giving the -bitmapDirectory (or -xpmDirectory) option.  There is no
way to compile in xim pieces.

Winboard Plus 4.0.8 -- Sat Aug 5 15:51 PDT 2000 
Mark Williams

* Bugfix release.

* Main menu now visible in Win NT.  Thanks to Microsoft for creating a 
compiler which auto-mutilates the WinBoard.rc file so that the menu is
displayed under Win 98 but not Win NT.  I fixed the problem by hand editing the
.rc file.  Sigh.

* "Reset game" now works properly when invoked from Analysis mode when two
engines are present. Correction of bug introduced in Winboard Plus. Thanks to
Mogens Larsen for reporting the bug.

* Pasting a FEN position with Black to move now works in all modes.
Thanks to Mogens Larsen for reporting the bug.

* Added gnuchess.dat to the distribution and a couple of text files that were
missing in 4.0.7.

*** Merged in the following changes from Tim Mann:

* Bugfix: Highlights would sometimes not be taken down, at least in
xboard.  The problem is that SetHighlight works in chess coordinates,
not view coordinates, so if flipView has been toggled since the
highlight was put up, it gets taken down from the wrong square.  Fixed
by being careful to call ClearHighlights before toggling flipView, not
after.  Ugh.

* Bugfix: HighlightLastMove would sometimes highlight square a1 after
a FICS "tomove black" command.

* Bugfix: a move list with initial position that has black to play was
not being handled correctly.

* Removed "decline abort" and "decline adjourn" commands from Zippy.
The syntax is changing on FICS, and Zippy really doesn't have to
explicitly decline the requests anyway; it's sufficient to not accept
them.

* ICC wild 26 is now recognized as VariantSuicide (the FICS name; ICC
calls it "giveaway").

* Bugfix: PGN games with black to move in the initial position weren't
being legality checked correctly.  I think this was introduced in
4.0.6 when I fixed a different PGN bug.  Thanks to Mark Williams for
spotting the bug and suggesting a fix.
        
* Bugfix: the initial "name(" of a kibitz wasn't being colorized.

Winboard Plus 4.0.7 -- Thu May 12 17:57 PDT 2000 
Mark Williams

* New white pieces with black borders.

* Major overhaul to Options menu. New dialogs for setting General Options,
  Board Options, ICS Options, Sounds, Fonts, Colors.

* Clipboard support for PGN game scores and FEN positions.

* Optional ICS Premove with user-specified first moves for White and Black.

* New sound events: ICS Win, ICS Loss, ICS Draw, ICS Unfinished, ICS Alarm.

* New auto-flip option when playing against a chess program.

* Added PGN Training mode.

* Certain menu options now disabled while program is thinking.

* ">>" button now always goes to the end of the game.

* In ICS mode, the board is automatically activated at the start of a
new game.

4.0.7 -- Sun Mar  5 17:17:49 PST 2000  Tim Mann

* Fixed a nasty bug in parsing ICS game histories that was introduced
in 4.0.6.
        
4.0.6 -- Fri Mar  3 16:20:11 PST 2000  Tim Mann

* Fixed bugs in Zippy's code for responding to messages sent with the
ICS "message" command.
        
* Fixed bugs where xboard could still create dialogs partially off the
top of the screen.  Fixed unwanted interactions between initial sizes
and positions of various unrelated xboard dialogs.  Deleted
borderXoffset and borderYoffset resources.

* Fixed restoring the xterm name on exit to work in more cases.
        
* Small fix to the WinBoard self-extracting installer: when it starts
up, it now always deletes any old wb-setup directory left over from a
previous installation.  Formerly this worked only on Windows 95/98, so
on NT the self-extractor would sometimes stop and ask the user whether
it was OK to overwrite the old wb-setup files.

* Modified WinBoard to be compilable with the free Cygwin tools
available from http://sourceware.cygnus.com/cygwin/.  Based on work by
Mark Schoenberg.  (His mods were enough to get WinBoard through the
compiler, but a few things were broken along the way, so I reworked them.)
        
* Merged changes from Don Fong to make WinBoard compilable with
Borland C++ 4.5.  I do not have Borland C++ and am unable to test with
it, but I did make sure the changes don't stop it working with MSVC++
5.0 and Cygwin.

* Forbade entering TwoMachines mode from MachineWhite or MachineBlack
mode when machine is on move.  This used to confuse the engine and
cause problems, because when we try to get the engine to stop, it
might make a move first, and it might not do that immediately.
        
* Variant classification now deals with new ICC wild 25, classifying
it as "3check".  Also, we are now conservative about new wild numbers
we don't know about, classifying them as "unknown" instead of
"normal".

* Replaced bad GNU Chess 5.00 book in WinBoard distribution with a
correct one.  1.e4 now no longer takes GNU Chess 5.00 out of book!
Thanks to Pete Galati.

* Made move parser accept and ignore nonstandard ep suffix on PGN moves,
e.g., "exf6ep" or "exf6e.p."

* Fixed bugs in parsing PGN files starting with no move number, or
with a move number other than "1." (including "1...").  Thanks to
Michael Soulier and Stefan Zipproth for example PGN files
demonstrating the bugs.
        
* Changed InitComboStringsFromString to not modify the input string;
this caused an exception when WinBoard was compiled with MSVC++ 6.
Thanks to Bert Tuyt for the report.

* In WinBoard, dragging a piece off the edge of the board without
releasing the mouse button no longer instantly aborts the move.  You
still must release the mouse button inside a square to complete a
move, but you can now drag off the edge and come back inside.
        
* WinBoard bugfix: If you went into EditGame or EditPosition mode
while observing or playing an ICS game, and a new move came in
(snapping you back to ICSClient mode) while you were dragging a piece
around, the drag state would not be reset, causing the piece you were
dragging to remain on the board wherever you dropped it.  Thanks to
David Brinegar for the bug report.

* WinBoard bugfix: If you moved or resized the board while your
opponent was moving, and you had AnimateMoving on, his move would not
be displayed until after your next move.  Thanks to DAV for reporting
the bug repeatedly until I finally understood it.

* Added a Zippy "farewell" feature; see zippy.README.
        
* Fixed a bug (xboard only) in processing -zippyReplayTimeout on the
command line.  Thanks to Steve Beer for the report and patch.

* Send "computer" command to both sides in TwoMachines matches.
        
* Fixed a Y2K bug in Evan's cmail code!  Thanks to Brian Mays for the
report and patch.
        
* Bugfix: Initialize variant field of GameInfo struct in gameinfo.c.
This could cause crashes in LoadGame.  Thanks to Andrzej Nagorko.
        
* Fixed several pattern matching problems: Finger notes and formula
vars no longer get replied to by zippyTalk or colorized.  Channel
tells by players with many titles are now colorized fully, and
zippyTalk handles channel tells by such players correctly.  The shout
"--> foo(99): bar" no longer gets colorized as a channel tell.  Code
is cleaned up a bit, and autocomment capture now uses the same
criteria as colorization.

* xboard fix: Enter key now closes error popups as in WinBoard.
Especially nice for the "Exiting: Connection closed by ICS" popup.

4.0.5 -- Tue Dec  7 10:30:40 PST 1999  Tim Mann

* Added missing documentation of SIGTERM usage into engine-intf.html.

* Bugfix: New -zippyReplayTimeout option was misspelled in WinBoard as
-zippyReplyTimeout.  Thanks to Francesco Di Tolla.
        
* xboard -ics now restores the xterm's original title upon exit, at
least if $WINDOWID is set.

* Bugfix: -matchGames mode wasn't terminating properly.  Final score
popup would come up twice and engines would not be killed off.  Thanks
to Frank Quisinsky for the bug report.  Bug was introduced in 4.0.4.

* Bugfix: "tell 50 foo shouts: bar" no longer gets colored as a shout;
similarly for "...s-shouts:" and "...c-shouts:".  Thanks to David Lee
for the bug report.

4.0.4 -- Fri Dec  3 17:51:27 PST 1999  Tim Mann

* Bug fix: formerly if you invoked SaveGame from EditPosition mode
with black to play, the position was not saved properly.  As a quick
fix, SaveGame now takes you out of EditPosition mode before saving.

* Changed -matchGames to keep track of the score (won-lost-drawn) and
display it in the banner.  The final score is displayed in a modal
popup before the program exits.

* Changed "Connection closed by ICS" to display in a model popup
before the program exits.

* Added -popupExitMessage option, default on.  Setting this to false
suppresses the modal popups that you sometimes get just before
xboard/winboard exits -- both the two new ones listed above and the
existing Fatal Error popup.  This is useful when running Zippy
unattended from a shell script (or .bat file) that loops and starts a
new copy after a time delay when there is an error.

* In xboard, added missing support for WM_DELETE_WINDOW to all
windows.  This makes xboard a better citizen, and is much needed with
the newer X window managers that put an [X] button on every window and
bind it to "kill application" if the applicationd does not support
"delete window".  We can't allow xboard to be thoughtlessly killed,
since that leaves the chess engine running in the background.

* We now avoid positioning tops of xboard dialog boxes offscreen.

* zippy.lines can now use the caret ('^') character as an inter-saying
separator. You no longer need to dig up a text editor that can put NUL
('\000') characters in text files.

* Bugfix: In traditional chess clock mode, White now gets time added
to its clock as soon as it makes time control.  Previously both White
and Black got their time added only when Black made time control,
which would cause some engines to make their 41st move as White too
quickly, thinking they were low on time.

* The zippyAcceptOnly feature no longer sends a decline command or
a tell.  Thus you can use it to put Zippy into a manual accept mode.

* Bugfix: "tell 50 foo tells you: bar" no longer gets colored as a
personal tell; similarly for "...whispers:" and "...kibitzes:".

* WinBoard no longer blanks the message line when you start a new
move, unless there was an error message there from the previous move.

* WinBoard now never puts the white piece drop menu on the middle
mouse button.  Both colors are always on the right button.  Windows
users often don't think to look for a middle button menu, or don't
really have a working middle button despite what Windows says.

* Guest login handles on ICC are now parsed again; an ICC wording
change had broken this feature.
        
* Autodetect when the engine does not support the "draw" command and
don't relay the command to it in that case.  Needed because sending an
unsupported command to GNU Chess makes it move immediately, so people
could make Zippy move without thinking by continually offering a draw.
Thanks to Frank Walker for the problem report.

* Small changes based on code from Robert Jurjevic: The WinBoard
analysis window can be minimized (though not to the taskbar).  The
WinBoard ICS Interaction window will autoscroll if you drag the mouse
above or below the window while selecting, and password entry mode
leaves the entry line as the background color instead of using 75%
gray.  Seeks are now colorized with a new color of their own.

* Added -zippyMaxGames feature.
        
* -zippyNoplayCrafty command line option was missing in xboard; fixed.
Thanks to John Perry.

* Fixed bug in detecting when Zippy loses his bughouse partner.
Thanks to Ben Dean-Kawamura.

* allobs fix in 4.0.3 didn't work; corrected it.

* WinBoard will now update the board immediately if a move is made
while the window is being moved or resized.
        
4.0.3 -- Sun Aug 15 18:44:39 PDT 1999  Tim Mann

* Bugfix: moved winboard wizard functions that send a command directly
to the chess engine from Ctrl+Alt+1 and Ctrl+Alt+2 to Alt+1 and Alt+2.
Ctrl+Alt is equivalent to the European AltGr key, so the old assignments
conflicted with typing an @-sign on some keyboards.
        
* Bugfix: prevent "If this message stays up, your chess program does
not support analysis" from appearing if AnalysisMode is selected when
already in Analysis or AnalyzeFile mode.  Also, deleting analysis window
in WinBoard now switches you to EditGame mode, to avoid the anomaly of
being in an analysis mode with no analysis window present.
        
* Change all Zippy environment variables to work as command line
options.  For now they still work as environment variables too, but
the environment variables are deprecated.

* Use unsigned long to hold node counts from engines.

* Print error instead of crashing on attempts to use xpm or xim pieces
in monochrome mode.  Thanks to Jim Torrance for the bug report.

* Tightened pattern recognition for ICS messages sent to Zippy; should
no longer match ICC allobs output.  Fixed recognition of when Zippy is
sent a message; small change to FICS output had broken it.
        
* Zippy now exits if the engine dies unexpectedly.

* Added Zippy "spoofedby" feature; see zippy.README.
        
* Narrowed the margins on some older WinBoard dialogs to make their
style uniform with the newer ones.

* Bugfix: Changing Ponder Next Move from the Options menu did not take
effect until the next game.  This bug was in WinBoard only.  Thanks to
Stefan Zipproth for the bug report.

* Got rid of "Internal error: bad move type" message when the engine
makes a move that we can parse but think is illegal.  Now the move is
just passed on (which also may be the wrong thing to do).

4.0.2 -- Thu Feb 25 19:55:32 PST 1999  Tim Mann

* Added one more built-in sound choice to WinBoard: move.wav from Pete
Galati.
        
* Added new larger piece sizes: big, huge, giant, colossal, titanic.
        
* Modified the xboard font searcher to be able to use scalable fonts.
The new -fontSizeTolerance flag controls how closely a nonscalable
font must match to be used in preference to a scalable font.
        
* Zippy environment variables ZIPPYABORT and ZIPPYADJOURN now control
whether Zippy will accept abort and adjourn requests.

* We now determine the variant type being played on ICS, send it as a
command "variant VARTYPE" to the chess engine (if in Zippy mode), and
put it in the PGN tags.  See engine-intf.txt for variant names.  See
ZIPPYVARIANTS in zippy.README to control what variants Zippy will
accept.  Note that this now applies to bughouse, too.

* Zippy now won't do -zippyTalk in a channel unless his handle is
mentioned there.

* Zippy now ignores bughouse holdings if not actually playing, instead
of stupidly trying to abort the game.  Let me know if other strange
things happen when Zippy is observing a game.

* Set ICS Interaction title bar to "user@server".  User name is
determined by parsing message during the login sequence.  Zippy now
uses this method to get its name too, so ZIPPYNAME is eliminated.
Problem: with xboard, the title bar gets set to "xterm" upon exit,
since we didn't know what it was to start with.

* Fixed xboard-only bug in -colorChannel1 command line option.  Thanks
to Fredrik Sandstrom for the bug report and fix.

* Check for overflow of MAX_MOVES.  Still not sure this is checked for
everywhere it needs to be.

* Added .epd to suffixes recognized by WinBoard as position files.
This is a bit of a kludge, as the .fen reading code was not written
with .epd files in mind, but reading them with it is said to work.

* Added ZIPPYGAMESTART feature, suggested by Jason Hoblit.  See
zippy.README.

* Restored the feature of soaking up the moves from a user-typed "moves",
"oldmoves", or "smoves" ICS command when xboard/WinBoard is idle, but
changed it so that the move list is also echoed to the screen.  Some folks
still liked the old feature since it is faster than examine mode, and
printing the moves to the screen should eliminate the confusion it caused
for other people.

* Bugfix: in WinBoard, starting a click-click move and then clicking on
a black line would cause the source square to be forgotten without taking
the highlight down.  Now the errant click is ignored.

* Corrected shortcut key labels for LoadNext/PrevPosition in WinBoard's
File menu.  Thanks to Andreas Stabel for the bug report.

* Bugfix: if you had a completely full ICS context menu in WinBoard
(90 items), the last one would do nothing if selected.

* The "computer" engine command added in 4.0.1 was not working; I had
left out part of the code patch.  Oops.

4.0.1 -- Fri Feb 12 21:24:15 PST 1999  Tim Mann

This is a minor bugfix release.  The first item listed below is the
only interesting new feature.

* You can now type in moves from the keyboard in WinBoard.  Either
select TypeInMove from Step menu, or (if you are not in ICS mode),
just start typing.  In ICS mode you can type moves into the ICS
interaction window anyway, so I've retained the feature that typing
switches you there.  Not implemented in xboard.

* Removed the "noise 1000" command that used to be sent to the chess
engine when entering analysis modes.

* Deleted the feature where we soak up the output of a user-typed ICS
"moves" or "oldmoves" command.  This ancient feature is not really
useful anymore, now that "examine" exists, and it caused a lot of
complaints.

* Increased maximum size of ICS text menu to 90 entries and added
overflow checking.

* The command sent to ICS by Zippy at the end of each game can now be
customized with the ZIPPYGAMEEND environment variable.

* We treat xboard.info as a pseudo-source since most people don't have
the tools to rebuild it.  But "make install" erroneously had a
dependency on it, even though "make all" did not.  Fixed.

* Miscellaneous minor updates and clarifications to engine-intf.txt.
In particular, documented the st and sd commands (support for command
line options -searchTime and -searchDepth), warts and all.

* Added a small code patch from Bob Hyatt to detect when Zippy's ICS
opponent is a computer and send the command "computer" to the engine.
Added this command to engine-intf.txt.  Thanks, Bob.

