Fix illegal drops

[xboard.git] / ChangeLog

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

(latest entries created by git log --no-merges --pretty="%ai %an <%ae>:%n%s%n%n%b" vlast.. )

2014-10-17 21:17:56 -0700 Arun Persaud <arun@nubati.net>:
updated NEWS


2014-10-17 21:17:28 -0700 Arun Persaud <arun@nubati.net>:
forget a few __APPLE__ ifdefs; changed to OSXAPP


2014-10-17 00:01:44 -0700 Arun Persaud <arun@nubati.net>:
fix typo and prefix

AC_PREFIX_DEFAULT was always set, even if AS_IF didn't get called? Some kind of caching?
Using just prefix=... seems to work though

2014-10-16 23:34:22 -0700 Arun Persaud <arun@nubati.net>:
USE OSXAPP instead of APPLE and fix withval->enableval in AC_ARG_ENABLE


2014-10-16 21:44:33 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix loading of engine-defined PGN games

Somehow there could be disagreement over what the official opening
position of an engine-defined variant was during loading of the game.
It then refused Betza-defined castling, which tests the corner pieces
based on this initial position. We now assume the FEN tag, which such
a PGN game will always contain, holds the official opening position,
so that castling will always be assumed possible (if there is a corner
piece).

2014-10-16 20:29:27 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Preserve parent variant for PGN of engine-defined game

Rather than decoding an unknown variant name, (which will result in
'normal'), we keep the currently set (parent) variant when an
engine-defined variant is currently set that matches the name in
the PGN variant tag of the loaded game.

2014-10-16 17:54:13 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix some uninitialized variable bugs

The writing of Seirawan castling rights in FEN was still dependent on a
now unused variable, and encountering a VariantMen tag in a PGN file
could have created the misconception the memory was full.

2014-10-16 17:41:01 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Update Game List after tag selection changed

When the Game List Options dialog changes the tags to be displayed in
the Game List lines, we now automatically redo the entire Game List.

2014-10-16 16:38:00 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Always accept piece commands in variant great

Great Shatranj belongs to the variants XBoard does not know the rules of,
and should thus always accept engine piece commands.

2014-10-15 21:28:36 -0700 Arun Persaud <arun@nubati.net>:
make all tests for strings in configure use the same scheme


2014-10-15 21:23:26 -0700 Arun Persaud <arun@nubati.net>:
updated Changelog, DIFFSTAT, and SHORTLOG

2014-10-15 21:06:36 -0700 Arun Persaud <arun@nubati.net>:
fix osxapp enable option in configure.ac


2014-10-15 21:06:00 -0700 Arun Persaud <arun@nubati.net>:
remove experimental from gtk build option


2014-10-15 21:02:38 -0700 Arun Persaud <arun@nubati.net>:
only enable osxapp build target on apple systems, clean up configure.ac a tiny bit


2014-10-15 20:57:28 -0700 Arun Persaud <arun@nubati.net>:
update NEWS file


2014-10-15 23:48:14 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Don't switch to engine-defined variant on game loading

When loading a game from PGN the variant tag will have been decoded as
'normal' in case of an engine-defined variant, and we certainly would
not want to switch to that. Better stay in the variant the user had
selected before, and hope for the best.

2014-10-15 21:50:44 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement displaying of variant tag in Game List

The variant tag was displayed as an empty string in game-header lines.
Processing it during PGN load was not able to handle engine-defined
variants anyway; they were recognized as 'normal'. A new field in the
GameInfo struct now holds variantName in text form, and this is the
primary place from which it is displayed in the Game List.

2014-10-15 21:06:53 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Update Game List when setting new Game List Tags

Using the Game List Tags dialog to alter the gae header lines now
automatically causes an update of the Game list according to the new
tags specification.

2014-10-15 20:26:31 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Reset move entry on Clear Board

The board markers would stay on when the board was cleared, and a
no-longer-present piece would stay selected, leading to deletion of the
first piece that you tried to select.

2014-10-14 09:38:05 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Split Tournament dialog in side-by-side panes

This seems to cure a sickness in some Xaw versions, which refused
to display text in the text widgets, or make them sensitive for mouse
clicks if the last three buttons were added. It also helps keeping
the 'OK' button on-screen in the GTK version.

2014-10-12 19:43:24 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Remember position obtained from setup

When the engine sends a setup command, it should be remembered as
initialPosition, in order for the castling 'rook' test of the Betza
move generator to work.

2014-09-13 12:35:12 -0700 Joshua Pettus <jshpettus@gmail.com>:
Add build script to configure for a XBoard.app for OS X

This relies on the the gtk build and adds some special code in
gtk/xboard.c so that XBoard can be installed inside the app directory
independ on where the app file is located.

2014-10-12 10:09:51 -0700 Arun Persaud <arun@nubati.net>:
replaced hardcoded pngdir with built-in ~~

don't need to create xboard.conf from xboard.conf.in anymore

2014-10-12 00:18:50 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Some improvement on new Shogi SVG pieces


2014-10-11 10:55:41 -0700 Arun Persaud <arun@nubati.net>:
updated spanish, ukranian, and dutch translation


2014-10-11 19:42:03 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix logos Xaw

The patch for logo standard left a typo and an excess line in the Xaw
version. This is fixed, and also implements the user-logo cascade for
the Xaw version.

2014-10-11 19:26:23 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Try also /home/<user>/.logo.pgn for user logo

The user logo now also has a cascade, first looking in the user's home
folder for a file .logo.png, and then in the -logoDir for <user>.png.

2014-10-11 16:38:03 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix ICS logo display


2014-10-10 18:14:09 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Show current variant on New Variant buttons GTK in bold

If an engine-defined variant is selected, we compare its name to the
inscription on the button, otherwise we compare the variant numbers.
(Test to trigger this is a bit awkward, to not inadvertantly apply it
to the ICS Text Menu, which also has colored buttoms.)

2014-10-10 17:01:51 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Increase number of engine-defined-variants Buttons XB

An third column is added to the New Variant dialog when necessary.

2014-10-09 23:25:26 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Force GTK logo size to quarter board width

This is what GTK will do anyway, because clocks and logos are in a hbox
that equally divides space, so telling XBoard different just misinforms it.

2014-10-09 17:08:45 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Replace default Shogi pieces

The SVG Shogi pieces are replaced by a new set, made by Joshua Pettus.
This set also contains and Elephant and Crown Prince.

2014-10-09 09:35:33 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement new logo standard

XBoard now looks in the system directories for plugin logos,
/usr/(local/)share/games/plugins/logos, according to the cascade
-logoDir, engine dir (-fd,-sd), /usr/local/..., /usr/...

2014-10-09 08:47:25 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Relocate OS X' LOCALEDIR


2014-10-08 21:56:28 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix check test with multi-leg moves

The check test after a multi-leg move was still using the same kill
square in the opponent's reply, making it blind to King captures in
a second leg. So far we got away with that because the Shogi multi-leg
pieces also can reach all their potential victims in a single step,
but with engine-defined multi-leg moves there is no guarantee this
will always be the case.

2014-10-08 21:55:17 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix infinite loop in cylinder moves


2014-10-07 19:13:54 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Let t on final leg in Betza notation forbid checking

To allow approximate definition of pieces that are too difficult to
describe exactly, we can revoke their capture rights on Kings,
to prevent imagined moves that happen to threaten a King forces the
opponent to evade this imagined check, making most of his moves illegal.
Then the piece can be safely replaced by an upward-compatible one,
leaving rejection of the extra moves to the engine.

2014-10-07 18:56:59 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement Betza hr and hr as chiral move sets

To easily separate right- and left-handed moves on oblique atoms
(which needs opposite use of l and r in their multi-leg continuations),
the direction sets hr and hl are added.

2014-10-07 08:20:06 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Preserve other Betza mode bits on setting default modality


2014-10-06 20:20:00 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Switch to new Betza orth-diag conversion standard

The interconversion from orthogonal to diagonal atoms is now possible
for any 4-way atom, based on specification of an 'in-between' direction,
irrespective of modality. Also take backward out of default continuation
direction-set.

2014-10-03 10:31:47 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement Betza t modifier for hop-own

The 't' modifier will restrict any form of hopping as specified by
p or g modifiers to friendly pieces. This only works on non-final legs.

2014-10-02 19:32:45 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement directional modifiers on KQ, and let y&g upgrade

The Betza y and g modifiers now not only change range, but also upgrade
the directional capabilities of the atom to full non-degenerate 8-fold.
Directional modifiers can now be used in that symmetry too, to maximally
exploit this, and implement 4-fold pieces that make 45-degree turns.
Let oblique continuation legs also use the non-degenerate system.

2014-10-02 15:19:09 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement Betza y modifier

The non-standard modifier 'y' is the spontaneous counterpart of 'g',
changing the range of the atom, and upgrading its directionality.

2014-10-02 10:21:43 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement Betza g modifier for non-final legs

The g modifier is implemented as a p that turns a rider into the
corresponding leaper when it hops, and vice versa. This is done by
stripping off any existing range indicator, and making the range
infinite by suffixing '0' when the old range was 1 (whether implied
or explicitly written). Except that for sliders/steppers this is done
by atom-name conversion after stripping off the range. Limited range
thus in general turns into leaper (D -> D0, Dn -> D), but with FWK
W3 -> R and R3 -> W, etc.