* Bugfix: configure was looking for usleep(), but config.h.in didn't
have an entry for it, so HAVE_USLEEP was never defined.  Also, in the
non-HAVE_USLEEP case, there was a possible race condition between the
signal handler and the pause() call.  Thanks to Ben Dean-Kawamura for
the bug reports.

* Bugfix: a "1..." preceding Black's first move (usually present if
there is a comment on White's first move) would cause the parser
to think a new game started there and report "End of game" on the
current game.  Thanks to Dell Garner for the bug report.

* Changed the abbreviation for -searchDepth from -sd to -depth to fix
a clash with -secondDirectory.

* Changed -searchDepth to be orthogonal to the various timing options.
Previously it would search for an unlimited time to the given depth;
now the engine is allowed to choose its search time in the normal way,
but is told to cut off the search early if it reaches the given depth.

* Bugfix: SaveSettings in WinBoard was saving both the long form
option highlightSquareColor and its abbreviation (hsc).

* Fixed a redisplay bug in DisplayBothClocks that could cause White's
clock to be highlighted in EditPosition mode when Black's should have been.

* Fixed Zippy's feature of passing on draw offers from the ICS opponent.
The colorization code was eating the offer before Zippy saw it, so I disabled
colorization of offers while in -zp or -zt mode.  (Other types of
colorization are already disabled in Zippy mode for similar reasons.)

* Added some more commands to the default ICS context menu in WinBoard,
and put "(name)" on the ones that insert a playername (or game number).

* Added missing documentation to WinBoard help file: PonderNextMove, the
-ponder/-xponder argument, and the up/down arrow history feature in the
ICS Interaction window.

* Fixed ZIPPYNOPLAYCRAFTY feature to fire only if the opponent is Crafty,
not if some other Crafty is observing the game and kibitzes.

* Fixed bugs in generating "partner" commands when Zippy is playing
bughouse.

4.0.0 -- Sat Jun 20 16:59:47 PDT 1998  Tim Mann

* Fixed WinBoard resizing so that the edges you do not drag remain in
place, even when the final size is different from what you dragged to.
(This doesn't work on NT 3.51 or earlier, if anyone cares.)

* Fixed problems with resizing WinBoard on Windows NT: you no longer get
whitespace around the board, a clipped board, or a lot of flashing during
the drag.

* Added customization for highlight-square color.

* Bugfix: backend.c wrote to stderr in a few obscure places, which doesn't
work in WinBoard.

* Added Index Number field back to LoadGame/LoadPosition dialog boxes in
WinBoard.  This required some extra code to work with both the
Explorer-style dialog boxes and the old-style NT 3.51 dialog boxes.  Ugh.

* WinBoard startup dialog now understands -zp mode.  The first chess
engine selector is enabled and the value is used.

* Bugfix: ResetGame would send "exit" to engine twice if it was in an
analysis mode.

* Bugfix: The response "Illegal move: bk" to the "bk" command, indicating
that it is not implemented, would confuse xboard.

* Bugfix: Ignore empty hints ("Hint:\n") instead of printing a message
saying that some bogus string (often "Hint:") is illegal.

* Bugfix: LoadNextGame/LoadPreviousGame/ReloadSameGame were broken.

3.6.12 -- Sat Jun 13 14:57:10 PDT 1998  Tim Mann

Beta test release of xboard and WinBoard.

* Spruced up xboard man page a bit.

* Added Highlight Last Move to xboard too.

* The revamped xboard mouse click code had several more bugs than the
previous version, so I revamped it again.  All fixed now, I hope.

* xboard bugfix: crash in CreatePieceMenu.

* Oops.  winboard-3_6_11beta.exe installed 3.6.10beta, not the new version.

3.6.11 -- Fri Jun 12 20:00:48 PDT 1998  Tim Mann

Beta test release of xboard and WinBoard.

* Major reorganization to engine-intf.txt.  Some new features added,
many existing features clarified and hints for engine programmers added.
Support for new features added to program.

* Changed defaults for -scp (etc.) back to being independent of -fcp, as
in 3.6.2. That is, the default is no longer to copy the value from -fcp if
-scp is not specified.  This will sometimes make command lines longer but
will remove some confusing situations.

* Split -reuseChessPrograms into two options: -reuseFirst and
-reuseSecond.  Fixed obscure bugs when reuse is off.

* Detect ICC and change some command usage slightly there: only ICC has
set-quietly; only FICS has iset and bsetup clear.  With this feature, we
can now safely use bsetup clear when appropriate instead of deleting each
piece one by one.

* Added indent to lines that are wrapped by WinBoard ICS Interaction
window.  This seems to tickle a Microsoft bug in the RichEdit control:
occasionally the start of the next line after the wrapped line is indented
too.  I think the wrapped indent feature is useful enough that it's worth
having in spite of the bug.

* WinBoard bugfix: bounding box for the board was computed incorrectly by
mouse code.  It would think you moved off the bottom of the board before
you really did, so it was hard to drag pieces along the bottom rank on
smaller board sizes.

* Fixed obscure parser.l bug: ambiguity between capture using : instead of
x and email header line.  Also removed or simplified a lot of complex
patterns for English (as opposed to PGN) end-of-game indications while I
was in there.

* Added PonderNextMove option.

* xboard EditPosition piece menus now let you set the side to play by
selecting the "White" or "Black" label at the top of the menu, like
WinBoard.  (Both xboard and WinBoard still let you set the side to play by
clicking on the clock.)

* Merged documentation of Crafty and GNU Chess thinking output in
engine-intf.txt.  Corrected code in backend.c to match documentation; it
used to require the PV to start in column 27 (!).

* Fixed bug that caused crashes in "-reuseChessProgram false" mode.

* Draw by agreement in Two Machines mode now works as it is supposed to.
Other minor bugs in handling game end messages from machines are also
fixed, and engine-intf.txt is revised in this area.

* Revamped xboard code for handling mouse clicks on the board.  In
particular, it had a bug if a click-click promotion move was made.

* Fixed and reinstated optimization for << in analysis modes: send
initString and re-send the board if needed.  Sending repeated undos
is bad because Crafty prints some analysis after every one.

* Don't go into a spin loop while pausing between xboard animation frames.
This caused problems, especially when compiling with -O2 on HP-UX.

* WinBoard Options/Colors/RevertToDefaults now also turns off Monochrome
if it was on, and similarly for Options/ICSInteractionColors/RevertToDefaults.
This seemed to confuse people before.

* Bugfix: ICS Interaction context menu items with neither of the two
optional flags set did not work.

* Bugfix: PGN move numbers are optional, but we were insisting on "1"
right after the tags.

* Bugfix: only the first game of an -mg match was loading -lpf or -lgf.

* Other minor fixes and code cleanup.

3.6.10 -- Mon Jun  1 01:32:31 PDT 1998  Tim Mann

* Added ZIPPYNOPLAYCRAFTY feature; see zippy.README.

* Added timestamps for chess engine interaction to -debug output.

* Now allow 10 seconds (was 5) between games in match mode for engines to
respond to "force" command that ended previous game.

* When a game is picked from the game list and the names of White and
Black are known, display them in the window banner instead of the filename
and game number.

* Changed default for -scp and -sd to be the values of -fcp and -fd.

* Added boxes for both engines to WinBoard startup dialog.  Changed to
dropdown boxes whose content can be set from winboard.ini, similar to
the box for ICS name.

* Accept "offer draw" (or "*offer*draw*") from engine to offer a draw.

* Accept "move mmmm" from engine as well as silly "1. ... mmmm" syntax.

* Added support for bughouse engines (Zippy mode only).

* Move bombproofing for engines sending illegal or out of turn moves.

* Always send "xboard\n" to engine as first command.  Eliminates need
to run crafty as "crafty xboard".

* Rewrote xboard code for input from engines to fix longstanding bugs.
Previously an engine that sent one line in two separate writes with 
some time lapse in between (or perhaps that sent several lines in one
write) could make xboard hang.  WinBoard did not have this bug.

* Fixed minor bugs in dealing with illegal moves.

* Added "iset startpos 1" for FICS.

* Removed special support to put "GNU Chess" or "Crafty" in the window
banner instead of the engine's base filename.  This gives a more level
playing field for other engines.  Also fixed bugs in the code that finds the
engine's base filename.

* Bugfix: Abort or Resign at the very start of a game in chess engine mode
would fail to send "force" to the engine.

* Fixed some minor bugs in new "result" messages, and changed format to
include a trailing PGN comment giving the reason the game ended.

* Changed syntax of /font options in WinBoard; old syntax was assuming
that a font name can't contain a digit, which is wrong.  Old font names in
WinBoard.ini files are still accepted.

* Serious further work on getting WinBoard to read/write files in the
"right" directories and documenting exactly what it does.

* Added LoadPreviousPosition, LoadNextPosition, ReloadSamePosition.

* Misc. tiny fixes and cleanup.

3.6.9 -- Sun May 24 20:53:08 PDT 1998  Tim Mann

xboard and WinBoard beta release.

* Tightened up illegal move handling.  Illegal or meaningless moves from a
file or chess engine should no longer get translated into a1a1 and blindly
applied.  Explicit moves to the same square (like a1a1) no longer make the
piece on that square capture itself and vanish.

* Added AlwaysOnTop feature to winboard.

* Fixed bug in winboard "@" command line indirection operator.

* Removed buggy "optimization" code for ToStart in analysis modes.

* Fixed typo bug in zippy.c that would often cause winboard to crash
right after starting a game in -zp mode.

3.6.8 -- Sat May 23 22:23:19 PDT 1998  Tim Mann

xboard and WinBoard beta release.

* Added -firstDirectory and -secondDirectory so that chess engines can
be started in a different directory from xboard/WinBoard itself.

* Added -matchGames to allow more than one game in -matchMode.  The games
are played with colors alternating between the two chess engines.  Had to
add a time delay at the end of each game to catch up with game ending
messages from both chess engines -- yuck.

* Changed TwoMachines mode to have firstChessProgram play White by
default.  Added -firstPlaysBlack to restore old behavior.

* Major rework of code in backend.c for dealing with first and second
chess engines.  Generalized the code and created a struct type with one
instance for each engine.  Several small bugs fixed along the way. 

* Fixed bugs in Abort in local chess engine mode.

* Tightened pattern matching to recognize moves from machine.  Should cure
bogus "Illegal move" errors while Crafty is pondering with Show Thinking
on, among other things.

* WinBoard keeps focus in board window after popping up a comment due to
Forward/Backward/ToStart/ToEnd commands.

* Finally got rid of warnings on compiling moves.c under Windows.

* Fixed so WinBoard will compile on machines where WM_MOUSELEAVE is not
defined.

* Removed various bits of unused code.

* Parse (and ignore) NAGs in PGN files.

* Updated engine-intf.txt.

* Send "result R" to the chess program at the end of each game, where
R is the PGN result 1-0, 0-1, 1/2-1/2, or *.

* Bugfix: suppress animation when loading a game with timeDelay=0.

* Bug fix: pondering output from Crafty that contained "..." could be
mistaken for a move, causing bogus error messages and general chaos.

* xboard "make clean" bug fix: moved things created by configure to be
removed only by "make distclean".

* Fixed a remaining case where move was being animated before being
relayed to opponent.

* Backend was reversing time/otim commands to the chess program (new bug),
and was often not sending them on the first move of a game or the first
move after a mode change (old bug).  Fixing the latter required a kludge
to work with both GNU Chess and Crafty, because Crafty swaps white and
black clock times when it swaps sides, while GNU Chess does not.  The
kludge is that we always make sure the chess engine thinks its opponent is
on move before we set the clock.  This will sometimes result in sequences
like "white\ntime xxx\notim yyy\nblack\ngo", or the same with white and
black interchanged.

* Removed last trace of old -whiteString and -blackString arguments.

* Fixed so that you can type "bd" into the Direct Command dialog when
debugging gnuchess, without having xboard parse the clock display
(starting with "White") as "White wins" and end the game.

* Changed moves.c to make a distinction between IllegalMove (where we
could tell what move was intended, but it violates the chess rules) and
MeaninglessMove (where we could not tell).  This is a step towards getting
rid of cases where we translate meaningless moves into "a1a1", but some
work is probably needed to track them all down.

* Fixed bug in WinBoard monoMode dragging.

* Changed WinBoard to new style file dialogs.  In the process, disabled
the feature that let you enter the index number of a game or position within
the file in the dialog.  This is no loss in LoadGame, as you can pick your
game from the game list dialog that follows, but it does remove some
functionality from LoadPosition.

3.6.7 -- Mon May 18 21:25:00 PDT 1998  Tim Mann

xboard and WinBoard beta test release.  I've been unable to reproduce some
reported drawing bugs, so they may still not be fixed.

* Changed PopUpErrors option to PopupMoveErrors, which is more useful, and
changed default to off.  Errors other than move errors should always be in
popups; they are often too big for the message area and too important to
put where they can be easily missed.  But it seems good to get rid of the
popups for move errors by default.

* xboard analysis window no longer forces itself to top whenever there is
new output.  Also, now uses built-in Xaw word wrapping instead of trying
to do its own; works better.

* Bugfix: Reset did not take Crafty out of analyze mode.

* A chess engine no longer has to claim to be Crafty to be able to use
AnalysisMode and AnalyzeFile; it just has to implement the analyze
command as Crafty does.  Put in heuristics to generate more informative
error messages if the engine does not support analyze, and checked that
they work at least with GNU Chess.

* Added click-click moving, HighlightLastMove, and HighlightDragging to
WinBoard.  The highlight features are unlikely to go into xboard unless
the xboard drawing code is cleaned up to do *all* drawing on the board
from DrawPosition, as WinBoard does.  This is low priority for me.

* Eliminated bogus "Error gathering move list: no header" popup.  You
could get this by observing two fast games at once without turning off
GetMoveList.

* Disable WinBoard Sounds menu entirely in -ncp mode, since not even
MoveSound is used in that mode.

* WinBoard bugfix: Several problems were caused if the user changed 
WinBoard's current directory in a Load, Save, or Browse dialog.  In
particular, WinBoard.ini would get saved in the wrong directory, and
sound .wav file names would not get saved with a full pathname.  I think
all such problems are now fixed.

* WinBoard bugfix: iconizing the board after an aborted mouse resize would
(partially) resize it to Tiny.

* WinBoard bugfix: board was not being drawn in color on 256-color
displays.  This bug was introduced in 3.6.6.

3.6.6 -- Tue May 12 17:43:43 PDT 1998  Tim Mann

xboard and WinBoard beta release.  Note: not all reported bugs are fixed.

* Move animation in WinBoard, programmed by Henrik Gram.  Great stuff!

* Animate backward moves too.

* xboard bugfix: -font (and friends) can now specify a font alias (such
as "fixed"), not just a full X Consortium name pattern.  Previously this
would give a resource conversion error message.

* Some coding style cleanup in xboard.c.

* Earlier error check for moving wrong color piece in EditGame mode.

* Completed fix to pattern matching; see 3.6.5.

* Fixed some software rot bugs in Zippy.

* Split AnimateDragging as separate option from AnimateMoves.

* Added FlashMoves to options menu and -flash/-xflash to command line
options (xboard only).

* Some preliminary work on visible bughouse holding support.  Most of the
code is not present in this release because I have not gotten it working
properly yet; my first approach went down a blind alley and I had to
remove most of the partially working code in preparation for a rewrite.

3.6.5 -- Fri May  8 14:22:09 PDT 1998  Tim Mann

Not generally released.

* Pack bughouse holdings display in banner more tightly when board is one
of the smaller sizes (when smallLayout or tinyLayout is true, that is).
An interim measure, but should help.

* Completed fix to make move list parsing recognize the end condition and
final comment.

* Fixed pattern matching for chatter (tells, etc.) to be more reliable.
This should reduce the incidence of colors starting in the wrong place and
of incorrectly matching things that look like other patterns inside
chatter, such as "shout <12>".  (These errors were already rare.)

* WinBoard now lets you customize the list of ICS's in the startup dialog.
For now, at least, requires editing WinBoard.ini with a text editor.

* Bug fix: WinBoard window sizing and position setting save/restore did
not work right when the taskbar was at the top (or left side) of the
screen and not in autohide mode.

* Merged xboard click/click mode and drag mode, and deleted
ClickClickMoving option on menu.  Now if you click on a piece, it
highlights and a further click on another square will move it.  (A second
click on the same square takes down the highlight.)  Or if you press down
on a piece and start to drag, you can drag it to a new square.

* When making click/click style moves, if your second click is on a piece
of the same color, instead of saying "Illegal move", we now cancel the first
starting square, replace it with the second, and wait for another click to
finish the move.

* Separate -secondInitString.  Default: same as -initString.  Suggested by
Remi Coulom.

* Patch from Frank McIngvale to make animation work with XIMs.  Yay!

* Changed move animation to happen after move is passed on to ICS or the
chess engine, where applicable.  Fixed problems with updating moveList
that this uncovered.

* Changed game list to show PGN result token for each game.