2014-10-01 15:32:14 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement Betza a modifier

This is pretty complex, and might overload the current primitive two-leg
legality testing when pieces are defined that can reach squares only in
two legs.

2014-10-01 09:28:32 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement Betza j on W,F as skip first square


2014-10-04 21:56:14 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Always accept piece commands for Falcon and Cobra

The wild-card pieces Falcon and Cobra do not have any moves assigned to
them, and XBoard would accept any move for them even when legality
checking is on. So for these pieces it is always better to trust the
engine's piece definition.

2014-10-04 21:52:51 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix pieceToCharTable of Falcon Chess

This variant was using the Falcon piece, but defining ID F for the
unused Lance instead.

2014-10-04 19:59:13 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Don't preserve setup position on board-size change

Normally the setup command sent by an engine is ignored if the user
already has set up a position, to not render it impossible to do the
latter. But if the setup command orders a different board size nothing
the user has set up can possibly remain valid.

2014-10-04 19:39:06 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Reset move entry on stepping through game

A forward or backward step through the game after selecting a from-square
did not erase the latter, so that it became impossible to select a
from-square for an opponent move, as it would be interpreted as the
to-square for the patially entered move. This was especially annoying
with legality testing off, as the legality test there, based on the
square markings, does not reset the entry state either, but waits
until you do click a legal to-square, so that you had to attempt
moving with pieces of the side not to move to get out of this state.
Now stepping through another position will reset the move-entry state
completely, always considering the first click a from-square selection.

2014-10-07 20:55:00 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix engine-defined variant as startup

Non-standard names were all recognized as 'normal' in InitBackend1().
Now if the -variant argument was not "normal", it is assumed to be an
engine-defined variant if we run with engine.

2014-10-07 18:53:50 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix clipping of board GTK

ResizeBoardWindow now measures the with of the window dressing by
brute force, interrogating GTK both for the width of the top-level
window and the board widget, and subtracting the two.

2014-10-08 18:28:09 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix writing of Seirawan960 virginity in FEN

When variant seirawan was played with Fischer castling, only the
castling rights were written, and not the virginity of other pieces.
Now all virgin files are dumped in Shredder-FEN style.

2014-09-29 21:07:17 -0700 Arun Persaud <arun@nubati.net>:
update po/pot files


2014-09-29 21:00:42 -0700 Arun Persaud <arun@nubati.net>:
new version number for release 4.8.0


2014-09-13 14:33:37 -0700 Arun Persaud <arun@nubati.net>:
fixed whitespace error in configure.ac for os x


2014-09-20 10:06:17 -0700 Arun Persaud <arun@nubati.net>:
update gettext configuration to not include any generated files in git

before we had a somewhat mixed setup, which seems to create problems
when a different gettext version was installed on the users machine compared
to the version used to generate the files in git. This should solve
the problem. The user will now have to generate these files using autopoint,
which we call from autogen.sh.

2014-09-28 22:19:00 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Configure XBoard for -size 49 in master settings

Now the code for auto-sizing works, it turns out this does more harm than
good in the GTK build, as it takes maximum size that fits the screen as
initial size, and the board window then cannot be shrunk below that.
So we configure a hard -size 49 for a cold startup.

2014-09-28 21:27:22 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement ff etc. in Betza parser

Dupliacated directional modifiers can be needed with oblique moves
to unambiguously specify a direction set, e.g. ffrrN.

2014-09-28 13:40:13 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Rename PGN Pieces tag to VariantMen

Following a suggestion from Steven Edwards, this indeed makes a much
clearer name. Also add code to skip the tag on reading, rather than
allowing it to end up with the 'extra tags', as the latter could lead
to duplication of the tag on subsequent saves.

2014-09-28 11:18:10 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Suppress use of promo-Gold bitmaps in Tori Shogi (WB)

In Tori Shogi the Swallow Pawn does not promote to Tokin, so the use
of a Gold General pictogram would only be confusing. So we stick to
the regular fairy bitmaps in Tori, which is recognized by the board
having 7 ranks in Shogi.

2014-09-28 10:55:10 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Prevent board-size oscillations

It could happen that after resize there was room for enlarging the
squares by one pixel, but that this crossed the boundary where the line
gaps also become wider, and there is no room for making both squares
and line gap. This would cause perpetual oscillations of the window size.
Resizing is now suppressed when a 1-pixel square-size increase would also
result in a line-gap increase.

2014-09-28 10:09:17 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Allow use of second-row pieces for non-promoted in drop games

Second-row pieces were automatically demoted on capture, but this
should really happen only when the piece-to-char table says it is a
shogi-promoted piece (assigned a '+' there). Pieces assigned a letter
should really stay themselves.
 PieceToNumber had to be adapted to skip '+' pieces in order to count
correctly for second-row pieces, as did the code in UserMoveEvent() that
mapped the from-square onto the piece type. (Why do we do that anyway?
The piece type was displayed on the from-square!)

2014-09-26 20:24:05 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix reset of 50-move counter on FRC castling

The FxR internal representation was taken a bit too seriously!

2014-09-26 20:15:33 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix abort of machine game on variant mismatch

When the second engine did not support the current variant, and we clicked
Two Machines, there was an error popup, but the Two Machines menu item
stayed checkmarked. Now we explicitly uncheck it, and switch to EditGame
mode.

2014-09-26 19:47:08 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Make unsupported variant on loading 1st engine non-fatal

Like in the startup code, when the loaded engine does not play the
currently selected variant, we now figure out what the engine does play,
and switch to that. The error message still pops up, but not non-fatal.

2014-09-25 23:37:24 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Print PGN Piece tag listing engine-defined pieces

A newly defined PGN tag is printed, which summarizes all non-standard
pieces. Black copies of identically named white pieces are suppressed.
In the future also non-standard naming should be reported here.

2014-09-25 19:11:21 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Always accept piece commands in partly supported variants

Normally piece commands are ignored when legality checking is on, as
XBoard's idea of legality should obviously overrule the engine's.
In variants that are not or only partly supported, however, XBoard has
no own idea of what is legal, so it is better to take the engine's
advice than enforcing wrong rules.

2014-09-25 12:29:15 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Size seek graph to also cover board rim WinBoard

The seek graph always used the plain board size, but in WB it is
possible to diplay a rim around the board, and the seekgraph would
then be positioned in the upper-left corner of the rim, leaving the
lower and right board edge and rim uncovered. Its size is now increased
to fill the entire window (as leaving a rim around it seems not useful).

2014-09-25 11:48:00 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Reparse ambiguous move under built-in rules

When a move parses as ambiguous under engine-defined piece locomotion,
we re-parse it with XBoard's native idea of the pieces, just in case
it is a move from a game that was saved without the engine having
redefined the pieces. This keeps games created before the introduction
of the 'piece' command loadable.

2014-09-25 11:14:06 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Prevent multi-path moves from parsing as ambiguous

In Cylinder Chess the MovesFromString() generator might generate duplicate
moves, one wrapping, the other direct. When they were both counted as
matching the SAN input move, Disambiguate() would think it an ambiguous
move. So a move that came from the same square as the one we already
have must not be counted.

2014-09-24 21:47:59 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix cross-edge e.p. capture in Cylinder Chess

ApplyMove() did not remove the e.p. victim when an edge-crossing e.p.
capture was made, because e.p. rights are only set to neighboring files
(after checking there is a Pawn there to capture). The e.p. heuristic
has now been changed to also assume e.p. on file-changing Pawn moves
that span a large number of files, assuming that these will be wrapping
moves. Eventually we should change to better e.p. logic that can be
integrated with engine-configured move generation, as the current system
would fail for any Pawn that captures more than just neighbor ranks.

2014-09-24 19:43:27 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement Betza o modifier for cylinder boards


2014-09-24 17:02:37 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Improve virginity test for engine-defined pieces

Two squares in the board are now reserved for flags that indicate
whether back-rank pieces have been touched. This allows MovesFromString()
to accurately test virginity of these pieces, rather than having to assume
it when the piece matches that in the opening position. For other ranks
the latter test is still used, as these are normally (irreversible) Pawns,
which cannot return there (and in drop games, when they would, would again
be considered virgin enough for the purpose of double-pushing!).

2014-09-24 13:20:31 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement Betza p and g modifiers in piece command


2014-09-24 08:28:40 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Repair piece defs with showTargetSquares off

The proxy legality checking based on marking of target square was
rejecting all moves when target squares were not highlighted.

2014-09-23 21:53:55 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Allow definition of castling in piece command

Castling is defined by the O atom, with a suffix indicating the number
of steps the King must step in the direction of the corner. Kings keep
their normal castling unless overruled by this. Can currently only be
used on the variant's royal piece, as XBoard's ApplyMove() is not smart
enough to understand displacing the corner piece is implied when it is
used on other pieces.