* The usual minor bug fixes.

3.6.4 -- Thu Apr 30 23:14:43 PDT 1998  Tim Mann

xboard and WinBoard beta test release.

* xboard documentation updated, both man page and info file.  An attempt
to generate both from the same source with LinuxDocSGML was abandoned, at
least for now.  WinBoard doc updated too.

* Merged updates to animation code from Hugh.  Fixed a couple of remaining
bugs, mostly to do with handling Pause mode correctly.

* Added SaveSettingsOnExit to WinBoard.

* WinBoard now saves the last screen position of the Comment, Edit Tags,
Game List, and Analysis windows (in addition to the board and ICS
Interaction windows) in WinBoard.ini.

* WinBoard now provides a way to customize the right-button context menu
in the output area of the ICS Interaction window.  For now, requires
editing WinBoard.ini with a text editor.  

* Various minor fixes to ICS message recognition, most affecting only Zippy.

* Colorize the notification when someone sends you a message while you are
logged in, not the output of the "messages" command.

* Fixed colorization to turn off at the right place more reliably.

* Added a new color and WinBoard sound for "requests": abort, adjourn,
draw, pause, and takeback.  The same color and sound are used for all of
these.

* Added GUI to change sounds in WinBoard.

* Suppressed WinBoard error popup when a sound can't be played.  They were
popping up when two separate WinBoards tried to play a sound at the same
time (such as during a bughouse match).

* Fixed WinBoard crash when a sound file didn't exist.

* WinBoard ports of timestamp and timeseal now produce decent error
messages if they fail to connect, and WinBoard captures the messages and
puts them into a popup.  Previously WinBoard would exit with no message on
such errors.

* WinBoard resizing with mouse improved; now accurately picks the largest
size that fits in the new area.

* Added option to turn off error message popups.

* Added sizes slim, dinky, and teeny.

* Fixed some minor problems with error message wording.

3.6.3 -- Mon Feb 23 19:08:57 PST 1998  Tim Mann

xboard beta test release only.  Man page is updated for the changes, but info
file is not.

* Bug fix: LoadGame could not load a PGN game whose first move was
castling (which is possible from setup positions).

* Bug fix: we were ignoring illegal move messages from chess engines in
modes where the engine is refereeing but not participating (EditGame,
LoadGame, AnalyzeFile).  The problem was in a workaround for a Crafty bug,
where Crafty generates a bogus illegal move message if a "." (used by
the PeriodicUpdates feature) is sent in analysis mode when the current
analysis reveals a forced mate.  Installed a different workaround (that
works in most but not all cases) and reported the Crafty bug.  Bob Hyatt
replied with a fix, so the Crafty bug should be gone in Crafty 14.12 or so.

* WinBoard only: enabled resizing board with the mouse.  It snaps to the
nearest (but not too much larger) predefined size.

* The xboard default font is now sized according to piece size, just as
clockFont and coordFont have always been.

* Added more sizes: bulky, moderate, average, middling, mediocre, petite.
        
* xboard only: Added Animate Moves and Click-Click Moving to Options menu.  

* xboard only: Added code to animate piece movement, from Hugh Fisher.

* New feature: autoflag is implemented in MachineWhite, MachineBlack, and
TwoMachines modes.  If it is on and either a chess program or the user
oversteps his time, xboard will automatically call the flag and end the
game.  This feature was requested by someone who runs matches between
programs and wants to penalize GNU Chess for its habit of running slightly
over at the end of a time control.  I personally think it's pointless.

* Fixed a bug in Zippy's pattern matching; he can now reply to ICC
messages again.

* Generate prettier notation for illegal moves.  If a move is illegal only
because it leaves the player in check, generate notation as if that were
not illegal.  For instance, if two knights are a knight's move away from
f3, one on g1 and one on e5 that is pinned, the notation would be Nf3 to
move the g1 knight, Nef3 to move the e5 knight.  (Suggested by Philippe
Schnoebelen.)  In addition, if a move is illegal because the type of piece
moved cannot go that way, would be jumping over another piece, etc., we
give fully disambiguated coordinate notation; for example, Ng1g3, Ke1xe8,
d2xd8=Q, etc.  An alternative would be to still write Ng3 if there is only
one knight on the board, or Ngg3 if the other knight is not on the g file,
but life is too short to code up all this stuff, and perhaps the way I did
it is better as it emphasizes that this move was really made despite being
illegal.  I wonder if the new PGN standard revision will speak to notation
of illegal moves?

* Fixed gross bug from 3.6.2: if a chess program said "checkmate" and
White was left on move, the PGN outcome was "1-0 {Black mates}" instead
of "0-1 {Black mates}"!  The bug occurred only for that specific string.

3.6.2 -- Wed Jul 23 16:47:29 PDT 1997  Tim Mann

* There was a bug in backing up from the end of a game against a chess
engine, introduced when -reuse mode was added.  Fixed.

* Recognize "{" as terminating an ICS move list.  Previously we needed to
see a prompt (containing %), but occasionally ICC can send more stuff
right after a move list with no prompt in between.

* In WinBoard, -mm now implies -cp, so you don't get the startup dialog.

* WinBoard startup dialog no longer sets -scp.  People used to get
confused because they would give the -fcp and -scp arguments, but omit
-cp.  This would bring up the startup dialog, which would set both -fcp
and -scp to the value in the engine name box.  Now you can change -scp
only with the command-line argument.  That will confuse some people too,
but hopefully fewer.  I think overall it's better than putting spaces for
both -fcp and -scp in the dialog, which would puzzle newbies.

* Bug fix: Analyze File mode now works in WinBoard.

* Bug fix: WinBoard window no longer cuts off part of the bottom row if
the menu bar wraps to a second line.

* Accept game end messages of the form "PGN-result {comment}" from the
chess engine, where PGN-result is 0-1, 1-0, or 1/2-1/2.  Accept "resign"
or "computer resigns" from the chess engine as a synonym for "0-1 {White
resigns}" or "1-0 {Black resigns}". Accept any message from the chess
engine containing "game is a draw" as a synonym for "1/2-1/2 {Draw}".
Accept "White resigns" or "Black resigns".

* Enable "Action / Draw" menu item in chess engine mode.  It sends the
command "draw" to the engine.  Accept any message from the chess engine
containing "offers a draw" as a draw offer.  WARNING: Draw offers from the
user and the engine (or from two engines) are not yet matched up by xboard
in chess engine mode.  Two engines could get into a loop offering each
other draws.

* Bug fix: Game end messages from the chess engine are always ignored in
Zippy mode.  Previously they could sometimes be processed, which was
sometimes making Zippy hang at the start of a new game that followed
quickly after an old one.

* Zippy can now respond to titled players in channels.  Zippy will avoid
talking to himself in a channel if he knows his own name.

* Bug fix: in colorization arguments, the bold setting was being ignored
if the background color was defaulted.  (xboard only.)

* Extended ParseFEN to accept Crafty 12.2's interpretation of FEN.
If fewer than 8 squares are given in a row, or fewer than 8 rows are
given, the uncovered squares are empty.  Also, there can be a / after the
8th row.

3.6.1 -- Sat May 17 01:02:33 PDT 1997  Tim Mann

* Bug fix: Hitting escape while entering a password on WinBoard no longer
sets the font color back to black (which made further typing visible).

* Bug fix: On Windows NT 4.0, WinBoard /ics /icshelper=timestamp would
often hang if you tried to exit with the Exit menu item, the [X] button,
etc.  The same fix may also have corrected the problem where running a
chess program with /xreuse would sometimes crash Windows 95 (but not NT)
when the chess program was killed and quickly restarted.

* Bug fix: Periodic Updates would not work if turned on after having been
off.  Also, the default was needlessly "off" in WinBoard.

* Zippy bug fix: Continuing an adjourned game was totally broken.  The
wrong moves, or no moves at all, were being fed to the chess program.

* WinBoard change: Giving -ics without -icshost now pops up the startup
dialog with "Use an Internet Chess Server" pre-selected, so that you get a
menu of chess servers instead of going directly to ICC.

* Bug fix: WinBoard's SaveSettings was saving -icshost, -icsport, and
-icshelper in winboard.ini.  This causes a problem if you use a different
icshost later that needs the default values for icsport (5000) and
icshelper ("", meaning none).

* Disabled complaints about unexpected "[Ii]llegal move" messages from
the chess program; we still get them at times.

* Greatly simplified the code for Move Now and SIGINT (ATTENTION).  Fixes
the bug where Move Now did not work with GNU Chess in WinBoard, and
several more obscure bugs as well.  Reintroduces a minor bug: GNU Chess might
not think on your time after a Move Now command.  Trying to fix this
causes too many other problems.  Documented change in engine-intf.txt.

* Cleaned up code for obtaining ratings from ICS.  It was kludgey and may
have had bugs on systems with no "Creating:" message.  It should fail
gracefully there, just not finding the ratings.

3.6.0 -- Thu May  8 19:55:58 PDT 1997  Tim Mann

* WinBoard now has right-button context menus in the ICS Interaction
window.  Some of the items do immediate commands on the player name you
have selected or pointed to.  (Suggested by Paolo Casaschi.)  Also, the
middle button or shift+right does a immediate Copy and Paste.

* Fixed a bug in Zippy's wild rejection.  The bug was causing Zippy to
reject all forms of wild.  Now it rejects only 1, 9, 16, 17, and 24, as
was intended.  These variants have different rules, while the other
variants defined so far only have different starting positions.
(Actually, wild 2, 3, and 4 never allow castling, but may sometimes by
chance have king and rook start on squares where castling appears legal.
This could cause a problem later.)

* Zippy now resumes adjourned games by feeding the chess program 
the move list instead of the position.  Setting -getMoveList to False 
restores the old behavior.  Feeding in the move list is better because 
it enables the chess program to correctly handle en passant legality, 
castling legality, draw by repetition, and draw by the 50 move rule. 

* Added a WinBoard startup dialog for people who run WinBoard.exe without
using the Start menu or reading the help file.  Instead of being dumped
directly into GNU Chess mode, you now get a menu of choices.  You now have
to give the new WinBoard -cp option to go directly into chess engine mode.
I did not add this feature to xboard.

* Handling of "illegal move" from Crafty is now back on.  (It was turned
off in 3.4.6; see below.)  Crafty 12.0 doesn't seem to send bogus ones
anymore.  I did have to suppress sending time and otim in analyze mode to
make this work, but there is no harm in that.

* Zippy now sends "gameend" to ICS at the end of each game.  You can alias
this to "seek" or whatever you like.

* Added Crafty support for Move Now, using the "?" command.  We test
whether this command exists by trying it once at the start of the first
game, before it really makes sense.  If we don't get a message like
"Illegal move ... ?", we assume it's OK.  (We're careful about this
because sending a "?" to GNU Chess on Move Now keeps it from pondering the
next move as it should.)

* Added support for a tellusererror command from the chess engine.  See
engine-intf.txt.

* Attempted to make engine-intf.txt complete.

* Minor fixes/cleanup to GameEnds code, to be sure we always interrupt
when needed, and always undo an extra move generated by stopping the chess
program while it is thinking.

3.5.7 -- Sat May 03 22:37:44 PDT 1997  Tim Mann

Beta release of both xboard and WinBoard.

* Updated xboard documentation.  I really wish the man page was built from
the info file instead of being separate!  It's a pain to update them both.

* Updated documentation of the interface between XBoard/WinBoard and chess
programs, and renamed it to engine-intf.txt.

* Added support for tellics, telluser, and askuser commands from the chess
engine.  See engine-intf.txt.

* Various bug fixes.  On WinBoard, 256 color displays are finally handled
right; no more color flashing when you switch windows unless the colors
really need to be updated.  Also, I think the problem of the console font
failing to change or changing back by itself is finally fixed.

3.5.6 -- Tue Apr 29 03:08:00 PDT 1997

Beta release, WinBoard only.  ChangeLog was out of date in release,
updated here.

* Added ICC timestamp and FICS timeseal to WinBoard distribution.  These
are my own ports to Win32, based on and used by permission of the owners
of the proprietary source code (Daniel Sleator for timestamp, Henrik Gram
for timeseal).  The owners permit these programs to be distributed only in
binary form (to help prevent cheating), so to avoid running afoul of the
GPL, I have kept them as separate programs, not linked into the WinBoard
address space.  Updated Start menu icons to use them.

* Miscellaneous smaller changes.

3.5.5 -- Fri Apr 25 03:06:00 PDT 1997

Beta release, WinBoard only.  ChangeLog was out of date in release;
updated here.

* Added simple sound support to WinBoard.

* Updated WinBoard documentation.

* Internal implementation of the telnet protocol is now more complete.
You can now hop through a VMS host on the way to ICS.  For example, do
"xboard -ics -icshost vms.host.edu -icsport 23", then log in.  Do not use
the -telnet flag; that says to use an *external* telnet program.  To avoid
double echoes, you need to force character mode in VMS telnet after
connecting from VMS to ICS.  Do "^]^Mset mode char^M^M".

* Display opponent names in the title bar for MachineWhite, MachineBlack,
and TwoMachines modes.

* WinBoard now has an installer, built with InstallShield.  The installer
works with a binary-only distribution.  Sources are still freely
available, but now separately.

* Added SaveSettings to WinBoard, which saves current option settings to
winboard.ini.  Made sure all options can be saved and loaded, including
fonts and com port settings.

* If the game ends while you are dragging a piece, we now don't let you
finish the move.  Formerly the move was accepted and your opponent's clock
would start running again.

3.5.4 -- Fri Apr 18 01:15:24 PDT 1997  Tim Mann

Beta test release of WinBoard and xboard.  The documentation remains
out of date.

* Many fixes and improvements to new WinBoard user interface code.

* In Zippy mode, avoid sending another copy of the same move to the chess
program if ICS sends us another copy of the board image.  Hard to believe
it took me until now to diagnose and fix this problem!

3.5.3 -- Sat Apr 12 19:49:33 PDT 1997  Tim Mann

Beta test release of WinBoard and xboard.

* Fixed fatal bug in WinBoard input handling.

* Made code to stop chess program for reuse a bit smarter.  Removed kludge
of "white" in initString; it didn't work well, and is unneeded
with Crafty 11.21 and later.

3.5.2 -- Sat Apr 12 15:40:01 PDT 1997  Tim Mann

Beta test release of WinBoard.

* Ignore check and permit promotion to King during suicide games; no need
to turn off TestLegality.

* Renamed CheckLegality to TestLegality to avoid confusion with the chess
term "check".

* Added Rematch to Action menu.

* WinBoard now has a custom ICS interaction window with scrollback, a
separate line for input, and colorization.  Accelerators that conflict
with normal editing keys were changed (by requiring Alt+) so they can work
in both the console and the main window.

* WinBoard error popups are now non-modal and disappear when you make a
new move, as in xboard.

* Configure now defaults to --disable-ptys on all systems.  If anyone has
an ancient SysV system where pipes don't work with select, they can still
do "configure --enable-ptys" explicitly.  I would appreciate getting a bug
report if this happens to anyone, with complete output from configure and
"uname -a".

* "make install" now makes the installation directories too.

* Fixed a bug that would cause an "Illegal move" message for Black from
GNU Chess to be incorrectly considered bogus.

* Handle name changes during an ICS game (FICS bname and wname commands).

* You can force both chess programs to be killed at the end of each game
by turning off the reuseChessPrograms option.  There are still some
problems with reusing Crafty even in version 11.20.

* Now the second chess program stays around by default too, if it ever
gets started.

3.5.1 -- Sat Apr  5 16:47:48 PST 1997  Tim Mann

Beta test release.

* Don't kill off and restart the chess program for each game; keep the
same one running, using the "new" command to start a new game.  This
change works around the problem in Windows 95 that makes WinBoard crash it
at times, and is generally desirable to make new games start faster.  The
second chess program (for TwoMachines) is still killed at the end of the
game, and unfortunately this can still crash Windows 95.  Temporarily
added "white" to the initString to make this work with Crafty 11.20, which
has a minor bug in "new".  Older versions of Crafty have worse bugs in
"new"; they should not be used with this version of xboard.

* Support for FICS suicide chess: Parse illegal moves (that leave King in
check) in game history.  If CheckLegality is off, allow promotion to King
and illegal moves in game files.  We still generate e2e4 style notation
for illegal moves.

* Handle FICS "has timeseal; checking" message.

* Changed the coords from white back to black; this was a bug.

* Fixed problems compiling with K&R compilers.

* Fixed an old bug in RegisterMove that was crashing cmail on some systems
(notably linux).

3.5.0 -- Thu Jan  2 16:59:49 PST 1997  Tim Mann

Thanks to Frank McIngvale for much of the work on versions 3.4.4 and above!

* Fix Crafty resumed game time bug (frankm)

* Word wrap text in Analysis window (frankm)

* More debug info for XPM loading (frankm)

* Replaced config.sub, config.guess, etc., with up-to-date versions from
autoconf 2.12.  Hopefully this will fix problems on Pentium Pro machines.

* Removed some Makefile gunk that was causing looping for one person.

3.4.7 -- Thu Dec 19 14:22:41 PST 1996  Tim Mann

All changes from Frank:

* Retrieve ratings from ICC (and FICS, etc., when they add the Creating:
message), save them in the PGN tags, and pass them to Crafty in Zippy mode.

* Add settable time delay between characters in ICS login script.

* Colorize messages like personal tells.  Fix false recognition of channel
tells.

* Pass "tells" from Crafty through to ICS (in Zippy mode).

* Implement ~/ filename convention from C shell for game and position file
names.

* ZIPPYACCEPTONLY feature for testing.

3.4, patchlevel 6 -- Sat Nov 23 16:58:50 PST 1996  Tim Mann

* Put recognition of "illegal move" messages from Crafty for
illegal castling, etc., inside an #if that is currently turned off,
because of a bug in Crafty that generates bogus "illegal move" messages
after some moves that are actually legal and accepted by Crafty.

* Added -checkLegality option; previously this could not be turned off.

* Fixed an old bug in finding default board size parameters.

* Differentiated among channel tell, kibitz/whisper, and personal tell/say.

* Fixed a bug in detecting Xpm in the configure script.

3.4, patchlevel 5 -- Mon Nov 18 16:22:53 PST 1996  Tim Mann

* Added zic2xpm to the kit; it had been omitted by mistake.

* Added some default colors for -colorize

3.4, patchlevel 4 -- Sat Nov 16 18:10:17 PST 1996  Tim Mann

This is meant to be a beta release in preparation for version 3.5.

* Added InfoXBoard and ManXBoard to the Help menu.

* Made Frank's ICS input box optional, defaulting to off.

* Merged in Frank McIngvale's XbKit.  Many new features, including
-clickClick mode, the Analysis modes, piece flashing, ZIICS import,
ICS text colorization, and the ICS input box.  Many thanks to Frank for
supplying and documenting this code.

3.4, patchlevel 3 -- Mon Nov 11 18:23:14 PST 1996  Tim Mann

Small set of changes made while Frank McIngvale was working on XbKit in
parallel.  Unreleased in this form.

* Updated zippy.README.

* Removed useless X event handler and removed strange code for copying
form translation table to board that worked around the bug it caused.
This makes changing bindings in .Xdefaults more straightforward.

* Now pressing the Control key steps back one move, and releasing it steps
forward again.

* Moved quit from "q" to "Q" for greater safety.

* Use "unobserve" instead of "observe" to stop observing; needed on FICS.

* Support for interface variable.

* Strip titles from people who are talking to us, so (for example) Zippy
won't try things like "tell Darooha(*) hello".  Needed on FICS.

3.4, patchlevel 2 -- Tue Jul  9 19:06:42 PDT 1996  Tim Mann

This patchlevel is not planned to be an announced release.  It's in
preparation for integrating Frank McIngvale's XbKit.  It mostly contains
minor fixes I've accumulated since 3.4.pl1.

* WinBoard /telnet option now fires up an external program, as
with xboard, instead of trying to use a feature of NT 3.1 that does not
exist in later versions of NT or in Windows 95.  ChangeLog for 3.4.pl1
said this had been implemented there, but it really wasn't.

* Indicating player to move by the icon color now works under Windows 95.

* WinBoard now kills the chess program when you exit using the system menu
or the [X] button.  Thanks to Michael Lowe.

* Minor changes to Zippy, including: Now understands wild challenges on
FICS; these used to crash it.  Removed limit on how fast a game will be
accepted.  Use "set formula ..." on the chess server to limit this if you
are using GNU Chess.  Now sends correct opponent name to Crafty.

* Added some missing default values to documentation.  Thanks to Stuart
Cracraft.

* Bugfix: Trying to print "No fonts match pattern" error message would
crash.  This can happen to Linux users who don't install Helvetica, for
example. 

* Bugfix: Defaulting feature in -size n,n,n,n,n,n would put us in an
infinite loop.

* Added more directories to search for Athena widgets on HPUX.

* New cmail (3.12) from Evan Welsh; includes small fix from Kayvan Sylvan.

* Recognize "illegal move" messages from Crafty for illegal castling, etc.

* Bugfix: "name" command added to zippy.c for Crafty was producing error
message if GNU Chess was in use.

* In bughouse mode, suppress holding messages from console window; show
holdings only in banner.

* Minor fixes/cleanup to Makefile.in and configure.in.

3.4, patchlevel 1 -- Mon Dec 11 13:43:12 PST 1995  Tim Mann

* This patchlevel updates WinBoard to match xboard, and includes a few
fixes and minor improvements.  "-size tiny" and "-size n,n,n,n,n,n"
are still not implemented for WinBoard.

* Installed support for Crafty based on code from Bob Hyatt.
Currently the only documentation for this is in the FAQ, and a few
things don't work with Crafty 8.23.  Please do not report these
problems as bugs in either xboard or Crafty.  Bob and I know about
them.  The worst ones (if not all of them) should be fixed in Crafty
8.24.

* Changed the kludge command we send when gnuchess wants to print
something that doesn't end with a newline, from "help" to "bogus".
This works because the error message gnuchess prints ends with a
newline.  It also improves compatibility with Crafty, which doesn't
need the kludge, but for which sending "help" causes a problem.

* Don't draw grid at all if lineGap is 0; previously we drew the grid
with X "0-width" lines (usually 1 pixel wide) and then overwrote it.

* Makefile: Removed xboard.info from "all" target so we won't try to
rebuild it, because this fails on hosts that don't have makeinfo
installed.  Other minor fixes to "clean" targets, etc.

* Fixes and cleanup to Auto Comment code that handles continuation
lines and highlighting.

* Auto Observe now tries to observe the game from the point of view of
the player who was on your gnotify list.  Requested by rng.
Limitations: We can't tell which player it was unless you have ICS
highlighting turned on.  Also, currently "observe foo" works as
required (observing from foo's point of view) only on ICC, but FICS
will probably implement this soon.

* "-size tiny" now makes the default font smaller automatically.  This
was implemented by introducing an extra Form widget in the hierarchy
for all xboard windows, named either normalLayout, smallLayout, or
tinyLayout.  So you can have resource specifications that apply only
to certain layouts; in particular, XBoard*tinyLayout*font.

* Bug fix: EditGame or EditPosition while playing or examining still was
not really being permitted.

* WinBoard bug fix: On the EditPosition menu, King did not work.

* Added text catalog of WinSock error messages to WinBoard, because
Microsoft still has not put them in the system message catalog.

* Removed support for older ICS game-ending messages that do not have
a PGN result token (*, 0-1, 1-0, or 1/2-1/2) after the closing '}'.
The code for older messages was sometimes firing on the newer
messages, due to parsing ambiguity.  If the current code sees an old
message, it will understand that the game is over, but will always
display * as a result token instead of trying to guess the result by
interpreting the text message.

3.4, patchlevel 0 -- Tue Nov 21 01:02:50 PST 1995  Tim Mann

* This patchlevel was released for xboard only.

* Updated the info file.  It should now be as up-to-date as the man page,
with good English except in the parts that pertain only to AmyBoard.

* Added "-size tiny", requested by Bob Hyatt.  Also cleaned up bitmap
support and added "-size n,n,n,n,n,n" to allow arbitrary-sized bitmaps,
if the actual bitmaps are supplied by the user.

* Updated bughouse support.

3.3, patchlevel 4 -- Sat Nov 18 02:27:21 PST 1995  Tim Mann

* Unreleased beta that works with preliminary FICS bughouse code.  Will
need changes before release to track FICS message changes.

* Removed use of .EX macro from man page.  It is not supported by some
nroff -man macro packages, notably the one on Slackware Linux.  Switched
to boldface for references to xboard and other commands within the man
page, as this seems to be the modern way.

* Bug fix: If ICS rejected a move, it was correctly undone on the board,
but the message widget still displayed the bad move.  Reported by DAV.

* Normally, xboard in ICS mode fetches the move list whenever the board
display switches to a new game.  Doing this is now an option
(getMoveList) that can be turned off, which is useful if you are watching
multiple blitz games.  Requested by rng.

* Move list fetching code is now smarter: it ignores a move list if it is
not for the right game.

* Added support for bughouse as implemented on FICS.  Holdings are shown
in the window title in place of the strength numbers.  A menu on mouse
buttons 2 and 3 (same on both) lets you drop pieces.  There is no checking
as to whether you actually hold the piece you are trying to drop; we rely
on ICS to check that.  Notation of the form P@f7 is generated and parsed.
The mate detector does not understand that non-contact mate is not really
mate in bughouse, but this does no real harm.  It results in a "#"
suffix being displayed on the move notation, but xboard does not assume the
game is over.

* Bug fix: Promotion to a knight was not working with ICC!  Thanks to
Wendigo for the report.

* Bug fix: Special pty code for host types *-*-aix3* and *-*-irix3*
(supplied from configure.in) had a bug that would cause childio.c to
fail to compile, due to a "continue" that was not within a loop.

* Bug fix: In pgntags.c, memory was being freed while still in use.  One
symptom this caused was that on some machines, cmail would fail with a
message that it could not find the BlackNA tag.  The error was in some
submitted code that I included in version 3.2.pl3 without reading
carefully enough.  Anders Forberg noticed the symptoms, and Evan Welsh
(who had nothing to do with causing the bug) found the bug and submitted a
fix; thanks to them both.

* Removed restriction against using EditGame while playing, observing,
or examining on ICS.  You still get a warning popup.

3.3, patchlevel 3 -- Sat Sep 16 11:44:05 PDT 1995  Tim Mann

* Bug fix: Going directly from MachineWhite to TwoMachines mode would kill
off the second chess program after Black's first move.

* Added -timeIncrement feature.  Thanks to Joel Rivat.

* Deleted code that tries to keep you from observing more than one game,
or observing while playing or examining.  There is actually no problem in
doing this, except that every time an update comes in from a different
game than is currently being displayed, xboard fetches the history of the
new game, which may be time-consuming if you are on a slow link.

* Fixed configure so as not to crash when neither lex nor flex is found.
lex or flex is needed only if the user wants to rebuild parser.c.  Thanks
to Phil Humpherys for reporting the crash.

* Bug fix: config.h.in used #define instead of #undef for some macros used
in the pty code in childio.c.  This causes the code to fail to compile on
some architectures, because the symbols are supposed to be undefined, not
defined to empty, when they are not set in configure.  Bug originated in
3.3.pl0 when config.h was introduced.  Thanks to Phil Humpherys for report.

* Bug fix: ShowThinking would not show anything when current position had
no move to display in the move window.  E.g., if position was created by
LoadPosition or EditPosition.

3.3, patchlevel 2 -- Mon Aug 28 11:11:11 PDT 1995  Tim Mann

* Zippy code was omitted from xboard-3.3.pl1 by mistake; now included.

* For WinBoard, added hint to help file that you may need to turn off
LocalLineEditing while typing dialing commands to your modem.

3.3, patchlevel 1 -- Sat Aug 19 15:13:30 PDT 1995  Tim Mann

* Zippy distribution is no longer separate from regular xboard distribution.

* Deal properly with Show Thinking output from GNU Chess when it is
thinking on its opponent's time.  In TwoMachines mode this output is
suppressed to avoid interfering with the output from the machine that is
on move; in other modes it is displayed (including the move that GNU Chess
is predicting the user will make next).  GNU Chess produces this output
only if it is built without -DQUIETBACKGROUND defined; this symbol is
defined by default in patchlevels before pl75, but undefined by default in
pl75.

* Bug fix: Handling of initial board position in move list for wild games
was broken, so wild games could not be observed and adjourned wild games
could not be continued.  Thanks to "Maximum Entropy" for the bug report.

* Added feature: algebraic notation now shows "+" indicator for check
and "#" for checkmate, as called for in PGN standard.  Thanks to Kevin
Maher for the suggestion.

3.3, patchlevel 0 -- Thu Jul 27 22:21:07 PDT 1995  Tim Mann

* Changed configuration to use a config.h file instead of passing zillions
of -D options on the cc command line.

* Merged a small fix and some updates to the texinfo file from Jochen
Wiedmann.  The texinfo file still needs work.

3.2, patchlevel 5 -- Tue Jul 18 20:29:39 PDT 1995  Tim Mann

* Beta test release of xboard only.

* Updated WinBoard code to include new xboard features.

* Added texinfo file from Jochen Wiedmann to the release, but not as the
primary documentation.  It needs updating, and the English needs work.
I did make a few improvements, mostly to change incorrect uses of @var to
either @samp or @code as appropriate.

* Merged in code changes to 3.2.pl4beta from Jochen Wiedmann.

* Fixed EditComment; did not pop up window in previous beta.

* Added AutoComment feature.

* Added GameListDestroy to disable the outdated game list popup in cases
where we load a new game file without building a new popup.

* Added yyskipmoves feature to parser.l to speed up building of gamelist.

* gamelist.c wouldn't compile with a non-ANSI compiler.  Fixed.

* Change to yy_text handling in patchlevel 3 still had problems.  Can't
use AC_DECL_YYTEXT in configure.in, because that defines YYTEXT_POINTER
according to whether the lexer on the current host makes yytext a pointer.
But most people will be using a parser.c that was generated on another
host and shipped with the package.

3.2, patchlevel 4 -- Sun Jun 25 19:13:43 PDT 1995  Tim Mann

* Beta test release of xboard only.  

* Added FIREWALLS section to man page.

* Changed -icsport to be a string.  Now with the -telnet option,
specifying -icsport "" suppresses the second argument to telnet.

* Added EditTags feature.  Removed AboutGame from menu, because EditTags
subsumes it.  EditTags suggested by Jochen Wiedmann and first implemented by
him in AmyBoard.  xboard implementation is my own.

* Fixed some missing or incorrect prototypes.

3.2, patchlevel 3 -- Sat Jun  3 18:57:38 1995  Tim Mann 

* Beta test release of xboard only.  

* New version of cmail from Evan Welsh, to fix compatibility problems with
perl 5.0.

* Added game list feature on Load Game, based on code from Jochen
Wiedmann.  Integrated it with cmail.

* Several bug fixes from Jochen Wiedmann, including one to my yy_text
workaround for the difference in the type of yytext between lex and flex.

* Handle clock pause on FICS.

* Suppress clocks in untimed FICS games (time control 0 0).

* Rebuilt configure script with autoconf 2.3.  This fixes a bug in
configuring for X11R6, where -lSM -lICE would not be added when needed.

* Fixed inconsistent type declarations on IntSigHandler and
CmailSigHandler.  Bug report from Josh Daynard.

* backend.c wouldn't compile with a non-ANSI compiler.  Fixed.

3.2, patchlevel 2 -- Tue Feb  7 14:50:30 1995  Tim Mann 

* Minor release of both xboard and WinBoard.

* Added recognition of some FICS messages.  On the other hand FICS is also
changing some of its messages to match what xboard already recognizes.

* Temporarily went back to using "promote" command on ICS instead of
"a7a8=Q", because FICS doesn't implement the latter yet.

* We now avoid using overlapped I/O on pipes in WinBoard, to make Windows
95 beta 2 happy.  This lets WinBoard work with GNU Chess on Windows 95!

* Installed patches from Jochen Wiedmann to coordinate with Amiga XBoard.

* Installed patch to cmail bug in LoadGame from Evan Welsh.

* Bugfix: checkmate and stalemate moves entered with EditGame in ICS mode
were not being handled correctly.  Bug was in GameEnds().

* Implemented EchoOn and EchoOff for xboard, using system("stty echo\n").
Now passwords won't be echoed when you connect directly to ICS.  Also,
telnet negotiation characters aren't displayed (when possible).

* Implemented more of the telnet protocol.  Now connecting to a telnet
server with "-icsport 23" should work even without giving the -telnet
option.  The telnet is in "old line-by-line mode".

3.2, patchlevel 1 -- Sat Dec 10 13:50:46 1994  Tim Mann 

* This patchlevel released for WinBoard only.

* winboard.c: Fixed ConsoleInputThread().  Needed to change CRLF to LF,
not to CR.  This was stopping normal /ics mode from working.  Thanks to
Asher Kobin for the bug report.

* winboard.c: Fixed Raw(), EchoOn(), EchoOff().  Now they take effect
immediately, not on the next console read after the one in progress.

* winboard.c: Attempted to make WinBoard work with gnuchessx running
directly on Windows (not remotely via rsh).  It now works on NT, but only
if gnuchess is told not to think on its opponent's time ("easy\n" removed
from initString).  The problem seems to be that GenerateConsoleCtrlEvent
is not doing anything.  On Windows 95 beta 2, we get error messages on
both reading and writing to gnuchessx; I didn't investigate why.