2014-09-23 21:49:44 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Update description of piece command in protocol specs


2014-09-23 14:52:16 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Improve SAN of Pawn moves and allow Betza e.p. definition

The move descriptions in 'piece' commands can now use 'i' and 'n'
modifiers to define a blockable double push. ('n' only work for moves
of stride 2!) The test for initial moves now is based on comparison with
the opening position, rather than on a way-to-generous heuristic. (This
still does not prove in generak the piece did not move, though, but it
does for Pawns.) A new 'e' modifier stands for e.p. capture; to implement
this the rank and file of the e.p. square are now stored in the board.
(EP_STATUS was holding the file of the moved Pawn, which in Berolina Chess
is not the file of the e.p. square!)
 The generation of SAN moves for Pawns can now reliably distinguish between
captures and non-captures even in Berolina, when the move generator informs
it whether the move was e.p. capture.

2014-09-22 22:48:50 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Retune -stickyWindows GTK

This still does not work well, and in particular windows refuse to stick
to the top of the board window.

2014-09-22 21:59:00 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix detection of screen size GTK

A gdk call that can get the screen without there first being a window
mapped to it is used now. This fixes initial sizing of the board window
on a cold start.

2014-09-22 21:57:23 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Configure a size for the Eval Graph

If no size for the Eval Graph window is configured in the master
settings file, the window will pop up as 2x4 pixels.

2014-09-22 20:39:21 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Prevent crash on specifying non-existent texture XB

By setting the texture width to 0 when the texture could not be loaded,
XBoard will never attempt to draw it.

2014-09-22 20:30:50 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Configure some themes in XBoard master settings

The -themeNames option is set to contain some themes that could be
defined with the default pieces and textures that come with XBoard.

2014-09-22 18:48:00 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Remove debug printf


2014-09-21 16:47:41 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Control Eval Graph with mouse

Let mousewheel control evalZoom, and right-click toggle between
normal view and differential view ('blunder graph').

2014-09-21 16:10:19 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Add extra font field to Option struct

Storing the font in the textValue field was too kludgy, as this field
is already heavily used in the various Option types, for storing handles,
or defining colors and callbacks. A 'font' field is now added at the end,
so that it can be optionally initialized in the rare case it is needed.

2014-09-21 13:44:11 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Repair damage done by merging with v4.7.x


2014-09-21 08:36:22 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Sort fail lows and fail highs below others

Every more recent line should always float up aove lines that were
indicated to be fail lows or fail highs through ? or ! PV suffixes.

2014-09-20 12:58:32 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Describe the new texture conventions in manual

The exact rules for how textures are scaled in response to their
filename is described in the manual.

2014-09-19 10:22:14 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Force exactly overlayed texture scaling through filename

A filename suffix -NxM.png on a texture image will cause XBoard to scale
the image as if it were the image of a complete board of N files and M
ranks, to make the square size equal to that of the current board,
and adjust the cutting parameters such that contiguous squares are cut
out of it. (If the current board format is larger than NxM, the remaining
space will be padded with white!)
 If N=M=0, N and M are assumed to correspond to the current board size,
and the bitmap will be scaled to fit the height of the board.
 If no suffix is present, the old cut-and-tile algorithm is used. It will
be assumed the texture bitmap is intended for a single square, unless the
pathname contains a file that starts with "xq", in which case it is assumed
that the file contains a Xiangqi board.

2014-09-18 23:30:09 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Allow two Pawns per file in Tori Shogi

When the board has 7 ranks it is assumed to be Tori Shogi, and the
requirement for maximally one Pawn per file is relaxed to two.

2014-09-18 23:08:36 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Test legality even when off if engine defined pieces

After reception of a piece command XBoard will mark target squares
even when legality testing is off, and now also will require the lifted
piece to land on a marked square (like when the engine highlighted them
through a highlight command).
 A problem is that this does not work for drops, as drops do not mark
target squares. So drops are always accepted. Which of course is bad
for Pawn drops. But there really is no way yet for an engine to specify
how pieces can be dropped, similar to the 'piece' command.

2014-09-17 22:32:43 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix texture scaling


2014-09-17 13:41:58 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement engine-defined pieces

When legality testing is off, XBoard now listens to 'piece' commands
to redefine its move generator. After such a piece command is received,
target-square marking also can be used when legality testing is off.
The engine keeps in charge of legality testing, though.
 Only a very basic version of Betza notation is currently understood for
the move description: mode modifiers have to appear in the order mcd,
and gpz modifiers are ignored. Multi-leg moves as in Betza 2.0 is not
implemented at all. A leading i modifier limits the moves to the first
two ranks, which is no good at all. (But XBoard does not keep track of
which pieces have moved, and this at least would work for Pawns in a
FIDE-like setup, in particular Berolina.)

2014-09-17 09:44:15 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Scale texture bitmaps that are not large enough

Texture bitmaps are now sized up by an integer factor to be at least
as large as the board (if they seem intended for a full XQ board, which is
assumed when they are larger than 256x256) or the square. This necessitated
the textures to be re-initialized in InitDrawingSizes whenever the window
size changes.

2014-09-16 19:04:19 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Ignore color arguments not starting with #

Old settings files from xpm-using version might still contain color
designations like 'white' or 'red'. Since Cairo would interpret these
as black, it is better to ignore them, so the default color is used.

2014-09-15 15:30:40 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Implement -autoInstall option

When -autoInstall is a non-empty string, XBoard will scan the system's
plugin specs directories for UCI and WB protocol, to see if new engines
were added since the last settings save. If so, they are added to the
list of installed engines.
 The variants listed in the plugin-spec files are currently ignored,
but the option is already configured to install engines for all variants.

2014-09-14 22:47:55 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix typo


2014-09-14 21:41:04 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Make inlined functions static

Clang apparently chokes on non-static inline functions,
and they were not used from other files anyway.

2014-09-09 20:29:15 -0700 Arun Persaud <arun@nubati.net>:
updated spanish translation, added new polish translation


2014-09-13 09:12:54 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Look for logo in engine dir first (GTK)

When setting engine logos, we look for a file logo.png in the engine
directory, before trying <enginename>.png in the -logoDir. This is useful
on OS X. Fix bug that checks for existence of -fd/sd rather than -logoDir,
when using the latter (also in Xaw).

2014-09-11 23:10:59 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix checkmarking of OS X menu items

The routine MarkMenuItem() was moved to xboard.c, where it can see
the OSX-specific #defines, and calls to sync_menu_bar were added
to it, and to EnableNamendMenuItem().

2014-09-11 22:48:10 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix crash of tournament dialog GTK

The font/color disambiguation of the Option.textField in Label and
ulti-line TextBox options cannot stand it if unused elements are
initialized as "" rather than NULL.

2014-09-11 19:17:55 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Let OS X display dock icon


2014-09-11 12:41:50 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix Xaw button color error

The textValue field, now also used to indicate font, was used in two
places when popping up buttons, and the font patch only addressed one.

2014-09-11 10:23:42 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Make SVGDIR a variable

Replace all occurrences of SVGDIR in the source by svgDir, a char array
that is initialized to SVGDIR, so that it can be altered at run time.
This is needed to make portable apps. Where SVGDIR was used in string
concatenation, the code had to be re-written to use snprintf to do the
concatenation at run time, through a new function LoadIconFile.
The extension of image files used there is now made dependent on the
__APPLE__ switch, so that it can use .png in OS X (which does not seem
to understand SVG icons).

2014-09-10 21:20:40 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix warning in dead code Show


2014-09-10 20:49:38 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Reorder variants, to comply with Polyglot book specs

XBoard's variant number is used as a key to access Polyglot opening
books, and thus cannot be changed without consequences. (As it was when
we introduced ASEAN and Chu Shogi.)
  We also renamed Chu Chess to Elven Chess (under which name it is now
described on chessvariants.org).

2014-09-10 19:52:12 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Reserve piece command in protocol specs


2014-09-10 18:54:45 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
suppress Alien Edition standard variants

Variants that are standard in the Alien Edition should not be
recognized as engine-defined variants: the engines that announce
them will not respond with the required setup command, so allowing
the user to select them just leads to trouble.

2014-09-10 13:51:44 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix SetWidgetFont GTK

This was testing for an empty font spec in the wrong way, with as a side
effect that some pointers to font strings would be interpreted as colors.

2014-09-10 15:56:59 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Update protocol specs


2014-09-09 21:43:39 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Let message field and button bar use GTK -messageFont

The -messageFont spec (using the default coordFont pixel size, like the
others) is now used to modify the fonts of message field and buttons
in the button bar. They pass a pointer to the appData font variable in
their Option's textField. This conflicts with the use of this field in
buttons for specifying a color, as done in the New Variant buttons.
This is resolved by checking if the first byte of the pointed-to value
is a #. The Xaw version must ignore the textField if it does not
specify a color.

2014-09-09 20:51:54 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix GTK fonts


2014-09-09 19:57:52 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Use coordFont default pixel size for other fonts

The message-font is way too large for the dialogs. A better solution
would be to redo the entire table of SIZE_DEFAULTS in xboard.h, as
this file is front-end-specific anyway.

2014-09-09 19:39:50 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Use -gameListFont in Game List

Not implemented in a very generic way, but hacked into the code that
added the handler. There aren't many other listboxes that would need
adjustable font.

2014-09-09 16:49:18 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix Chat window for Xaw build

The ICS output field and Hide button are removed from the ICS Interaction
window, to make it similar to the old Chat window. This required adding
some tests in the handlers of this window, to prevent segfaults due to
accessing it.