3.2, patchlevel 0 -- Wed Dec  7 13:23:36 1994  Tim Mann 

* Thanks to all the beta testers who gave me feedback: Josef Nelissen,
Steve Booth, Evan Welsh, Dima Dakhnovsky, Chris Petroff, Peter Jansen,
Derek Terveer, Michel van der List, Richard Lloyd, Shelly Mistry, and Mike
Lee.  Sorry if I forgot anyone.  Thanks to Virendra Kumar Mehta for
information about DYNIX/ptx.

* Don't exit on keyboard EOF unless we get two in a row.

* WinBoard only: added -localLineEdit switch to allow turning off local
line editing if you really want to.  It is still a bad idea to let the
echoing be done remotely, however; see below.

* Fixed some configure problems on HP-UX.  [Steve Booth]

* Fixed (I hope) configure problem on SunOS 5.3/Solaris 2.3.  [Josef Nelissen]

* cmail bugfix from Evan Welsh (cmail 3.4).

3.1, patchlevel 9 -- Fri Dec  2 23:54:56 1994  Tim Mann 

* Beta distribution only

* Always do local echo/edit of user typing in ICS mode.  Doing the echo
downstream may seem nicer in some modes, and it can be hard to turn that
echo off, but the echoed characters can be interleaved with ICS output
and make it impossible to parse correctly.  For xboard this involved only
a change to recommendations in the man page, as Raw() isn't implemented.
For WinBoard, removing Raw() made a real difference.  Added code in 
WinBoard to change /r/n back into /r on keyboard input, as we get the 
former when Raw() is not called.

* Do not issue ICS "refresh" command after we start to observe a game
unless we get to the next prompt without seeing a board image.  Newest
version of ICS doesn't require this refresh, but old versions around still
do.

* cmail bugfix and small code cleanup in LoadGame, from Evan Welsh.

* Added keyboard accelerators N/P for LoadNextGame/LoadPreviousGame.  Evan
Welsh request.

* Using "-" on the command line as a filename for loading (saving) games
or positions specifies the standard input (standard output).  Alain Picard
suggestion.

* On WinBoard only, a command line option without a leading '-' or '/' is
now taken as the value of -lgf.

* Changed to not use stdin, stdout, stderr as initializers in backend.c;
needed for GNU libc compatibility.  You also must build parser.c with flex
(not lex) if you are using GNU libc, to avoid having the same problem
there.

* Changed WinBoard to avoid using "overlapped" input on the console.  It
now seems to fully work on Windows 95 beta 2.

* Improved comment popups on WinBoard.  Now newlines are handled properly,
and the plain Comment popup window doesn't disappear and reappear when we
step to a new move with a new comment.

* Fixed bugs in detecting the absence of the time and otim commands.

* Added built-in implementation of rcmd protocol to WinBoard.  Windows NT
does not implement passing signals through rsh, and Windows 95 does not
have rsh at all.

* Added -remoteUser option.

3.1, patchlevel 8 -- Mon Nov 28 15:26:07 1994  Tim Mann 

* Beta distribution only

* Rearranged ChangeLog file into reverse chronological order to be closer
to GNU standards.

* Integrated new cmail code from Evan Welsh (including cmail RCS rev 3.2).
Includes a bug fix to TruncateGame.

* Updated ICS address to be chess.lm.com.

* Bug fixes to handling the aftermath of FatalError.  Thanks to Chris
Petroff for the bug report.

* Test for remsh before rsh, other fixes for HP-UX.  Thanks to Richard
Lloyd.  I wasn't able to do all the things he suggested, so there may
still be some rough edges in building on HP-UX.  See the FAQ file for hints.

* Bug fix; added missing check for HAVE_SYS_SYSTEMINFO.  Thanks to Josef
Nelissen for testing on Solaris 2.x.

* Updated WinBoard to match xboard.  (WinBoard still has a few option
dialogs that don't exist in xboard.)

* Changed Hint output to a popup.

* ShowThinking output and move output no longer overwrite each other.
ShowThinking output won't appear if the displayed position is not current. 
ShowThinking output in TwoMachines mode made clearer and documented.

* Implemented --enable-ptys and --disable-ptys arguments to configure.

* Fixed Book and Hint code to work over a pty with echo enabled and tabs
expanded to spaces.  Thanks to Dima Dahknovsky for the bug report.

* Moved Attention calls from all over xboard to one place, inside
SendToProgram.

* Added bulletproofing to ShowThinkingEvent.

* Added code to handle "refresh N" boards that come in from ICS properly,
assuming ICS is changed to mark them with a new relation code (-3).

3.1, patchlevel 7 -- Sun Nov 13 22:16:01 PST 1994 -- Tim Mann

* Beta distribution only

* Changed ShowThinking to just show the current best line in the
DisplayMessage area, instead of dumping everything to stdout.

* Installed new cmail (RCS rev 3.1) and cmail.man (RCS rev 1.10), and
changes to cmail code in backend.c, from Evan Welsh.

* Miscellaneous minor fixes.

3.1, patchlevel 6 -- Fri Nov  4 12:53:53 PST 1994 -- Tim Mann

* This patchlevel was not actually released to anyone.

* Updated the pty code to be based on GNU Emacs 19.24, and moved it to a
separate file.  It was hard to split out just the pty configuration from
all the stuff emacs does with its custom configure script and .h files,
but I did my best.

* Converted from imake to GNU autoconf.  This was a serious upheaval.  

* Put in code to help trap "error gathering move list" problem reported by
Michel van der List if it recurs.  I couldn't reproduce it.

3.1, patchlevel 5 -- Mon Oct 31 21:12:00 PST 1994 -- Tim Mann

* Beta distribution only.

* We now test for checkmate or stalemate in EditGame mode after every user
move, and in LoadGame mode whenever we hit the end of a game without
seeing a PGN end marker.  cmail needs an update to deal with this
correctly; Evan promises one.

* Bugfixes in new move generator.  Thanks to Mike Lee for reporting one of
the bugs.

* Imakefile was omitted from patchlevel 4.

3.1, patchlevel 4 -- Mon Sep 19 18:19:46 PDT 1994 -- Tim Mann

* Beta distribution only.

* The move generator includes a mate tester.  Initially this is used only
by cmail, and even that usage needs further work.

* Wrote a true move generator and used it to replace all the move
disambiguation and legality checking code in parser.l.  The move generator
is capable of dealing correctly with en passant and castling availability,
but the rest of the program still does not keep track of this information.

* Bug fix: xboard did not handle "foo has made you an examiner of game 23"
message.  Thus you could not use examine features until the next board
came in, showing your new relation to the game.  Thanks to POOKIEWOOKIE on ICS
for the bug report.

* Added AutoObserve feature.  Thanks to Chris Petroff for the idea.

* Added Book feature to use new gnuchess "bk" command.  Mike McGann request.

* Redid code to handle missing "time" command in gnuchess, because latest
gnuchess no longer sends a response to this command.  

* Eliminated need for -DFLEX.  Thanks to Michael Shields (Vladimir?) for
the idea.

* Added missing code to implement MoveNow in TwoMachines mode.

* Added ShowThinking feature.  Thanks to Richard Lloyd for the idea.

* Applied patches from Evan Welsh; some fixes and improvements to the
cmail code.

* Fixed bug in moving from EndOfGame mode to MachineWhite or MachineBlack,
introduced in previous patchlevel.  Also fixed related bug in ending a
game in ICS mode; was entering EndOfGame mode instead of IcsIdle.

* Added patch to implement internetChessServerLogonScript flag, from Kevin
O'Connor.  Thanks!

3.1, patchlevel 3 -- Wed Sep  7 13:22:07 PDT 1994 -- Tim Mann

* Beta distribution only.

* Merged in new cmail code from Evan Welsh.  He added the ability to have
more than one game per message, needed for official IECG matches.  I added
the ability to resign or offer/accept/decline a draw in a cmail game.

* Bugfix: invalid -tc option caused segmentation fault; DisplayFatalError
was called too early in initialization.  Georges Honore reported this bug.

* Decided to keep EndOfGame mode as an element of the user interface---it
means that a gnuchess game or loaded game has ended, and the user must
explicitly select EditGame to edit it, rather than just being able to
enter more moves freely.  But internally there is no longer an invariant
tying this mode to whether the chess program is running.

* Zippy bugfix: Was saving only the final position in the -sgf file when
the game ended by something other than resignation or flag.  Also fixed
the longstanding bug that Zippy would think the final board of such a game
was a new game and restart the chess program.

* Now does a better job of faking castling availability in FEN.  We still
don't really keep track of it, but now at least we don't say that castling
is still available when the king or rook is not on its home square.

* Bugfix: Initial board of game history for wild games was going through
too much processing, causing us to forget the game length, which is now
needed by ParseGameHistory.  This was causing problems with resuming
adjourned wild games.

* Updated ICS host to ics.onenet.net.

* Zippy now accepts challenges where the opponent specified his color.

* Added ZIPPYPASSWORD2 to let operator give commands directly to gnuchess.

* Bitmap directory can include alternative icons now, too.

* Handle "Game * (*) has no examiners" message from ICS.

* Revamped window title and icon name selection.

3.1, patchlevel 2 -- Sun Jun 12 17:16:28 PDT 1994 -- Tim Mann

* "Beta" distribution for Zippy users only.

* Bugfix: common.h was assuming that X11 type Boolean is char, which is
not always true.  This might have caused all sorts of obscure bugs!

* Installed new bitmaps from Elmar Bartel as the default.  Many thanks!
The old bitmaps can be used by changing the "bitmaps" symbolic link before
compiling xboard, or by using the -bitmapDirectory option at runtime.

* Loading a game that ends with the PGN unfinished symbol ("*") now always
leaves you in EditGame mode.

* Added documentation of -icscomm to man page.  Thanks to Maarten Remkes
for the linux script.

* ^C now kills gnuchess as well as xboard.  Thanks to Dima Dakhnovsky for
reminding me how to do this.

* Bitmap icon color now indicates player to move.  Stuart Cracraft
suggestion.

* Changed piece bitmap flags to have just one flag, which points to a
directory full of bitmaps, instead of a flag for each bitmap.  Also
changed bitmap naming convention.

* "Connection closed by ICS" is no longer a FatalError popup.  This was
too annoying in the normal case where the user typed "quit".

* Changed default font to 14 pixels instead of 10 points.  This seems to
make it close to the size I want it to be on more displays.

* Major mode (GNU Chess, etc.) appears in title bar.  Stuart Cracraft
request.

* Bug fix: EditGameEvent, MachineBlackEvent, MachineWhiteEvent, and
TwoMachinesEvent were calling PauseEvent to get out of pause mode, which
now has undesired side effects, such as advancing the display to the
forwardMostMove.  It works fine to just set pausing = FALSE instead.

* IcsExamining mode now lets you access ICS edit position commands using
the same popup menus as xboard's EditPosition mode.  Thanks to DAV on ICS
for inspiring the idea.

* Corrected test for gcc on HP in Imakefile; thanks to Richard Lloyd.

3.1, patchlevel 1 -- Wed Jun  1 16:25:11 PDT 1994 -- Tim Mann

* Added parser.c.lex and parser.c.flex to the distribution.

* Added HP gcc options to Imakefile, from Mats Nylen.

* EndOfGame mode no longer highlights EditGame mode indicator, because the
modes really do differ.  I would like to get rid of EndOfGame mode in the
future, at least as far as users can see.

* Larger %a in parser.l, needed for RS/6000 users.

* Minor improvements to Mail Move error messages.

* When -debug flag is given to xboard, -v is passed to cmail.

* Old Save Style uses "1. ..." instead of "1..." when black moves after a
comment; more like xboard 3.0's actual style.

* New version of cmail from Evan Welsh.  Looks for UCB Mail in a more
portable way and has some minor bug fixes.

* Bug fix: Saving a FEN position while in EditPosition mode with black to
play was showing white to play in the saved position.

3.1, patchlevel 0 -- Fri May 20 16:36:15 PDT 1994 -- Tim Mann

* This is the first general release since 3.0, patchlevel 9.  Releases
since then have been limited-distribution or beta releases.

* Thanks to my version 3.1 beta testers: Dmitry Dakhnovsky, Ed Hanway,
Richard Lloyd, Mike McGann, Shelly Mistry, Josef Nelissen, Chris Petroff,
Jack Robertson, Michel van der List, Ky Macpherson, Derek Terveer, and
Evan Welsh.  Sorry if I've forgotten anyone who gave me feedback.

* Updated man page.

* Minor cleanup on menu sensitivity code.

* Integrated another even better version of cmail from Evan Welsh,
including some code he supplied in xboard itself.

* A few more fixes to cmail support.

* Automatically update clocks after an ICS "moretime" command.

* Handle ICS automatic examine mode (set examine 1) after a game.

3.0, patchlevel 14 -- Tue May 17 13:41:44 PDT 1994 -- Tim Mann

* Beta test release only.

* Don't suppress prompt on first board of game being examined; otherwise
it looks like nothing happened at all (i.e., like you're lagged).

* Added StopExamining and StopObserving (Chris Petroff suggestion).

* Made Reset do a refresh on ICS.  John Chanak's original ICS code for
xboard tried to do this; I finally decided it was a good idea after all.

* Fixed bugs in handling updates that come in while you are pausing in
examine mode.  This is pretty tricky to do right.

* Made un-Pausing immediately pop you to the current position.  This is
always necessary in examine mode, since the game may have changed under
you; your remembered moves might no longer be right.  So I made it do the
same in all modes.

* Took Detach Examine mode back out...used Pause mode for this instead!

* Put in a trap for the GNU Chess bug of printing an Illegal Move message
when its own hint move is illegal.

* Attempted to fix a Zippy bug; bogus "exited unexpectedly" messages after
it checkmates its opponent.  This involved adding an argument to GameEnds
to say who says it ended (ICS, GNU, etc.)  and taking different actions in
different cases.  Unfortunately, this just restored an older Zippy bug, where
Zippy restarts gnuchess when it gets the board with the final position,
because GameEnds still puts xboard in EndOfGame mode.  Sigh.

* We now capture elapsed time on last move when getting game history.
(Josef Nelissen bug report)

* Rewrote code to redisplay last file title so it really works.

* Fixed bug in oldSaveStyle.

* Larger %a and %o were needed for lex.

3.0, patchlevel 13 -- Mon May 16 16:26:22 PDT 1994 -- Tim Mann

* Beta test release only.

* FatalError popups stay on the screen now.  If the error is really fatal,
all functions are disabled, and the program exits when you press OK.

* Redisplay last file title when using Load (Next/Previous/Same) Game.

* Zippy understands new match challenge message format on ICS.

* Added Revert command.

* When examining a game on ICS, < > buttons do ICS backward/forward
commands, unless you set the Detach Examine option.  Thanks to Dima
Dakhnovsky for the idea.

* Fixed various minor problems in cmail mode.

* cmail generates the tags now.  New version of cmail from Evan Welsh.

* cmail mode won't let you do MailMove unless the currently displayed
position is exactly one move past the end of the game you loaded.

* Fixed building of man pages in Imakefile

* Updates to INSTALL file

3.0, patchlevel 12 -- Sat May  7 21:10:03 PDT 1994 -- Tim Mann 

* Beta test release only.

* Brought man page up to date.

* Added Shift+R to resign from keyboard.

* Some items on Action menu available in GNU Chess mode now.

* Revamped Action menu.

* Reordered functions in backend.c as a small step toward reorganizing
this whole mess.

* Generate TimeControl PGN tag.

* Get type of ICS game (e.g., rated blitz) and save in PGN tags.

* Bug fix: Clocks were not redisplayed when entering EditGame mode.

* Bug fix: Clocks were not being redisplayed after loading a game file
with -td 0, so they could show the wrong color active.

* Bug fix: Chess programs would be killed and match mode would exit
prematurely when loading a PGN game fragment ending with "*".

* ICS command "sposition" no longer confuses xboard.

* Integrated new version of cmail (with support code in xboard) from Evan
Welsh.

* Added TruncateGame, MoveNow, RetractMove, and QuietPlay.

* Bug fix: switching between MachineWhite and MachineBlack was not calling
Attention().  Reported by Dino Dini.

* More improvements to INSTALL and Imakefile.

* Pack moves into 79 character lines in PGN output.  We don't generate
check indications, and we always break the line before the result,
so this is not quite PGN export format.

* Use FEN tag in PGN.

* Improved finding and counting of game starts in save files.

* Negative position or game numbers in -lpi/-lgi mean to seek to that byte
offset.  Hook for possible future features, not in man page.

* Detect and handle absence of either "time" or "otim" commands in
gnuchess.

* Don't use "promote" command to ICS anymore.

* Handle switching sides and taking back moves on FICS.

* Handle flip state flag in style 12 board.

* Handle examine mode on ICS.

* Improved error popups for various kinds of illegal moves.

* Suppress unasked-for hints from gnuchess (which it generates in post
mode).

* Load/save position functions use FEN.  Old style also supported.

* Added detailed error messages if loading a bitmap file fails.

* Small board has 2-pixel lines between squares instead of 3-pixel.