2014-09-09 15:53:29 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix infinite-regression problem on OS X

Apparently the OpenFile handler can already be called during
gtkosx_application_ready (or gtk_signal_connect), so we have to make
sure 'suppress' is set before we call those.

2014-09-09 10:47:41 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Correct for .Xresources form->paneA renaming in manual

In the Xaw build the name of the form widget of dialog.layout.form...
hierarchy was changed to paneA, paneB, ... when allowing multi-pane
dialogs through the Break Option (such as the Engine Output window).

2014-09-09 10:36:42 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Use GTK fonts in Engine Output and Move History


2014-09-08 20:42:11 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Include some conditional OS X fixes

DATADIR must be corrected to ~~ in the OS X app, and the info command
must be started in a different way.

2014-09-08 20:12:08 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix WB DoEvents error


2014-09-08 19:06:35 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Limit auto-extending to click on first move of PV

Even when auto-exted on clicks on the tail of a PV will no longer cause
any moves to be added to the game. Only clicks on or before the first
move can do that now.

2014-09-08 16:20:26 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Animate multi-leg in auto-play and forward event

Replaying multi-leg moves is now also done by animating both legs.
In addition, the highlights are set on final and intermediate square,
rather than from-square.
(There seems to be a great deal of code duplication here. Why doesn't
AutoPlayOneMove simply call ForwardInner?)

2014-09-08 14:45:51 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix key bindings for non-menu functions

The accelerator keystrokes for next game, previous game and debug mode,
which don't have menu items, and are grouped in the menu table 'noMenu',
were not treated in the building of the translation table, as 'noMenu'
does not occur on the menu bar. An extra iteration is done now to
treat this menu group.

2014-09-08 14:24:59 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix Xaw key bindings

The routine to automatically construct the translation table for
accelerator keys did not respect the rule that more specific keystrokes
(like :Meta<Key>Home:) have to be defined before more general ones
(like :<Key>Home:). As a result Alt+Home would activate the Revert
item, intended for plain Home.

2014-09-08 14:20:08 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Define default font names


2014-09-08 09:25:18 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix MSVC problems

Book.c did not include config.h, and thus missed the redefinition of
snprintf (which is not known in MSVC). The use of 'inline' in parser.c
also needed an MSVC redefinition, now added to winboard/config.h.

2014-09-08 09:22:55 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix GTK font patch


2014-09-07 23:35:22 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Replace strcasecmp by StrCaseCmp

The MicroSoft compiler does not know strcasecmp, so we better use
our own function for this everywhere.

2014-09-07 23:28:29 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Let GTK build pay attention to font arguments

Multi-line text edits now use the initial value of the Option.textValue
field as a pointer to a (char*) holding a font name, to be used for
the text displayed in it. (After which they overwrite the field with
the textView handle.) The font options -tagsFont, -commentFont, icsFont,
-moveHistoryFont and -gameListFont are now activated in the GTK build,
to control the corresponding memos.

2014-09-07 13:58:15 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix auto-display comment option in General Options

This options had the same ID as auto-Queen, making it ignored.

2014-09-07 13:43:48 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Call DoEvents during time-consuming operations

Processing of a game file for reading it, searching positions, saving
selected games or creating a book from them can take minutes. To keep
the application responsive during that time, the main loops of those
tasks calls DoEvents(). A possible risk is that the user can now do
too much, and will initiate actions that interfere with the tast in
progress.

2014-09-07 13:32:49 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix GameListHighlight WB

The routine to highlight a game line in the Game List listbox of WinBoard
used an extremely inefficient linear search to find the game amongst the
selected games. This has now been replaced by bisection, after finding
an upper limit to the number of entries in the listbox by doubling
a trial size.

2014-09-07 12:10:50 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Provide DoEvents function in front-ends

A routine to process all pending events in a non-blocking way is now
provided in the Xaw, GTK+ and WinBoard front-ends, so that time-consuming
tasks (such as PGN loading or opening-book creation) can call it to keep
the user control over the application alive.

2014-09-06 23:13:01 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix syntax error in bitbase code


2014-09-06 22:35:22 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Describe Save Selected Games menu in manual


2014-09-06 22:29:43 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Let Save Games as Book only use selected games

The book-creation code now only takes games into account that are
selected for display in the Game List.

2014-09-06 22:20:19 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Rewrite key-binding section of manual

This section still wrongly described the key-binding procedure from
before the rewrite of the XBoard front-end.

2014-09-06 13:21:13 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Connect scroll event to Graph Option in GTK

The mouse wheel is now detected in graphics widgets, and passed to
the user handler as button 4/5 mouse events.

2014-09-06 13:15:55 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Make sure node count is positive

Engines that print the node count as an int overflow into negative values
when the count exceeds 2G. This is really an engine bug, but as it
interferes extremely annoyingly with the alignment of the Thinking Output
XBoard now repairs it by adding 4G to undo the overflow.

2014-09-05 14:03:49 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Connect mousewheel to Forward/BackwardEvent (XB)

The mouse buttons 4 and 5 are now processed in the handler for the
board widget.

2014-09-05 13:41:34 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix size of time in Engine Output window

When the thinking time exceeds an hour, the format hr:min:sec is now used,
in stead of min:sec.cents. This prevents the field from extending beyond
the tab stop when thr hours kick in, and should work upto 100 hrs.
(Going beyond that at the moment does not seem useful.)

2014-09-04 22:57:07 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Verify if font-spec looks like one in Xaw

To make XBoard resistant to mixed use of Xaw and GTK builds, which
would not understand each others font specs stored in the settings
file (they have in common), we now determine if the value looks like
an X-font. If it doesn't, the Xaw version ignores it. The GTK build already
ignored it if it did lool like an X-font spec.

2014-09-04 17:03:55 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix alignment in Engine Output window

Numbers above 1000M would become too wide for their field. This is
solved by introducing the abbreviation 'G'. Values below 10M or 10G
now get an extra digit in their fraction, so that the precision is
at least 1%.

2014-09-01 14:44:54 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Add Save Selected Games menu item

A menu item to save all selected games in the Game List on a single
file is added.

2014-09-01 12:03:58 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Add final piece count to search criteria

The Load Options dialog now has a text field in which a range can be
entered (like 8-10). Position search will then only select games that
had their final number of pieces in this range.

2014-08-29 21:15:29 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix castling rights on using -lgf

The dummy InitPosition to force loading pieces was taking place after
InitBackend2(), while the latter could have already loaded a game
from an -lgf argument (which would then be spoiled by the InitPosition).
It has been moved to before InitBackend2() now.

2014-07-29 15:03:04 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix crash on double-click in Game List Tags

A double-click in the Game List Tags listbox has no function, but it
crashed XBoard, because the textValue field of teh ListBox Option definition,
supposed to contain a pointer to the user-supplied handler, contained a
pointer to an empty string.

2014-07-02 14:34:53 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix typing of null moves

Entering a null move through the move type-in could fail because the
to-square coordinates (not used in its encoding) could have invalid values,
which would reject the move in an early stage of UserMoveEvent().
Null moves are now exempted from this vaidity test, while variants where
null moves are legal now accept them in any gameMode.

2014-07-02 14:04:22 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix adjusting clocks in Xaw version

The handler attached to Label options failed to flip the sign of the
option number when a button other than Button1 was pressed, so that time
would always be subtracted on Shift + ClockClick and never added.
Shift now also suppresses interpretation as null move when playing.
(In Analyze / EditGame mode it is still null move starting a variation.)

2014-07-02 12:37:18 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix null-move entry during play

Null moving by clicking the clock was only accepted in Edit Game and Analyze
mode, while in Chu Shogi and spin-offs it is also allowed during play.
An F_NULL_MOVE flag was introduced to make it easy to test for such variants,
and Machine White/Black mode now also interpret clock clicks as a null move
if the opponent's flag has not yet fallen.

2014-04-09 09:53:56 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Let EditPosition double-click on piece promote it

For pieces with Shogi-like promotion (i.e. if the promoted piece has
pieceToChar defined as '+'), let a second click on it in Edit Position
mode promote or demote it (as the case requires) when the piece menu is off,
rather than remove it (as was the old behavior.

2014-04-09 09:19:43 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Make EditPosition pallette work in Asian variants

Construction of the pallette position was not working in variants that
did not have Pawns start on 2nd rank, or that had multiple ranks of pieces.
Now the algorithm is applied to all ranks.

2014-04-05 23:00:17 +0200 H.G.Muller <hgm@hgm-xboard.(none)>:
Fix crash on opening Tags window Xaw

Suppressing the optional book button made the Xaw generic popup crash,
because the following button had the SAME_ROW modality, and was thus
positioned w.r.t. a non-existent widget.

2014-04-05 20:39:55 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix writing of Chu-Shogi FENs

Promoted pieces were getting the wrong ID letter, of the piece demoted
by 11, rather than 22.

2014-04-01 18:20:29 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Allow pseudo-engines to adjust the clocks

Implement new options -first/secondPseudo, which declare the corresponding
engine to be a pseudo-engine, which will then be granted super-powers
through some new engine->GUI commands. For instance, to adjust the time
remaining on the GUI clocks through wtime and btime commands (using msec).