* Added OldSaveStyle and AboutGame.

* Renamed ForceMoves to EditGame.

3.0, patchlevel 11 -- Tue Sep 21 15:25:36 PDT 1993 -- Tim Mann 

* The following changes were present in xboard 3.0, patchlevel 11, but the
first group did not make it into WinBoard 3.0 until later.  xboard
3.0.pl11 was a limited-distribution release only, mostly to Zippy users.

* Man page minor fixes.

* Added a missing file close.

* Removed automatic error popdown on Reset, which was destroying some
error messages before they could be read.

* Fixed char vs. unsigned char warnings on bitmaps.

* Use ICS board style 12.  Some improvements to ICS parsing.

* Comment window is now labelled with the move the comment is on, and
comments don't pop down when you step to the next move.

* Save files now in PGN format.

* Support for loading PGN files.  PGN tags pop up when a PGN game file
is loaded.

* More info in INSTALL and Imakefile about building for Suns (and
other systems).

* WinBoard 3.0 patchlevel 11 split off from an early version of xboard
3.0 patchlevel 11, so it does not have all the features of that
patchlevel.  The following changes made it into both xboard and
WinBoard:

* Made Comment dialog non-modal in WinBoard.

* EndOfGame mode is now more transparent.  It looks like ForceMoves
mode except that there is no chess program running.

* Small bug fixes in clock management.  Most noticeably, pausing when
it is gnuchess's move now works as documented (again).

* Add minimal support for -icscomm option.  Not documented yet because
I haven't sorted out the issues with setting the tty modes on the comm
port device, or with locking it properly.  But it's usable by wizards.

* Don't restart the chess program upon Backward event from EndOfGame mode.

* Suppress extra prompt after ICS sends us a board.

3.0, patchlevel 10 -- Sat Sep 11 18:44:03 PDT 1993 -- Tim Mann 

* Beta test release only.

* Zippy now plays chess.  zippy.c and zippy.h are still not included
in the standard distribution, but are available on request.

* Switched to using style 12 on ICS.

* Updated man page, and documented use of XBoard*form.translations to
add more shortcut keys.

* Added shortcut keys "d" to claim/offer/accept a draw, and "t" to
call flag.  Suggested by venu on ICS.

* More explicit instructions in INSTALL file.

3.0, patchlevel 9 -- Tue Sep  7 14:02:00 PDT 1993 -- Tim Mann

* General release, minor update to 3.0.pl8.

* Loosened checking on whether it's okay to start a move, to satisfy
ICS ultra-blitz players.  Now we don't check whether it's your turn
until you let go of the piece.

* Parser now recognizes "+-+" as meaning the game ended in a draw.

* Got rid of S_NONE symbol, which seems to conflict with some symbol
Sun defines.

* Tweaked the man page.  Clarified that there is currently no way for
two people running copies of xboard to play each other without going
through the Internet Chess Server.

* Fixed a bug in color name conversion.  Asking for two different
colors whose names were the same in the first four characters would
get you two copies of the first one.  Thanks to Volker Zink for the
bug report.

* Improved confusing Usage() message.

* Added a bunch of Sun information to the Imakefile and INSTALL file.
Thanks to Ed Hanway, Arik Klingensmith, and others who responded.

* Test for defined(WIN32) instead of !defined(unix).

* Avoid using (void *) type with non-ANSI C compilers.  Thanks to
James Altucher for the bug report.

3.0, patchlevel 8 -- Thu Sep  2 12:23:01 PDT 1993 -- Tim Mann

* Note: Patchlevel 8 was the first non-beta release of xboard 3.0

* Added cmail to distribution.  Contributed (and still maintained) by
Evan Welsh.

* Bug fix: -queen option wasn't initializing menu check.  Reported by
Pat Surry.

3.0, patchlevel 7 -- Thu Aug 26 13:23:24 PDT 1993 -- Tim Mann

* Sent WinBoard 3.0.pl7 to Torre on ICS

* Added parser.h to hold the interface to parser.l

* Upgraded COPYING file and copyright notices to GNU GPL version 2.

3.0, patchlevel 6 -- Tue Aug 24 15:16:13 PDT 1993 -- Tim Mann

* We no longer display intermediate positions or intermediate comments
while loading a game file with -timeDelay 0 or loading an opening to
start up a -matchMode game.

* Loading an empty game from an xboard save file now gives a status
message "No moves in game" instead of an error popup saying "Game not
found in file."

* Added comment.awk to distribution.

3.0, patchlevel 5 -- Tue Aug 17 16:45:54 PDT 1993 -- Tim Mann

* Bug fix:  xboard would crash if it couldn't get all its colors; now
it switches to monoMode instead.  Also, xboard was trying to convert
color resources even when using a b/w display.  Reported by Larry
Rogers.

* Bug fix: Declared fields of TimeMark as signed so that we get signed
instead of unsigned arithmetic.  SubtractTimeMarks was breaking on
Alpha AXP (which has 64-bit longs) with old declarations.  Reported by
Michel van der List.

* Bug fixes:  Keyboard accelerators now work after EditComment window is
popped down.  Iconize keyboard accelerator now works even if xboard
was started with -iconic flag and later deiconized.

* Bug fix:  The routine that tests whether a move is illegal because it
would leave you in check was not handling e.p. captures properly.
Reported by Patrick Surry.

* Bug (?) fix:  Was adding time to clocks at time control even when
loading a game file.  Actually it's not entirely clear how time
controls should be dealt with when some moves are loaded from a file
or clicked in with ForceMoves.  For now ForceMoves mode does add the
time (because it can be used to change moves during a live game), but
LoadGame mode does not.

* Updated usage message.

* Previous attempted fix to matchMode had broken TwoMachines mode and
generally needed more work.  Also simplified command line interface to
matchMode.

3.0, patchlevel 4 -- Thu Aug  5 14:17:18 PDT 1993 -- Tim Mann

* Thanks again to the beta testers listed for 3.0 patchlevel 3, and also
Desnogues, Steve Cariglia, Niklas Engsner, Mark Silver, and Roger Rowe.

* Reorganized man page, splitting OPTIONS into subsections.

* matchMode was very broken; fixed.

* Changed convention for turning off command line options from --opt
to -xopt, to be less inconsistent with GNU standards.  Also changed
the long command line options to take True/False arguments like
resources; seems to make more sense this way.

* Added AlwaysQueen option -- suppresses promotion dialog and always
promotes to a queen if you move a pawn to the last rank.  Has no
effect on gnuchess (or your ICS opponents!) -- they can still
underpromote.

* Subtracted an extra fudge term when determining how wide message and
title widgets should be.  This fixes a problem some beta testers had.
Wish I knew why it's needed.

* Bugfix in parser; symptom was that you couldn't do LoadGame after
observing a game on ICS.

* Bugfix in Forward; didn't work after game ended while Pause was turned on.

* Removed bogus execute bits on .h files.

* Use REMOTE_SHELL and TELNET_PROGRAM definitions.

* MachineWhite and MachineBlack now work from TwoMachines mode.

* Popping down an error message with the [ok] button was not turning off
the errorUp flag, so the next move would cause xboard to try to pop it
down again, resulting in a wild memory reference and sometimes a crash.

3.0, patchlevel 3 -- Tue Aug  3 17:40:27 1993  -- Tim Mann

* Thanks to my beta testers: Patrick Surry, Takuya Kojima, Robert J. Luoma,
Chris L.  Petroff, Richard K. Lloyd, Michel van der List, Craig Metz,
Antoon Frehe, Simon Clift, Shelly, Eric Peterson, Christopher Mitchell,
Martin Koch, Ed Hanway, Steve Booth, Udo, Ken Hobday, and Joseph Duhamel.

* Improved error messages for trying to move the wrong color pieces or
to move when it's not your turn.

* Special code for monoMode on 1-bit displays now understands displays
where 1=white and 0=black.

* Declare getenv() if <stdlib.h> not included; avoids a compiler warning.

* Documented borderXoffset and borderYoffset.

* Added -titleInWindow option for use with X window managers that
don't let us set the title in the window banner.

* Fixed error message printing in WinBoard; system error messages no
longer appear as numeric codes.

* The error message popup is now non-modal; you don't have to press the
[ok] button before you can do something else.  In addition, the popup is
positioned so that it doesn't cover up the board (too much), and making a
move or otherwise clicking on the board pops it down.  (Not implemented in
WinBoard.)

* You can now call your opponent's flag in ICS mode by clicking on his
clock. 

* Fixed minor bugs in -flipView option and documented exactly how xboard
decides which way to flip the view.

3.0, patchlevel 2 -- Fri Jul 30 22:20:23 PDT 1993 -- Tim Mann

* Added Autosave to Options menu.  Would be better to have Save
Options dialog as in WinBoard, but this was quick to do and gives the
most-needed functionality.

* Changed "Reload Game" on menu to "Reload Same Game".

3.0, patchlevel 1 -- Thu Jul  8 21:22:59 PDT 1993 -- Tim Mann

* Sent a copy of patchlevel 1 to Patrick Surry to beta-test.

* Added -cmail option that sets appData.cmailMode.  Currently a no-op.
In the future this may set special modes for use by the cmail script
for playing chess by email.

* Added LoadNextGame, LoadPreviousGame, and ReloadGame to File menu.
ReloadGame suggested by Patrick Surry.

* Added -flipView command-line option.  Suggested by Patrick Surry.

* Fixed bugs in parser.l:  (1) Pattern for "# xboard game file ..."
needed to match to end of line.  (2) Start of a new file was not matching
the ^ start-of-line character.  The fix for this is a kludge.

* Made game counting code in LoadGame more robust, and made LoadGame
able to detect the end of a saved partial game (by noticing the start
of the next game) in game files created by XBoard itself.  We don't
try to find the start of the next game that way in other kinds of game
files, because the only way I can think of to do that is to look for
another move #1, and that technique gets too many false hits.

* Fixed recently introduced bug in LoadGame when game starts with a
position diagram.

3.0, patchlevel 0 -- Fri Jun 25 14:17:17 PDT 1993 -- Tim Mann

* Changes in this patchlevel were too numerous to list.  Larger ones are
listed below.

* Added a popup dialog to enter and edit comments.  Inspired by some
code from Patrick Surry.  Changed the normal read-only comment popup
to the same style.

* Added ICS init script feature from Karl Schwamb.

* Added some ESIX fixes and OMIT_SOCKETS ifdef option, from Kayvan Sylvan.

* Revamped code to allow use of flex instead of lex on parser.l.
Using flex requires adding -DFLEX to defines in Imakefile.

* Source code is split into front end (xboard.c), which knows about X
and Unix, and back end (backend.c), which knows about chess, gnuchess,
and the ICS.  There is also a front end for Windows NT.

* Boolean command line options now use "-foo" to turn on and "--foo"
to turn off instead of "-foo true" and "-foo false".  [Later -xfoo;
see above.] 

* Added menu commands to control autoflag, bell, and coords options.

* User interface has a new look: (1) Menu bar instead of array of
buttons.  A few very commonly used features have small buttons in
addition to being on the menus. (2) Large font for clock.  (3) Pop-up
dialogs for errors.

2.1, patchlevel 11 -- Sat Jun  5 00:01:01 PDT 1993 -- mann@src.dec.com

* Added code to deal with "wild" games on ICS.  This includes allowing
castling with the king on d1 or d8, which is allowed in wild(1) games
if the king started there.  Notation is o-o to castle "short"---to
whichever side the king is closer to---and o-o-o to castle "long."
Right now wild castling is always allowed by xboard; we rely on ICS or
gnuchess to reject it when we aren't really in wild mode.

* memcpy call had arguments in wrong order.

* Removed April Fool code

2.1, patchlevel 10 -- Mon Feb 15 10:19:31 PST 1993 -- mann@src.dec.com

* Avoid trying to select on a pipe when using System V.  Needed to
copy some code from InitChessProgram() up into establish().

* Bug fix in disambiguating pawn moves like "ed".

* Fix to error handling in ReceiveFromProgram.

* Bug fix: entering EditPosition mode with black to play highlighted
White's clock instead of Black's.

* Added telnetProgram resource in case "telnet" is not the name of the
telnet program.

2.1, patchlevel 9 -- Fri Jan 22 19:08:27 PST 1993 -- mann@src.dec.com

* Entering Force Moves mode clears "flag has fallen" messages from display.

* ICS host name can now be in numeric format; for example 128.2.232.4.

* Bug fix: LegalityTest was failing to test whether a pawn move was
illegal because the move would discover a check.

* Handle ICS message "mann asserts a win over manntest, who disconnected."

* Bug fix: A recent change to PromotionCallBack had broken
underpromotion to a knight.

* Bug fix: In EditPosition mode, dragging a piece onto a square border
would make it vanish.  xboard wasn't distinguishing this case from
dragging the piece off the board.  Thanks to Matthew Kidd.

* Removed "static" declaration from yywrap for compatibility with IRIX
version of lex.  Thanks to stiller@blaze.cs.jhu.edu.

* Added substitutes for bzero, bcopy, and gethostname for Solaris
SVR4.  Thanks to Michael Grant.

* Bug fixes to queen move disambiguation.  Bugs showed up only when
promotion resulted in more than one queen on the board.

2.1, patchlevel 8 -- Fri Dec 11 17:54:18 PST 1992 -- mann@src.dec.com

* parser bug fix: It now really works to leave off the piece a pawn is
promoting to and let it default to queen.

* When starting to observe or play an ICS game, don't draw board in
initial position and then immediately redraw it in the current
position.

* Handle ICS messages when an "abuser" forfeits a game by
disconnecting and when a game is aborted ("removed") by an
administrator.

* Bug fix: A user move when in LoadGame+Pause mode was resuming the
game load instead of putting us in force mode.

* It seems that crashes inside sscanf when xboard is compiled with gcc
(as on IBM PS/2 AIX, mentioned below, and also on VAX Ultrix 3.1) are
caused by an incompatibility between gcc and the sscanf implementation
on these platforms.  A workaround is to specify -fwritable-strings to
gcc (see the gcc documentation).  Added info on how to do this to the
Imakefile.  Thanks to Tom McConnell for this information.

* Clocks are allowed to go negative.  This is mostly for ICS
compatibility, but it affects gnuchess mode too.  Also, we give the
time bonus when a player hits the time control boundary even if his
flag is already down.  This choice is a bit debatable, but it makes
things look better when you are in TwoMachines mode and gnuchess's
time management screws up causing it to exceed its time limit.

* Implemented move legality checking code.  Moves made with the mouse
or parsed from a file are checked for legality before being made.
This is mostly in preparation for future extensions, such as
human-human play.  For now it makes -noChessProgram mode more useful.

* If the user takes back moves, we restore the clocks to the earlier
settings.

* We now handle the output of the ICS oldmoves command, including
parsing the game end condition.

* Added autoCallFlag mode.

* Fixed glitches in the medium size outline king and rook bitmaps, and
touched up the medium size solid king and outline queen.

* bug fix: DisplayMove would not display backwardMostMove - 1.

* Added CallFlag button and removed AcceptMatch.  CallFlag is more
important with new ICS, and accepting the current match offer is easy
with the new ICS command "accept" (can be abbreviated "ac").

* LoadGame and LoadPosition display tail of file name (plus index
number, if any) as title.

* MachineWhite, MachineBlack, and TwoMachines now work properly from
LoadGame and LoadGame+Pause mode.

* Use o-o and o-o-o to castle on ICS, not 00 and 000.

* Stripped out code for old ICS messages; new ICS is now installed.

* Bug fix: don't offer autosave when paused during LoadGame.

2.1, patchlevel 7 - Fri Dec 11 17:40:56 PST 1992 - mann@src.dec.com

* Track change to "Illegal move" message in new ICS.

2.1, patchlevel 6 -- Tue Dec  8 10:48:44 PST 1992 -- mann@src.dec.com

* Kludged around bug in keeping comment popup where you put it by
adding borderXoffset and borderYoffset resources that give the width
of the borders added by the window manager.  Yucch.

* Changed searchDepth kludge back to using "help" instead of "bd"; the
latter didn't work because the output contains a line starting with
"White", so we think gnuchess is telling us that White won.  Oops.

* Removed some code that uses an X11R5 feature, XrmGetDatabase.
Without this, the auto font sizing code is harder to make work, so to
keep my sanity I had to change it to affect only the fonts that appear
in the main window and the comment popup, not the other popups.  Maybe
this is better anyway.

* AcceptMatch button now works after a counterchallenge, too.

* Added code for more variants of messages about games being adjourned
or aborted.  These messages need to be unified in new version of ICS.

* Added workaround for minor bug in ICS; game number on first board
can be wrong.

* Removed code to say "refresh n" instead of "refresh"; was broken and
wasn't needed anyway.

* Added missing casts for compilers that distinguish enums from ints
and missing cast in connect call.

* Added code to recognize "both sides are out of time" draw message
proposed for next version of ICS.

* Thanks to Danny Sleator and Joe Peterson for bug reports.

2.1, patchlevel 5 -- Sun Dec  6 19:52:40 PST 1992 -- mann@src.dec.com

* Removed gcc and CDEBUGFLAGS from Imakefile.

* Corrected setting of mode to ForceMoves when game file ends or
contains an AmbiguousMove or BadMove.  Previous bug was harmless.

* Added AcceptMatch button and mention of 50-move rule for Draw button
to man page.

2.1, patchlevel 4 -- Sun Dec  6 02:55:42 PST 1992 -- mann@src.dec.com

* Now handles revised messages from new version of ICS.  New ICS is
not released yet, so this code might change further.  Next patchlevel
will remove support for old ICS; both are there now.

* Initial processing for a new game being watched or played now
happens when we see the first board image.  Removes the need to parse
some messages and unifies some code.

2.1, patchlevel 3 -- Tue Dec  1 19:40:40 PST 1992 -- mann@src.dec.com

* Handles ICS messages "Draw : neither player has mating
material" and "Draw : White has no material, Black has no time."

* Added AcceptMatch button for ICS mode.

* Fixed bug in previous fix to yylexstr().

* Fixed bugs in code for loading old position files that don't start
with "#" and for handling case where user asked for nth position in
file but there aren't that many.

* Handles ICS message "* has restored your old game"

* ResetProc always clears title line now.

* Don't issue just "refresh" when watching a game; use game number.
Upcoming new version of ICS may need this.

* Added code to try to prevent user from watching and/or playing more
than one game at a time.  There is a race condition inherent here; if
we get more than one board from the game before our command to stop
watching it takes effect, we'll think the user started it again.  Not
clear how to fix this.

* Redid ParseBoard8 to use sscanf.  Code is a good deal cleaner now.
Also, we now parse out the game number too, though we don't make much
use of it yet.

* (Tried to put in a feature that recognizes the current game in a
game file must have ended when we see the start of a new one, but had
to disable it because we have too many false hits with the current
parser, especially in gnuchess listing files.)

* Moves read from game files or received from ICS are now translated
into canonical algebraic form just like all other moves.  Minor nit:
If you use Reset while playing or observing a game in ICS mode, the
game history (including the current board position) is lost, so the
next move of the current game can't be translated.

* Default fonts now vary with board size, and are chosen by pixel size
instead of point size (since piece bitmaps have a fixed pixel size).

* Bug fix: An extra Forward was required to skip over time indications
in game files.

2.1 patchlevel 2 -- Fri Nov 27 23:30:00 PST 1992 -- mann@src.dec.com

* If you move the comment popup, the next time it pops up it will come
back where you put it.

* Fixed yylexstr() so calls to it can be interspersed with calls to
yylex().

* Fixed bugs in Forward/Backward while in LoadGame mode.

* Changed Save{Game,Position} functions to append instead of
overwriting if file exists.  Changed Load{Game,Position} functions to
deal with multiple games/positions per file.

* Changed load{Game,Position}File resources to trigger automatic load
on program startup.  Changed save{Game,Position}File resources to
trigger automatic save after every completed game and on program exit.

* Added autoSaveGames mode.

* Fixed bug with PromoPiece in call to MakeAlg from MakeMove.

* Many changes to get rid of picky compiler warnings and generally
clean up the code.

2.1 patchlevel 1 -- Fri Nov 27 02:45:00 PST 1992 -- mann@src.dec.com

* A last minute change before patchlevel 0 went out broke
ParseGameHistory().  Immediately fixed in patchlevel 1.

Version 2.1, patchlevel 0 -- Fri Nov 27 02:00:00 PST 1992 -- mann@src.dec.com

* Uses "time" command of gnuchess 4.0 and following to keep clocks in
sync.  Still works with older versions without this command; we test
whether the command is present the first time each chess program is
started.

* File name dialog pops up under the mouse cursor, so that it's got the
keyboard focus if the window manager is using pointer focus.

* Attempts to move the wrong color piece or an empty square are filtered
out in xboard instead of being passed on to GNU Chess or the ICS.

* The reason a game ended is now remembered even if you move backward
and forward after it ends.  It is forgotten only if you make a move
(which is as it should be---this is now a different game, which hasn't
ended yet).  The message saying why the game ended no longer wipes out
the last move.

* Pause, Backward, Forward, ForceMoves, and EditPosition now work from
TwoMachines mode.

* Forward and Backward now change only the board display unless you
are in ForceMoves mode.  Pause mode keeps new moves that are received
from being displayed on your screen until you unpause (or use
Forward).

* Added option to ring the bell after opponents' moves.

* Saved games that start from other than the standard initial position
now begin with a postion diagram as in saved position files.  The
loader is modified to understand such save files.

* Improved error checking when trying to read from gnuchess.

* Parser understands things that look like time indications, e.g., (0:12),
instead of popping them up as comments.

* Send "quit" to gnuchess before trying to kill it.  I needed this
locally because sending a SIGTERM to rsh was not killing the remote
program on some internal field test systems we have.

* Merged in code from John Chanak to make xboard a front end to the
Internet Chess Server, and added several improvements of my own.
(These include parsing end of game messages, loading the current state
and previous history of a game that's joined in progress via "watch"
or "load", automatic switch from board style 1 to 8 when needed,
removing irrelevant buttons and adding some new ones, adding the
useTelnet and gateway resources, and miscellaneous code cleanup, bug
fixes, and documentation.)  Many thanks to John for writing and
contributing the initial version of this code.

Fri Oct 30 20:16:40 PST 1992 (patchlevel 25) mann@src.dec.com

* gnuchess now castles by sending us "o-o" or "o-o-o" (starting with a
preliminary version of 4.0.pl60 that I have).  Added code to handle
this.  The old format ("e1g1", etc.) still works too.

* Added code to format moves in normal abbreviated algebraic notation
(for example, e4, exd4, f8Q, Nf6, 0-0) instead of coordinate algebraic
(for example, e2e4, e3d4, f7f8q, g1f6, e1g1).  Moves entered with the
mouse or received from gnuchess are translated into this canonical
format for display on the message line or in saved game files.  Moves
read from game files are not translated, however; they are shown and
saved just as they appear in the file.

Mon Sep 14 13:19:01 PDT 1992 (patchlevel 24) mann@src.dec.com

* It's not really correct to write an ep capture of a pawn on e5 as
fxe5; this should be written as fxe6 because e6 is where the capturing
pawn ends up.  Nevertheless, the parser now interprets fxe5 as the ep
capture f5xe6 unless there is a pawn on f4, in which case it
interprets fxe5 as f4xe5.  By design, the parser does not flag fxe5 as
ambiguous if there are pawns on both f4 and f5; instead it prefers the
more legitimate f4xe5 interpretation.

* A move like fxe6 or fe6 is now interpreted as an e.p. capture of the
pawn on e5 if there is one there and e6 is empty.

* A move like fxe5 can no longer be interpreted as an e.p. capture if
there isn't a pawn on e5 to be captured.  Previously this could happen
erroneously if there were pawns on both f4 and f5, and the f4 pawn was
capturing a piece on e5.

* The parser no longer munges the move it is parsing; in the past, for
instance, it would remove the "x" from a move like dxc5 before echoing
it to the screen.

* The parser now detects when a move is ambiguous and returns an
error, instead of arbitrarily choosing one possibility as it used to.

Mon Aug 10 18:40:47 PDT 1992 (patchlevel 23) mann@src.dec.com

* Added code so parser can handle fully qualified algebraic, e.g.,
Ng1-f3, N/g1-f3, Ng1f3, or even Pe2-e4.  The "P" and "/" work only for
fully qualified moves, not generally.

* Fixed parser bugs in handling moves with rank or file disambiguator,
e.g. N1f3 or Ngf3.

Mon Jul  6 17:55:32 PDT 1992 (patchlevel 22) mann@src.dec.com

* AIXV3 patch from Tom McConnell; thanks!

* After hitting the Backward button we see the last move made instead of
the word "Pausing" displayed.

* Fixed a couple of problems compiling on IRIX; thanks to Michel Arsenault
and Alan Walsh for bug reports and help with fixes.

* Added optional display of algebraic notation coordinates along left and
bottom edges of board.  This was inspired by some code from Jean-Christophe
Engel; thanks!

* Swapped EditPosition and Pause buttons to reduce danger of losing the
game in progress by hitting EditPosition when you wanted Backward.  Thanks
to Ove Lundberg for complaining (though it bugged me too).

* Despite one complaint, I kept the feature of changing both clocks when black
makes time control, rather than changing each individually when the player
involved makes it.  This is to avoid having it look like White is way ahead on
time while Black is thinking about the last move of the time control period
(i.e., when White has made time control but Black hasn't yet).

* Corrected clocks to work as in real tournaments.  Time is *added* when you
reach time control; previously the clocks were simply set to the length of the
new time period.  Also, when a player's flag falls, he still does not receive a
new time allotment at the next time control, but his opponent does still
receive more time if *his* flag has not fallen; previously, time controls were
ignored for both players if either flag fell.  I forget who reported this bug.

* Added a small ESIX patch from Kayvan Sylvan.

Mon Jun 22 13:24:38 PDT 1992 (patchlevel 21)

* Added some SVR4 support code from Ronald Cole.  It's Greek to me.

* Note:  Stephen Meatheringham reports a bug causing xboard to crash with an
error from the X server when run on a Sun Sparcstation IPX with Solaris 1.0.1
and OpenWindows 3.0.  This bug has not been tracked down yet.

* Fixed a problem with ATTENTION code.  On some operating systems, you have to
be even more cautious about when you send a SIGINT to gnuchess, because the
signal handler gets deinstalled each time it is used, and gnuchess does not
reinstall it immediately.  In particular, if you send a SIGINT and make an
illegal move or ask for a hint, gnuchess does not reinstall the handler until
*after* you make your next move, so you had better not send a SIGINT before
it.  I put a workaround for this case into xboard; I hope there are no more.

* The ATTENTION code is needed when compiling for the ESIX operating system,
because ESIX has a bug in the FIONREAD ioctl, so the code in gnuchess that
lets you interrupt its thinking on your time just by typing in your move does
not work.  Perhaps gnuchess will have a workaround for this eventually.

* Fixed yet another bug in using XtGetValues; code to inhibit shell resizing
was leaving a null pointer as the place to store the current sizes.

* Man page fixes: corrected -sl to -sd, updated LIMITATIONS.

* Added include of <sys/time.h> conditional on HAS_GETTIMEOFDAY.  I hope this
works for all systems; on mine, <time.h> and <sys/time.h> are the same file,
but with an #ifndef that makes it harmless to include both.

* Thanks for reports from Mike Pearlman, Stephen Meatheringham, Kayvan Sylvan,
d87-mal@nada.kth.se, Urban Koistinen, and Ronald Cole.

Mon Jun  8 11:39:18 PDT 1992 (patchlevel 20)

* Documented default value of initString and explained what it means.  This
paragraph also serves to inform the user that xboard by default will tell
gnuchess to think on his time, something one user apparently didn't understand.

* Redid -searchDepth (formerly -searchLevel) and -searchTime code.  They both
work correctly with gnuchess 3.1 now.  The -searchDepth option also works
correctly with gnuchess 4.0.  The -searchTime option does not work with
gnuchess 4.0.pl50 or earlier because those gnuchess versions do not have a
working searchTime feature.  Instead of having xboard do a half-baked emulation
of this feature by using a 1-move time control, I will be submitting a patch to
the gnuchess folks to restore the feature as in 3.1.

* Improved clock code.  It now keeps accurate track of fractional seconds of
clock running time, if your system has gettimeofday(); if not, it is generous
about them---if a clock has been going for n.m seconds when it is stopped, only
n seconds are charged.  Previously, xboard would charge one second immediately
when a clock was started, so if n.m had elapsed when it was stopped, n+1 would
be charged.  This should stop xboard from thinking gnuchess has let its flag
fall on the last move of a time control, as it often did before.  I also
improved the comments and condensed the code a little.  I hope there are no
systems that have X11 and don't have gettimeofday(), but if there are, I'd like
to know what to replace it with.

* "Flag dropped" messages now show up on the top line, beside the clocks, so
that they don't obscure more interesting messages on the second line.

* Fixed confusion of NULL and '\000'.

* Include time.h to get ctime() declaration.

* Fix to "parser bug" messages in parser.l.

* Changed code to work with old-fashioned tolower() macro that requires its
argument to be an uppercase letter.

* Minor fix to some usage error messages; was reversing program name and
erroneous argument value.

* Thanks to Mitch Wright, Ranier Orth, and Henrik Schmiediche for bug reports.

Fri Jun  5 13:06:08 PDT 1992 (patchlevel 19)

* Note: xboard does not seem to work when compiled with gcc 1.40 on a
PS/2 running AIX.  It crashes at the sscanf in line 768 of xboard.c.
You can work around the problem by specifying the -timeControl option
on the command line instead of defaulting it.  This does not seem to
be a bug in xboard itself; it runs okay on PS/2 AIX when compiled with
MetaWare High C.  [But see 2.1.pl8 above.]

* Removed bogus files from distribution.

* Changed variable name that conflicted with SVR4 ttyname function.

* Thanks to Christoph Strozyk for another bug report, and to several
of the folks mentioned under patchlevel 18 for testing it.

Wed Jun  3 17:50:37 PDT 1992 (patchlevel 18) mann@src.dec.com

* Added an include of sys/fcntl.h to pick up the definition of O_RDWR for
System V users.

* Fixed a bug in CatchPipeSignal; was testing the wrong variable to determine
which chess program got the signal.

* Removed the declaration of strrchr, which was causing problems for some
people, and added an include of <string.h> or <strings.h> instead.

* Fixed a bunch of type errors uncovered by compilers that check more carefully
than the one I was using.  I just patched many of them with casts---typically
cases where a function didn't take the same number and type of arguments as the
X header file prototype said it should---because I didn't see any other way to
fix them.  The versions of the X header files that I have don't actually have
full prototypes for these function types; they declare only the return type and
just give the argument types as comments.

* InitChessProgram returns immediately if chess program died while we were
waiting for it to say "Chess".  I think this fixes a bug that could make xboard
crash or go into an infinite loop printing "Warning: Select failed; error code
9" if gnuchess crashed there.

* Removed code that tries to put the file descriptor used to read messages from
gnuchess into nonblocking mode.  I don't see how this kludge could ever have
fixed the problem it was intended to (see below), and it seemed to be causing
a race condition when we were waiting for the initial message from gnuchess
after it started up.

* Thanks to Richard K. Lloyd, Konstantinos Konstantinides, Mark
Seiden, Erik Schoenfelder, Andreas Stolcke, Rainer Orth, Henry Thomas,
and Kris Van Hees for bug reports and help with fixes.

Sun May 31 08:00:00 PDT 1992 (patchlevel 17) Bart Massey bart@cs.uoregon.edu

*  'xboard' suffered from a couple of serious bugs.  An initial
   size needed to be given to the message label to avoid a race
   condition (!) so I gave it some text (otherwise occasionally
   xboard would fail because the message widget was 0 width under
   MIT X11R4).  The sizes needed to be Dimension, not int, because
   they were written in by Xt -- on my big-endian box Dimension
   is short, causing chaos.  Also, I hacked a declaration in for
   strrchr() to shut the compiler up -- this is arguably a kludge,
   but I couldn't think of anything better; maybe you can.

Fri May 22 13:40:50 PDT 1992 (patchlevel 16)

* First real (non-beta) release of version 2.0.

* Added GNU copyleft notices covering version 2.0 enhancements.

Thu May 21 12:30:09 PDT 1992 (patchlevel 15)

* Removed -g from Imakefile.

* Added casts in parser.l to prevent warning messages from some C compilers.

* Bugfix: code to exit if chess program couldn't be started had bug in error
message printing.

Tue May 19 11:58:06 PDT 1992 (patchlevel 14)

* Fixes one small bug: If LoadGame does not find a game in the file, it resets
xboard to BeginningOfGame mode (as if the Reset button was pressed), instead of
leaving it in LoadGame mode.

Mon Apr 13 18:32:27 PDT 1992 (patchlevel 13)

* Improved error handling when gnuchessx crashes.  Now displays a message and
enters EndOfGame state without exiting.  This should help with diagnosing
gnuchessx problems, since the position and move history are not lost.

* Debugged man page.

Version 2.0 -- Thu Apr  2 16:04:54 PST 1992 (patchlevel 12)

* Added small 40x40 pixel pieces; board size now specified as small, medium, or
large. 

* Improved man page.

* Saved position files now say "white to play" if it's white's move.

* SavePosition while in EditPosition mode now gets side to play right.

* Made "Pawn" the default on edit position menus.

* Removed "To play" from edit position menus, since it's redundant.  Left
support code in so it can be restored if someone wants it.

* Tightened parser's definition of a move number to help further with skipping
over stuff before the first move.