2014-04-01 12:00:14 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix piece inscriptions


2014-04-01 11:16:48 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix reading Chu Shogi FENs

The X (for Phoenix) was interpreted a 10 empty squares.

2014-03-17 23:26:08 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix loading positions in engine-defined variant

Paste Position did not work when an engine-defined variant had redefined
the pieceToCharTable, as the position would be interpreted and sent to the
engine before receiving the setup command. Copying the pieceToCharTable
received from an earlier game to appData cures this, because it will then
be used by InitPosition, which would reset the table to default otherwise.
 Another problem was that setupSpoiledMachineBlack would not be reset
after a game with no moves, leading to a spurious 'go' command to the engine.

2014-03-17 23:05:33 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Render inscriptions on Chu-promoted pieces in red


2014-03-17 22:56:07 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Allow writing text on pieces

A new persistent string option -inscriptions allows the user to define
characters to be written on the various piece types. The string is
interpreted as UTF-8. This option is useful for on-the-fly synthesis
of Shogi pieces, where the string can contain the various kanji in
the canonical pieceToCharTable order. Except that only a single color
has to be given.

2014-03-12 22:08:04 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix initial sizing of board

Previous patches had destroyed the ability to obey the -size option.

2014-03-10 21:15:30 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix width of menu bar

The truncated name might not have been terminated with '\0'.
TinyLayout now specifies the number of characters kept in the menu names.
In the OS X version truncating is never needed, as the menu bar is
at the top of the screen, independent of board size.

2014-03-09 20:38:51 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix warnings fonts patch


2014-03-02 23:10:50 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Implement (clock-)font handling in GTK

The options -clockFont, -coordFont and -messageFont are activated,
and supposed to have string values like "Sans Bold 12". The default
values have the point size as 'wild-card' %d, however "Sans Bold %d",
and are expanded by a printf with the point size taken from the defaults
table. The expanded versions are saved back into the settings file,
prefixed with a size tag, as usual.
 So far -clockFont is the only font that is actually applied while
writing.
 The 'clock kludge' was interfering with the initial sizing, as for some
sizes the initial clock-widget height happened to be equal to the final
height. It was therefore abandoned, and we seem to be able to live without
it.
 Sizes smaller than 49 were dominated (for 8-wide boards) by the menu bar.
Therefore tinyLayout now uses single-letter menu labels.

2014-03-02 20:45:57 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Also search indirection files in user's .xboard tree


2014-03-02 18:53:17 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix error Engine Output text highlighting

In wengineoutput.c higeTextStart and End where decalred both static and
extern, which somecompilers justifiably don't like.

2014-03-01 16:09:09 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix Board-dialog bug WB

The drawing code for the sample square did ot restore the correct brush.

2014-02-28 21:48:19 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix various warnings

Mostly written but never read errors. Also removed a debug printf that was
causing a format warning.

2014-02-26 16:34:52 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Kludge repair of expose after startup resize

Redraw of the main board after resize is delegated to DragProc, which is
scheduled after a delay, to make sure the user is done sizing (mainly to
reposition docked windows). But at startup the program-requested resize when
an engine does not support variant normal and needs a different board format
takes so long that the finale size expose occurs after the DragProc timeout.
As a result, the board was never drawn on the resized canvas.
  This is quite hard to fix in a satisfactory way. For now the suppression
of board redrawing during resize is disabled during the first 10 seconds
after startup, as any resize during that time is likely to be the result
of a program request.

2014-02-26 13:08:50 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Use first engine as default for second

Rather than having the same fixed default for both engines, we noww only
have a fixed default for the first, and use the actual first as default
for the second. One advantage is that this guarantees the second engine
plays the same variants as the first by default.

2014-02-23 23:51:20 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Allow adding played move to book

A button is added in the Edit Book dialog to prime XBoard for adding the
next move entered through the user interface to the book window, without
actually playing it.

2014-02-23 22:54:55 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Play move right-clicked in Edit Book dialog

A handler for button-3 down-clicks is added to the Edit Tags dialog,
which fakes the clicked word was typed in the move type-in box.

2014-02-23 19:23:28 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix pop-down of ChatDlg and TextMenuDlg from menu

Unlike other windows these two would not react to clicking the menu item
once they were up. Now they pop down in that case.

2014-02-21 11:10:41 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Preserve window width on board-format change

When a variant switch would cause recalculation of the board window,
we now first recalculate square size to keep the window width fixed.
Also adapt initial width on high boards.

2014-02-18 22:37:31 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Equip Board Options dialog with themes listbox

The option -themeNames is made persistent for XBoard, and its value is
displayed in a listbox added to the View -> Board dialog. The user
can select themes from this, and manipulate the contents similar to
the engine list in Load Engine. A textedit allows the user to define a
theme name for the current settings, and doing so adds the theme to the
-themeNames list.

2014-02-07 11:22:44 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Implement copy function in ICS Text Menu

A command consisting of the word $copy will place the clicked word on
the clipboard. When there was already text selected during the click,
that selected text will be used for $copy and $name, rather than the
right-clicked word.

2014-02-07 09:48:31 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
When ICS Console open EOF from keyboard is no error

In case the user started without a terminal, we just close the input source,
and suppress future writing to the xterm.

2014-02-06 18:46:21 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Leave xterm at start of new line after quitting XBoard


2014-02-06 18:36:41 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Also display channel tell in ICS Console during private chat

Channel tells of players for which a chat wa opened were displayed in that
chat, as well as in the channel chatbox, if the latter was open. But when
there was no channel chatbox it would not appear as colorized channel tell
in the console. This is now fixed, and in the private chat box such a
message is indicated as 'broadcast'.

2014-02-05 23:08:50 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Write broadcasts also to private chatbox of talker

The shouts or channel chatboxes capture messages first, but if a private
chat with the talker is opened, it now goes there too. Prefixed with
the type of broadcast.

2014-03-02 13:18:34 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix Xaw Chat Console


2014-02-05 22:38:09 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Let Ctrl-O key open chat for last talker


2014-02-05 22:15:03 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add End Chat button

This button clears the current chat, so that it becomes free for
assignement to another partner.

2014-02-05 21:59:10 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Ctrl-N in chat opens empty chat

This key isequivalent to pressing the left-most New Chat button.
If no unused chat present, it the command will be ignored.

2014-02-05 21:29:42 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Preserve unfinished input lines during chat switch

This is problematic, as apparently we cannot write in widgets that are
active with key hits, so attempting to write it back in response to Tab
did not work. This is now done with a delay, for lack of a better method.

2014-02-04 13:57:16 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Recognize Esc and Tab in ICS Console input

Esc is used to close the chat, or transfer focus to the board window.
Tab is used to navigate between chats (or open the first). Priority
is given to dirty chats.

2014-02-03 20:35:31 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Let ICS Console pop up GTK in stead of ICS Input Box


2014-02-03 19:45:44 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add context menu to ICS console XB-GTK

A right-click handler is added to the output memos of the ICS Interaction
window. It pops up the ICS text menu. It is remembered whether the menu
was already up; if not, it will be popped down after a command is selected
from it. Like the WinBoard context menu, the commands will be able to use
the clicked word.
The -icsMenu 'command' $chat is now recognized as a special case, not to
be sent to the ICS, but executed by XBoard. It will assign one of the
five chats to the clicked name. If no empty chat is available, the last
chat will be used (but not cleared).
 The -icsMenu in the master config file is now configured to contain
this command, but unfortunately this will not become effective for users
that already have a settings file.
The Text Menu now puts $input commands in ICS Console Input field when the
ICS Input Box is not up.
 The text placed in the input field for completion (as per $input directive)
turned out to be selected in GTK, so that when you stated typing, it was
erased again!
 By making the Text menu dialog subject to WindowPlacement control,
and interrogating the poition of the ICS Console window, the Text Menu
can be popped up such that the mouse pointer is on the bottom-left button.

2014-02-02 21:17:00 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Integrate ICS output into Chat Window

An extra pane has been added to the Chat Window to display the ICS output.
In GTK this output is colorized just as in the xterm. For Xaw this is not
possible. This pane is always visible, and the chat output can be hidden,
to double its size. Entered input will go to the selected chat (when not
hidden) as a tell or shout, or to the ICS as a command. The Chat-Partner
field is hidden together with the chat pane and the Hide button. Each of
the five chats now has a dedicated button; for unused chats (empty Chat-
Partner field) these now gets the text 'New Chat' inscribed on it.
Focussing is made more intelligent (to the Partner field when shown and
empty, otherwise to the input field).
The Up and Down arrows now work in the input field to recall old input.
 ScrollToCursor() was broken in GTK, because a mark was made for only
a single buffer. Now each buffer has its own end mark.
 Another problem was that using ScrollToCursow immediately after a memo
was shrunk (because a neighboring one expanded) it would still use the
old memo size, so that the bottom part remained out of view. Delaying it
50 ms solved that problem.
 A new option -icsUp was added to store the visibility in the settings.
It is updated only in ICS mode. The ICS Console pops up automatically
when the stored parameters say so. It is also made sticky.
 Unlike the xterm the textView does not extend background color to unwritten
positions behind the line. A separate GTK call is now added to set the
unwritten part to the same color as the background in normal text.
 A  control was added in the ICS Options dialog for setting the colors
of normal text (-colorNormal).

2014-02-02 17:27:26 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add logo-size control XBoard

This is mainly for the benefit of OS X useres who cannot find their
settings file. It requires a restart to become effective, which is
really sub-standard.

2014-02-02 17:18:56 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add Edit Engine List menu item to XBoard

In OS X the settings file is hard to find...

2014-02-02 16:41:13 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix warnings

Some unread variables are commented out, rather than removed, because they
can be useful for future development.

2014-02-02 16:33:34 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix moves of Spartan Captain

The rewrite of the move generator did no longer pay attention to rookRange,
so that the Spartan Captain moved like a Rook.

2014-02-02 16:18:07 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add Option type 'Skip'

The value -1 used to suppress rendering of option controls was drawing
error messages, because enums are unsigned. So a new member Skip was added
to the enum.

2014-02-02 14:40:16 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Detect Fischer castling in FENs

Castling rights with unorthodox piece placement imply Fischer castling.
Make sure shuffling is done after detecting this, because it means K & R
can also be shuffled (as long as K remains between both R).

2014-02-02 14:09:19 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Allow shuffling indicators in FEN

If the back ranks are enclosed by < > in a FEN, the route to shuffle
initial positions is invoked on it.

2014-02-02 13:51:37 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Allow wild-cards in FEN

A question mark in a FEN will now be interpreted as a piece randomly
chosen from the holdings. (From which it will be removed.) Intended
for use in symmetric black/white pairs; when used on asymetric locations
or with non-identical holdings the results will be undefined.

2014-02-02 10:40:18 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix Seirawan reverse-castling animation

In Chess960 animation of castlings is suppressed, because the move is encoded
as KxR, so that the to-square in general is not the final destination of the
King. So it is just confusing to the eye to see the King slide to the Rook,
and then jump to an entirely different square. In S-Chess castlings can
similarly be encoded as RxK (to indicate gating at the Rook square), and
animation of those moves should be suppressed for the same reason. But it
wasn't, and in Fischer castlings denoted as RxK this even caused the Rook
to disappear when it didn't need to moved.
 Perhaps one day a smart method of animating Fischer castlings can be
designed. (E.g. remove Rook, animate King move to true to-square, place Rook?
But what if the King doesn't move at all? Animate Rook then? Normal castlings
now animate the King, and then just displace the Rook.)

2014-02-02 10:22:25 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Make Fischer castling generally available

A new volatile option -fischerCastlings allows every variant to be played
with Fischer castling, which so far was only allowed in variants FRC/CRC.
This only makes sense in shuffle variants, where King and Rooks might not
be in their orthodox locations. But even when they are, it has consequences
for the user interface, which will require dragging King on Rook as the only
way to enter castling. Like -shuffleOpenings the setting will be cleared
as soon as you switch variant (unless you switch to FRC/CRC, where they
Fischer castling is on by default). It can be set from the New Shuffle dialog.

2014-01-19 22:41:52 -0800 Arun Persaud <arun@nubati.net>:
updated po/pot files


2014-01-19 22:19:10 -0800 Arun Persaud <arun@nubati.net>:
new version number for developer release


2014-01-19 22:15:33 -0800 Arun Persaud <arun@nubati.net>:
updated Dutch translation


2014-01-19 20:44:33 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Put ponder checkbox in Common Engine dialog WB

This is yet another duplicate of the one in General Options and Adjudiactions.

2014-01-19 20:27:59 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Expand %s in -openCommand to DATADIR and fix OSX settings-file name

This can be used in OS X to biggy-back a cd to a directory in the App
on the open command.

2014-01-18 23:42:01 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add some checkboxes in General Options dialog WB

Checkboxes are added for -variations, -memoHeaders and -autoExtend.

2014-01-18 23:26:03 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix warning in CheckTest


2014-01-18 22:40:27 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix printing of engine-output headers

They were not printed on analyis of the first position after opening
the Engine Output window.

2014-01-16 23:07:28 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix sticky windows on Win8


2014-01-16 22:54:32 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Drag touching edges together (WB)

When an auxiliary window is dragged by an edge, the snap code now checks
if that edge was touching another window. If so, that other window is sized
to keep touching the dragged edge. Unless it was also touching the main
window.

2014-01-15 22:35:23 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Implement new mate-score standard

Matescores are now indicated in the protocol as +/- 100000+DTM,
and in the Engine Output window as #DTM or #-DTM.

2014-01-15 21:28:51 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Put fail-high/fail-low indicators in protocol specs


2014-01-15 20:43:55 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
switch to new tbhits protocol

The extended info is now expected as 3 bare integers, the last one
separated from the PV by at least one tab character.
The protocol extension is described in the CECP specs document.

2014-01-14 22:37:30 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix sorting of Engine Output

The moveKey of the top line was never stored, so that more recent
lower scores could not pass it even if they had the same move.

2014-01-14 22:35:22 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Handle fali-low & fail high

A ? or ! at the end of the PV are now also written behind the score.

2014-01-14 20:31:46 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Implement extended thinking output

The PV field can now start with three quantities, selective depth,
engine speed and tablebase hits, disguised as PGN comment. XBoard will
strip that off, and allow them to be displayed in separate columns of
the engine-output window.

2014-01-14 14:46:47 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Allow hide/show of columns in Engine Output

A new persistent Boolean option -memoHeaders enables a new feature:
text headers are shown above the depth, score, nodes, time columns in the
engine output. Right-clicking them will subsequently hide the corresponding
data, and move the column header to above the PV in a 'not shown' list,
where it can be clicked again to re-open the column.
Configure maseter settings to switch -memoHeaders on by default.

2014-01-14 09:55:40 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix ManProc for OS X


2014-01-14 10:02:22 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix access to ~~/themes/conf for OS X

In OS X DATADIR is a variable.

2014-01-13 19:23:30 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add 'Continue later' button to Tournament dialog XB

To do GenericReadout without triggering a call to the OK function,
we introduced a new value for the option-number parameter, -2.

2014-01-13 17:32:28 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix Tournament Options dialog

Deleting a Label option had renumbered the remaining Options, breaking
the dialog and causing a crash when an engine was selected. A considyency
check is now added to catch this condition, and throw an error popup
when we mess up again.

2014-01-13 17:09:01 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Allow popup of TC and Common Engine from Tournament dialog

A new dialog class MasterDlg was added, similar to TransientDlg,
but to allow two dialogs to be up simultaneously. The Tournament Options
dialog was migrated to this class, so it can now get buttons to pop up
the Time Control and Common Engine dialog, like WinBoard. It even has
a button for the General options dialog.
Special care had to be taken to prevent popdown of the MasterDlg when
a Transient child is still up (similar to browser or error childs).

2014-01-13 16:21:59 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add --show-config special option

This prints a list of configure parameters, and then exits. If a second
argument is given, it only prints the value of that config parameter,
so that tricks like cp *.svg `xboard --show-config Datadir`/themes/chu
can be used in install scripts of packages with piece images.

2014-01-13 14:25:30 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Overhaul kill code

Engines were not always forcefully killed in XBoard, which could make
the tournament manager hang when an engine became unresponsive to "quit".
The various levels of killing are now revised, and the -delayAfterQuit
is absorbed in DestroyChildProcess(). There is a hard killing mode 9,
which immediately sends SIGKILL, intended for engines that have already
proven to malfunction or could not start at all. In other cases SIGTERM
is sent according to specs, but a SIGKILL is scheduled -delayAfterQuit + 1
sec later just in case the SIGTERM would be ignored.

2014-01-13 13:38:47 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix name of master settings file in OS X

The name needs to be relative to the bundle path, but ~~ expansion
did not work for compiled-in defaults, as the expansion takes place
when parsing an option. The dataDir is now also kept in an array rather
than as a pointer, to allow it to appear in statically initialized tables.

2014-01-08 16:25:25 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix Chu promotion of L, HM and GB

The promotion popup would not be shown for HM and GB with legality testing
on because the piece types used for them were wild-card pieces. The Lance
was treated as a Pawn in HasPrmotionChoice().

2014-01-08 16:09:15 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Remove redefine of DATADIR that leaked in from v4.7.x

2014-01-05 12:52:46 -0800 Arun Persaud <arun@nubati.net>:
updated po/pot files


2014-01-05 12:51:50 -0800 Arun Persaud <arun@nubati.net>:
new version number for release 4.7.3


2014-01-05 10:58:48 -0800 Arun Persaud <arun@nubati.net>:
updated copyright to 2014 in menu.c


2014-01-05 10:37:56 -0800 Arun Persaud <arun@nubati.net>:
removed .DS_Store file from git


2014-01-05 10:34:08 -0800 Arun Persaud <arun@nubati.net>:
Updated copyright notice to 2014


2014-01-05 10:27:53 -0800 Arun Persaud <arun@nubati.net>:
cleanup some trailing whitespaces


2013-12-29 13:46:40 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix crash on use of dialog Browse buttons GTK

The code to set a filter did not expect multiple extensions, and would
cause a buffer overrun when there were.