* Put back feature of skipping over stuff in game files before a start of game
marker.  This helps when loading electronic mail messages that have things
like "00" in the header (looks like castling), or that talk about moves in
running text before the actual game score.  Tradeoff is that we can't load
games that don't either have move numbers or say "gnuchess game" just before
the moves.

* Changed version number to 2.0 because others have used "1.3" for patched
versions of 1.2.  Display version number and patchlevel at startup.

Thu Apr  2 15:30:22 PST 1992 (patchlevel 11)

* Bugfix:  Match started with TwoMachines button now handles the case where
user loaded/edited a position *and* forced some moves.

* added error message feedback to many cases where the user's button press is
rejected because it isn't allowed in the current mode.

* added -noChessProgram option to allow use as a passive chessboard without
gnuchess, replacing old buggy scheme that tried to do this automatically when
gnuchess was not found on the search path.

* parser accepts "=" in front of promotion piece

* parser accepts ":" notation for captures

* Removed SwitchSides button to make room for EditPosition.  Left code in
place, ifdef'ed out.

* Added EditPosition mode.

* Reformatted xboard.c to eliminate long lines.

* Promotion popup now appears near rank to which pawn is moving.

* UI improvement: Button representing current program mode is kept highlighted.

* Bug fix:  Hint button is now disabled when machine is on move.

Sat Mar 21 14:57:37 PST 1992 (patchlevel 10)

* TwoMachines now works from MachineWhite or MachineBlack mode.

* ForceMoves and Backward/Forward now work at the end of a game started by
the TwoMachines button.

* Sends o-o or o-o-o to gnuchess instead of O-O or O-O-O.

* Bug fix:  SwitchSides button is now disabled when machine is on move.

* Added -searchLevel option and corrected handling of -searchTime to reflect
the way gnuchess really handles command line options instead of what the
documentation used to say.  Allowed optional ":seconds" on times.

Tue Mar 17 21:00:35 PST 1992 (patchlevel 9)

* Removed non-working code that tried to recover when a remote chess program
could not be started; now exits with a message instead.

* Now waits for chess program to print its first message (normally "Chess")
before sending it any commands.

* Matches between two machines can now start from a black-to-play position or
black-to-play opening.

* Bug fix: The program would sometimes lose track of the board position after
the end of the game, drawing a board full of white pawns next time it got an X
redraw request.  Now fixed.

* Cleaned up handling of illegal moves in game files.  Now we just stop reading
and go into ForceMoves mode.

* Saved games against gnuchess indicate who won (restored and debugged 1.2
feature). 

* Setting -searchTime turns off -clockMode.

* My trick for setting up a black-to-play position broke in the transition from
gnuchess 3.1+ to 4.0.  Installed a different trick that works with both.

* Bug fix: Was clobbering memory by mallocing one byte too few for the file
name in SetupPositionFromFileProc.  This usually resulted in a crash on the
second attempt to load a position from a file.

Mon Mar 16 16:49:00 PST 1992 (patchlevel 8)

* Game and position files are saved in CHESSDIR as well as loaded from there.
We actually chdir there, so gnuchess listing files also wind up there.

* Underpromotion moves from gnuchess really work now.

Sat Mar 14 14:55:24 PST 1992 (patchlevel 7)

* Cleaned up lots of bugs and kruft in parser.l.  It's a wonder the old version
worked at all.  Improved rejection of random text interspersed with the moves.

* Relabeled buttons to reduce confusion (Play From File --> Load Game;
Setup From File --> Load Position).

* Bug fix:  Doesn't exit at the end of a game that was started by pressing
the Two Machines button.

* Bug fix:  Doesn't get confused if you hit Machine Black (White) while
the machine is already playing black (white) and is on move.

* Prints "White to play" or "Black to play" after loading a position file.

* Bug fix:  No longer hangs if no start-of-game is found in a game file.

* More permissive about game file format.  The first line is not assumed
to be a header comment unless it starts with "#".  We don't insist on
seeing a "1." before the first move.

* Added pop-up dialog to allow underpromotion.  Moving a pawn to the last rank
pops up a dialog to let you select what piece to promote to.  Also added code
to handle underpromotion moves from gnuchess or from a game file correctly.

* Added code to send SIGINT to gnuchess before trying to move, if gnuchess
might be busy thinking on our time.  The code is ifdef'ed out because it's
not necessary; only beta-test versions of gnuchess 4.0 at patchlevel 19 or
below require it.

Mon Feb 24 20:10:22 PST 1992

* Merged fixes received from Stuart Cracraft and Mike McGann, to iconic mode
and to ShutdownChessPrograms.  (They both sent the same fixes.)  I did not
install the change that removed "easy" from the initialization string, so
xboard will still try to run gnuchessx in "hard" mode (thinking on opponent's
time) by sending this command.

Version 1.3 -- Thu Dec 19 18:19:00 PST 1991  mann@src.dec.com (Tim Mann)

* I've fixed a slew of bugs in this version of xboard:

* You can now back up from the end of a game, even if gnuchess was playing
against you.  Since gnuchess exits upon mate or draw (when it's not in force
mode), this feature works by starting up a new copy of gnuchess and feeding it
the saved starting position and moves.  Drawback: the new gnuchess starts with
clocks at the beginning of a fresh time control.

* Note: the "force" command to gnuchess is a toggle.  I fixed a couple of
places in xboard where "force" was being issued when gnuchess was already in
force mode, which resulted in chaos.  I don't know if they were in 1.2 or if I
added them myself in the course of putting in new features.

* ForceMoves button is now legal in PlayFromFile mode.  It closes the game file
and lets you force moves (or go Backward) from whatever position you've
reached.

* The parser now recognizes "white resigns" and "black resigns" (in any
combination of upper and lower case, with or without the closing "s").  These
used to be misinterpreted as meaning "white wins" or "black wins" respectively!
Added "1 - 0" and "0 - 1" as equivalents for "1-0" and "0-1".

* I removed the parser feature of considering "!" as a comment delimiter.
"!" is so common as an annotation in game scores that this feature seems like a
very bad idea.

* Added a wait() after each successful call to kill() to avoid leaving zombie
processes around.

* Removed the feature of putting an indication of how the game ended into the
internal move list when playing from a file.  The only benefit of this was that
it would get written back out if you saved the game at that point.  This
feature doesn't seem too useful, since you are unlikely to save a game you've
read in without making any changes, and it caused a lot of problems in the code
and user interface.  The end condition is still displayed as a message when
it's first read in from the file.  End conditions when playing against gnuchess
are still saved, too.

* programName is now the last file name component of argv[0], not all of it.

* xboard.c got three compiler warnings due to arguments that didn't match the
prototypes on X toolkit functions.  I got rid of them by adding some casts, and
adding extra unused arguments to the ReceiveFromProgram function.

* As I received it, the distribution had an extra period glued on the front
of every line that started with a period.  This broke the manpage in an
obvious way.  It also broke parser.l in a nonobvious way that stopped
PlayFromFile from working.

* Fixed one or two bugs where the clock was started when it shouldn't have
been.  I think one example was if you tried an illegal move in ForceMoves mode.

* Now starts the machine's clock if you hit MachineWhite when white is on move
or MachineBlack when black is on move.

* Now updates the to-move indicator (highlight on clock display) when you use
Forward/Backward or ForceMoves.

* gnuchess 3.1+ has a bug that makes it try to think on its opponent's time
even when in force mode; this causes problems with xboard's SetupFromFile
command because gnuchess's saved hint move is garbage after the new position is
loaded, so gnuchess prints out an error message when it tries to think about
what it would do if the user made this move.  Here is a patch for gnuchess
3.1+:

*** nondsp.c~   Mon Dec  2 12:37:18 1991
*-- nondsp.c    Thu Dec 19 15:36:06 1991
***************
*** 1223,1229 ****
    ok = flag.quit = false;
    player = opponent;
    ft = 0;
!   if (hint > 0 && !flag.easy && Book == NULL)
      {
        fflush (stdout);
        time0 = time ((long *) 0);
*-- 1223,1229 ----
    ok = flag.quit = false;
    player = opponent;
    ft = 0;
!   if (hint > 0 && !flag.easy && Book == NULL && !flag.force)
      {
        fflush (stdout);
        time0 = time ((long *) 0);

* SavePosition/SetupFromFile now handle black-to-play positions.  If it's
black's move in a saved position, the line "black to play" is printed
immediately after the board display.  Upon loading a position, xboard
looks for "black" at the beginning of this line to determine whether it's
black's move.

* SetupFromFile now leaves you (effectively) in ForceMoves mode.  This
eliminates some cases where the program would get into an inconsistent state,
and hopefully is more intuitive for the user, too.

* PlayFromFile now leaves xboard in ForceMoves mode after playing through all
the moves in the file.  You can use Backward/Forward, make moves, get gnuchess
to join the game with MachineWhite or MachineBlack, etc.  The game file can end
with either white or black to play.

* Backward/Forward now seem to work in all cases.  Previously using these
buttons during a game against the machine would get xboard out of sync with
gnuchess's idea of what was going on, resulting in chaos.  Now using these
buttons puts xboard in ForceMoves mode; you can bring gnuchess back into the
game using the MachineWhite or MachineBlack buttons.

Thu Jun 27 15:06:25 PDT 1991

  * Remove comment popups during a reset.

  * Bug in my forward move fix.  Didn't work for playgamefile mode.

Thu Jun 20 15:04:06 PDT 1991

  * Can't do a backward move when it isn't your turn because gnuchess
    gets confused.  Similar bug for forward move.  Do a bunch of moves,
    back up move and go forward.  These get filtered out now.

  * Put easy back into the initString.  This DISABLES easy mode which by
    default is ON.

  * New-line wasn't always being added to forward move strings sent to gnuchess.

  * Parsed comments as [ ..............text follows ] into d7d5

  * Long game file comments cause core dumps on HP systems.

  * Multiple uses of the get positition file button got confused
    on the second click.

  Minor fixes to the 1.2 release

Version 1.2 -- Tue Jun 11 17:14:12 PDT 1991

  * For version 1.2, Jeff Kenton, Richard LLoyd, David Jensen, Martin D.,
    Bill Schmidt, Scott Hemphill, Paul Vaughan and Bill Shauck all found
    a lot of bugs that we put into xboard just to see if they were paying
    attention.  They were.

  * Changed the protocol that xboard uses to work with the new version of
    gnuchessr.

  * There was an off by one error with the clock resetting code.  Also,
    the clock display highliting was wrong.  Thanks to Bill Shauck for
    pointing these out.

  * enumerations are not really integers on some C compilers.  I added casting.
    The new version of gnuchess expects the "go" command for two machine
    and machine plays white cases.  The whiteString resource is for
    compatibility with previous gnuchess versions which get confused by go.
    Thanks to Martin D. for catching these.

  * Fixed a bug where the checkmate message wasn't being parsed and xboard
    kept playing.  Also, the message was added to game file if saved.
    Thanks to Scott Hemhill for pointing this out.

  * Fixed a bug where saving a long game, resetting and saving a short game
    resulted in appending the end of the long game to the short one.
    If a game is just being played out and there is no reason for gnuchessr
    to be used, ignore pipe signals.  This allows people to use xboard
    as a chess board without need for gnuchess.  Also, trivially bad moves
    such as e2e2 are not passed on to gnuchessr but instead ignored out of
    hand.  This allows people using xboard as a chessboard with gnuchessr
    to pick a piece up, think and put it back down without making a move.
    Thanks to Jeff Kenton for pointing these out.

  * Added the CHESSDIR environment variable.  Game and position files are
    found in this directory.  If this variable is not declared, then the
    current directory is used.  File names starting with / are treated
    specially.

  * The bitmap file names were changed so that none exceeded 14 characters.
    This is necessary for R5.

        fcntl(from_prog[0], F_SETFL, O_NDELAY);

    The xboard fix was to set non-blocking i/o on the read pipe for gnuchessr

        printz ("Illegal move (no matching move generated)");

  * fixed a problem with the new version of gnuchessr where xboard wasn't
    getting gnuchessr error messages for illegal moves.  The problem seems
    to be fixed *without* any gnuchess changes but this is likely to be
    highly system dependant.  There should be a new line on line 246 of nondsp.c

  * -iconic doesn't work at all now.  The Iconify() does work.  This was
    a tradeoff and it is really an Xt bug.

  * Fixed a compiler warning for gcc and an error for the IBM RT compilers.
    This is the VOID_PROC macro in xboard.h.  Thanks to David Jensen for this.

  * If the DisplayWidth or DisplayHeight is less than 800, use small size mode.
    Also the size of the name widget was reduced from 500 pixels to 400 pixels
    because in small size mode there was a gap on the right.
    Changed the default font from fixed to helvetica_oblique14 to:
        **-helvetica-bold-r-normal--14-*-*-*-*-*-*-*
    helvetica_oblique14 is a font alias not on all R4 systems.
    Curiously enough, the 17 point is not available on 75dpi systems
    and the 18 point font is not available on 100dpi systems.
    Thanks to Richard K. Lloyd for pointing these out.

  * Minor hacks to work with R5 alpha.  Had to add an event handler to the
    boardWidget to get translations to work.  This may go away with the real R5.
    Added <Message>WM_PROTOCOLS: QuitProc() for R5 ICCCM compatibility.

  * If the first character of the file is not `1' then the first line
    of a game or position file is displayed as the name in a label widget.

  * Added a PopUp dialog for getting file names.

  * Modified the code to use the R4 routines rather than R3 compatibility.

  * Changed the forwards/backwards/readgamefile code to allow a play to step
    back and forth in a game.  If he steps all the way to the beginning
    he has to restart the game.  If he gets into a mated position, same problem.

  * Moved the man page from xboard.1 to xboard.man.  This makes imake man page
    installs work correctly.  Thanks to Richard K. Lloyd for pointing this out.

  * Corrected the queening code.  XBoard was sending the wrong syntax.
    It was sending for example h8(Q) when it should have been sending h8q.
    Thanks to Paul Vaughan and Mike McGann for pointing this out.

  * Added a lex parser for algebraic games.  It should be easy to use the parser
    for other interfaces such as the Microsoft Windows version.
    It parses comments of the form [anything] and ! to the end of a line.

Version 1.1 -- Mon Jan  7 14:46:03 PST 1991

  * Roger Dubar, Konstantinos Konstantinides, Wolfgang S. Rupprecht,
    Paul Scowen, Mvh Smidt and Kayvan Sylvan all helped immensely during
    beta-testing.

  * older non-ANSI versions of Sun compilers complain vociferously.

  * the remoteShell resource was added for HP-UX systems
    and other systems where the remoteShell isn't rsh.

  * -iconic doesn't work properly.  If XBoard is opened iconic then
    iconifying it later with a keystroke doesn't work.  I think
    this is an Xt bug.

  * For systems with smaller screens, XBoard can use smaller pieces,
    and a smaller board.  -bigSizeMode False uses a smaller set of pieces.
    These are scaled versions of the large pieces.  They look ok but could
    be improved.

  * getlogin() does not work in all cases.
    It was replaced by getpwuid(getuid())->pw_name).

  * Warning messages get displayed in the message box.

  * Any button restarts a paused game.

  * Kayvan Sylvan contributed a patch for ESIX.
    It seems that select() on pipes is broken on his system.  We declined
    to incorporate his patch though, because it was a work-around for
    something that was broken on one system, and selfishly, that system
    was not my own.  Nevertheless, it is likely that other System V users
    on PC's may need to use this patch and it is is included as the file
    ESIX.patch.  To use it, type

        patch xboard.c ESIX.patch

  * Fixed a bug where hint didn't work.  The local version of gnuchessr
    had been hacked.  The fix was to clone stderr and stdout for gnuchessr.

  * Fixed a bug where a player could play after a game had been finished.

  * The bitmaps have been pushed into the bitmaps directory.

  * Substantially rewrote and simplified the Two Machine code.

  * Added StrStr() because many systems don't have this ANSI function.

  * Added keyboard equivalents.  Added an iconify keystroke, C or c.

  * Works with gcc now.  gcc complained about casting float constants
    as XtPointers.

  * Detect if the visual doesn't support color.  If so, run in monoMode.
    An alternative would be to detect grayscale visual and use a collection
    of gray user interface colors.

        XBoard.clockMode:    False

  * For people who don't like or use chess clocks a clockMode switch
    has been added.  It is on by default.  It can be turned off in the
    .Xdefaults file with

  * Added a declaration, xtVersion, which will quickly break on R3 Intrinsics.

  * Hal Peterson contributed a new Imakefile.

  * Rewrote DisplayClocks().

  * Rewrote TwoMachinesPlay mode.

  * Konstantinos Konstantinides added the -searchTime option.

  * Substantially rewrote and simplified the clock code.  If a game was paused
    and then resumed, the clocks were out of sync.

  * Fixed a bug in HandleUserMove() where the user could make a move while the
    machine was thinking.  The fix detects and ignores these moves.  onMove
    was not being used and was removed.