2013-12-27 20:12:23 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Make building of Windows .hlp file optional

By using && and || operators in the build recipe, the building of WinBoard
with Cygwin is now not aborted whe the Help Compiler is not installed.

2014-01-05 11:47:18 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Use __APPLE__ compile switch for OS X

It seems that on Apple machines the compiler will always define the
symbol __APPLE__, so this is now used instead of OSX, which had to
be defined with a -DOSX compiler flag. The gtkmacintegration header
is now using <> in stead of "" in its #include. (Strange enough both
were working.)

2014-01-04 12:18:30 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Expand ~~/ to bundle path (OSX)

DATADIR is #defined in WinBoard as "~~" to make the expansion a no-op
there. Same for Linux compiles. For OS X it is #defined as a variable,
obtained from GTK-OSX at startup.

2013-12-29 23:27:12 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Activate ManProc in GTK


2013-12-29 22:14:06 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Remove use of strndup

The use of strndup was actually unsafe, as the length was passed using
an ordinary strlen. strndup was not availableon OS X 10.6, and is now
replaced by an ordinary strdup, which should be save enough.

2013-12-29 22:09:10 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Put some OSX code into gtk version

Under control of #ifdef OSX code is added to integrate the GTK front-end
into OS X. This involves moving the menu bar to outside the window, and
catching the signal that OS X sends to running applications when opening
another instance was requested. Som files are renamed to conform to the
file tree of the OS X App package.

2013-12-27 21:47:55 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Cure weirdness when dragging outside of board

Dragging outside the board (when people maximized the window) did not
properly restore background, (leaving a trail of dragged piece). Filling
the canvas with white when it is creatd cured this. The problem most likely
was that unitialized cairo surfaces had transparancy, and writing back
transparancy doesn't helpmuch to erase things.

2013-12-09 14:52:32 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix sorting of lines in Engine Output

Lines tat were fail-lows or fail-ighs should get the scores on the basis
of wich they are sorted corrected to that of te line (for the same move)
superceding them, so that other moves can rise to above them, when they
score higher than the corrected score.

2013-12-05 20:08:35 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Make filler variant button inactive


2013-12-05 17:50:21 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix packing of FRC castlings

The routine for packing moves to the internal database format was not
working for FRC castlings, totally messing up position search for FRC
and CRC.

2013-12-05 16:23:53 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix Xaw file-browser New Directory

Prevent people from using the error messages that will be printed here
as directory name by pressing New Directory a second time after first
pressing it with an empty or existing name.

2013-12-05 16:14:02 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Widen Xaw text entries for larger square sizes

Presumably people will use larger fonts if they use larger board squares.
(And in any case te default fonts would be larger.) This means they need
more space to see what is the current value of text options.

2013-12-05 15:47:26 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Remove width limiting of shuffle checkbox


2013-11-28 13:30:40 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Allow castling and e.p. to be edited in opening book


2013-09-16 22:32:32 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Repair FRC A-side castling legality testing

XBoard allowed A-side castling when a rook started on the a-file,
and the b-file was blocked.

2013-09-16 22:27:50 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Repair XBoard from node-count patch

u64 was #defined in backend.h, which unfortunately was almost alwats
(Not sure why WinBoard did not suffer from this!) This, and related
definitions have now been moved to common.h (which is always #included
first).

2013-09-15 20:13:06 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
WinBoard multi-monitor support

A patch provided by Tim Kosse

2013-09-15 19:28:54 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix node count range

The node count was still clipped to 32 bit because one of the structs
it passes through (FrontEndProgramStats) had the nodes field declared
as an unsigned long.

2013-09-15 19:12:42 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix adjudication of Giveaway stalemates

The case where the side-to-move would win by stalemate was not treated
in the Adjudicate switch.

2013-08-30 17:33:59 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix buffer overflow in parser

When a line ended with a comment, the parser had to read a line ahead
to check if there was a PGNresult following, without being able to
throw away the previous line yet. This could lead to buffer overflow
in a long game where every line ended in a comment. The buffer now is
cleansed of already parsed data when starting on a new syntactical unit.

2013-08-27 22:55:52 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix -zippyVariants option

Due to the change of cps->variants from (char []) to (char *), copying
zippyVariants to it on behalf of v1 engines would no longer work. This
is fixed by using ASSIGN() in stead of safeStrCpy().

2013-08-26 18:25:55 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix initial board sizing WB

The initial board size was based on the assumption the board would be 8x8.
This could push auxiliary windows off screen if the board was smaller,
after which EnsureOnScreen would undock them and park them in the upper-
right corner before the board assumed its true size. InitPos() is now called
before the initial sizing to et the board format correctly, InitDrawingSizes()
had to be made resitant to calling before a board size was picked for this
(because InitPos() might call that too).
  ActivateTheme() now uses the new board size, not the previous one (as was
indicated by size -2), so that -boardSize specifications inside a theme
will take effect.
  Board size petite should not be forbidden for Shogi; all pieces exist there.

2013-08-26 14:40:57 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix grabbing of selected piece

A second static click on a selected piece should deselect it, but an attempt
to drag an already selected piece should not 'bounce off'. It was already
possible to move a selected piece to another square, but the dragging was
not animated. This is now fixed, by only skipping the code to start dragging
when the second click is part of the only-move double-click for only capture.
(Before it was skipped on any second click.)

2013-08-26 13:35:47 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Allow entry of negative numbers in spin control (WB)

Windows numeric controls do not accept negative numbers, so in case the
range of a spin option can go negative, the style ES_NUMBER should not be
set on the control.

2013-08-26 11:37:27 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Allow drops / promotions/ deferrals to be edited into book

For no real reason only NormalMove would be accepted when parsing the list
of book moves.

2013-08-26 11:28:54 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix GUI book after setup position

The routine to send move to the engine would automatically append a 'go'
is a preious setboard had put the engine in force mode, but this should
not happen on a book hit, where the engine should even be put in force mode
if it was not yet there.

2013-07-05 18:32:46 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix book creation

Due to duplicate use of the same file pointer variable the game file would be
closed before any games were read from it.

2013-05-20 11:14:01 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Make PGN parser immune to unprotected time stamps

The parser choked on time stamps of the form dd:dd or dd:dd:dd if one
of the fields happened to be 00, because it wouldmistake that for a
non-compliant king-side castling. By excluding this interpretation when
the 00 is immediately preceded or followed by ':' this is now prevented.

2013-05-05 08:03:00 -0700 Arun Persaud <arun@nubati.net>:
updated po files for new release (make distcheck)


2013-05-05 07:57:22 -0700 Arun Persaud <arun@nubati.net>:
new version number for release 4.7.1


2013-05-05 07:53:54 -0700 Arun Persaud <arun@nubati.net>:
updated Changelog, NEWS, etc.


2013-05-02 22:47:44 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Update zippy.README

It is explained how to generate multi-line commands for -zipyGameEnd.

2013-04-01 20:43:36 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix Xaw file browser

The 'next page' entry of the Xaw file browser contained a \177 character
to keep at sorted at the bottom. This caused trouble for gettext.
Now the sorting range is adapted so the 'next page' is never part of it.
  More seriously, the listbox fill code had a wrong cast when testing for
an empty listbox on the first element, so that it only tested the low byte
of the pointer, in stead of the entire pointer for NULLness. As a result
the contents of the listboxes could suddenly disappear when even when it
was non-empty, because the first filename was allocated at an unlucky address.

2013-03-22 19:23:25 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Make Chat Box window obey -topLevel option

It was always popping up as a dialog of the main board, covering it.
One could argue it should always be to-level.

2013-03-19 13:20:24 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix scrolling of Chat Box

Seems the problem is that scroll_to_iter does not work until after some time,
needed to calculate line heights. Use scroll_to_mark instead, after clumsily
adding a text_mark.

2013-03-17 10:33:31 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix GTK SetInsertPos

SetInsertPos() was not yet ported to GTK, with as a result that the cursor
was positioned at the start of a line recalled in the ICS Input Box,
rather than at the end. The code also called SetInsertPos to force scrolling
to that point in the text (which in Xaw is an automatic side effect). So for
multi-line text edits this function calls ScrollToCursor(). For some reason
this does not work when replacing texts in the Chat Box memo, when switching
chat, although it does work in the same memo when a line is added to the
existing text.

2013-03-07 21:38:43 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Print score of final position in Analyze Game

The score/depth of the final position of the analyed game is added
(as an explicit comment) behind the PV of the forelast position.
To make this work, a new 'addBraces' mode (3) had to be added to
AppendComment, which suppresses stripping of the score/depth from
the comment.

2013-03-12 18:50:08 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix horrible bug in reading scores from PGN

Reading scores like -0.94 flipped the sign, because the integer part
-0 would read as 0, and thus was no reason for sign flipping.

2013-03-10 12:43:59 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix disappearence of pieces that were moved illegally

Dragging damages the from-square, but it is not marked as such, because
selective board redraws to update highlights and such would draw the piece
that is really still there, making it flash on the from-square before
the position after the move is drawn. This, however, backfires when the
move was illegal, as then there is no draw of a new position, and the
from-square stays empty. For now this is solved by making DisplayMoveError()
order a total redraw. (As this is rare, only done in case of an error,
and might even throw up a pop-up, the performance cost of this is acceptable.)

2013-03-20 17:50:33 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix quoting of book name in tourney file

The argument of -polyglotBook was written to the tourney file unquoted,
which effectively made it disappear if it was an empty or blank string.

2013-03-07 18:11:31 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Provide compatibility with Alien Edition setup command

The setup command of the Alien Edition has a board-size specifier
(possibly as prefix on a parent-variant name) between pieceToCharTable
and FEN. The regular version would choke on that, precluding engines
designed for the Alien protocol extension to have their setup processed
when running under regular XBoard. This patch makes it skip the extra info.

2013-03-07 17:52:02 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix saving of analyzed game

Anotated games after batch-mode Analyze Game were not auto-saved,
because the GameEnds() call supposed to do that was called with a NULL
argument for resultDetails. Now it is called with the resultDetails as
pre-existing in the gameInfo, or an empty string if none exist. This bug
was masked by the fact that in standard game ends (mate, etc.) the
analyzing engine would declare game end and supply a message.
 Also makes a distinction between Analyze Game and regular Analyze mode
when resuming analysis after loading a game; only the former now would
trigger a batch analysis.

2013-03-07 17:48:45 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix spurious popup after batch-mode Analyze Game

A batch analysis of games was not properly terminated, because the game
mode was tested after it had already been changed. No oldGameMode is
tested in stead.

2013-03-02 22:21:29 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Let XBoard -autoBox option also affect move type-in


2013-03-02 22:17:31 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix GTK box popup

Key values > 255 are used for PgUp, PgDown etc., so to pop up only on
printables needs a two-sided test.

2013-03-01 12:50:06 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix fatal error on unsupported board size

When a board size unsupported by the engine was requested through the
New Variant dialog, this was discovered while the dialog was still up,
causing the error message to pop up as a child of the New Variant dialog,
so that it would immediately be taken down as the latter closed.
As a result it could not be confirmed, so that XBoard would not exit, and
was left in a pathologic state (ignoring window-close). This is fixed
by performing Reset (which detects the error) after PopDown of New Variant.

2013-03-01 12:39:40 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Accept setup command for non-standard board size

The code to test if board-size is standard for the current variant
has been moved to a subroutine, to allow its calling when receiving
a "setup" command from the engine.

2013-03-01 11:57:01 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix buffer overflow in feature parsing

String features (variants, egt, myname and option) relied on a buf[MSG_SIZ]
for processing their value. The Nebiyu engine had combobox options that
required more than that. All string features are now stored in allocated
memory rather than in static arrays, and StringFeature allocates a buffer
of sufficient size for them. Only limitation now is the low-level input
buffer in the InputSource threads, whih is a static buffer of 4096
(=INPUT_SOURCE_BUF_SIZE) characters.

2013-02-27 23:44:55 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix vertical sizing of GTK board

The clipping of the boad at the bottom was due to resizing of the clock
height when we first write it with a big font. This patch delays calculation
of the final size until after the clock height changes.

2013-02-26 21:13:32 -0800 Arun Persaud <arun@nubati.net>:
Updated German translation


2013-02-26 20:27:24 -0800 Arun Persaud <arun@nubati.net>:
Updated Ukrainian translations


2013-02-26 17:59:43 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix of argument error

Oops. Pevious patch for skipping unknown multi-line arguments
was still buggy.

2013-02-26 17:44:10 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Let popping up of WinBoard chatbox for channel open it

WinBoard now issues a +ch command for a channel when you pop up a chatbox
for that channel. A work-around is added for the fact that the public
FICS seems to ignore the first such command.

2013-02-26 17:27:59 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Make skipping of unknown option smarter

To make XBoard better resistant to the presence of disbanded or future
options in its settings file, encountering those is no longer a fatal error.
But the error recovery only consisted of skipping the rest of the line.
This fails badly on string options with a multi-line value, which is
enclosed by {braces}. We now detect that case, to skip to } instead.

2013-02-26 17:15:39 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add -backupSettingsFile option

For better configurability of WinBoard, an option is added to optionally
read a settings file, namely if no other settings file (other than the
master file) was read so far. This can be used to import settings from
a previous version, without destroying its settings file.

2013-02-26 16:55:20 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Repair WinBoard compile error

__GITVERSION was not defined in WinBoard, as it was only defined with a
compiler flag by ./configure, which WinBoard does not use.

2013-02-26 16:47:58 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Work-around for Xt selection bug

In some Xt implementations calling extend-end with 0 params causes a crash,
although the mmanual specifies this should default to a call with 'PRIMARY'.
The default selection is now explicitly mentioned in all extend-end calls.

2013-02-24 09:31:57 -0800 Christoph Moench-Tegeder <cmt@burggraben.net>:
fix bug #38401: xboard.texi doesn't build with texinfo-5.0 (tiny change)

xboard-4.7.0's documentation is not buildable with texinfo-5.0, it fails with a lot of error messages from makeinfo like
./xboard.texi:1026: @itemx must follow @item
makeinfo seems right, @itemx is not used as described in the texinfo manual.
Attached patch puts each item of the "Options" menu in it's own @subsection and formats the description of each item as a @table as neccessary. This way, xboard builds again and the manual doesn't look too bad.

2013-02-24 09:09:54 -0800 Arun Persaud <arun@nubati.net>:
updated po/pot files


2013-02-24 09:04:48 -0800 Arun Persaud <arun@nubati.net>:
new version number for developer release


2013-02-20 21:28:36 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Try to make life more bearable in Xaw menus

The alignment still sucks, but not as much as before.

2013-02-20 19:26:17 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix -topLevel option

This option was copletely ignored, after the refactorig.

2013-02-20 18:16:00 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix testing for valid window placement Xaw


2013-02-20 12:15:08 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Suppress printing of status line in dual-board mode


2013-02-20 11:56:12 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix sizing of slave board GTK

For lack of a known method to specify a default size for Graph widgets,
we have to use the same system as for the main board: create with 0x0 size,
and measure the size of the outer window to determine margins.
After every popup we then resize to the required board size plus these
margins. (Awful...)

2013-02-20 10:08:09 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Better handling of undefined window parameters

Repositioning and resizing a newly created dialog was based on the width
parameter of the stored sizing info being valid. Now it is separately
tested if the position is valid, so that sizing and moving can be done
independently. This allows us to store position info, but not size info
for the slave board, so that it automatically will use the size of the
main board when first opened.

2013-02-19 19:14:49 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Limit debug print to debug mode


2013-02-19 19:14:00 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix repositioning of GTK windows

After 'hide' and 'show' GTK would not position the window in the same place,
but center it on the parent. Now we reposition the window after 'show' based
on the WindowPlacement data (if available).

2013-02-19 18:52:41 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Remember window params of slave board

The slave board now has been assigned a WindowPlacement, which is also
saved in the settings file, through new options -slaveX/Y/W/H.

2013-02-19 17:48:39 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix repairing of arrow damage

The '2' bit in damage[][] to indicate the arrow was erroneously cleared,
rather than kept.

2013-02-18 19:55:20 -0800 Arun Persaud <arun@nubati.net>:
fix OK-response in gtk dialogs, see c7f8df124

the fix in c7f8df124 needs to be applied to both of those if cases

2013-02-18 17:11:32 -0800 Arun Persaud <arun@nubati.net>:
removed two more translations from debug output


2013-02-18 17:09:30 -0800 Arun Persaud <arun@nubati.net>:
fixed segfault of g_markup_printf_escaped which needs utf-8 strings

defined text-domain as utf-8 solved this

2013-02-18 13:38:51 -0800 Arun Persaud <arun@nubati.net>:
fixed gtk-warning


2013-02-18 10:06:36 -0800 Arun Persaud <arun@nubati.net>:
removed some translation calls for messages in the debug log


2013-02-18 09:46:53 -0800 Arun Persaud <arun@nubati.net>:
add keyboard shortcuts back into Xaw version


2013-02-17 14:13:30 -0800 Arun Persaud <arun@nubati.net>:
in debug mode also print the git-version if available during build


2013-02-18 21:19:05 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Print game-list timing messages only in debug mode


2013-02-18 21:01:35 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix title of top-level windows


2013-02-18 18:51:44 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Cure GTK warning in top-level windows

There is no OK button in top-level windows, so the 'response' signal is
not defined, and should not be connected.

2013-02-18 18:33:34 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix graying of Revert menu item

The new code was checkmarking it in stead of enabling it!

2013-02-18 11:38:29 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add -onlyOwnGames option

This suppresses auto-saving of ICS observed games.

2013-02-18 11:17:16 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Allow display of 50-move counter in zippy mode

No chance it could be considered cheating there.

2013-02-18 10:48:53 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add new vertical pixel fudge

With the new GTK window type the board is 13 pixels too small.

2013-02-18 10:29:17 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Don't add PV moves on board clicking in AnalyzeMode

Playing PV moves by right-clicking the PV was only intended for clicks
in the engine Output window, selecting from multi-PV analysis output,
where by default it plays only the first move. PV walkig by right-clicking
the board, however, would start the walk at the end even in AnalyzeMode,
and thus add the entire PV. Which is probably not useful. So adding moves
in that case has now completely been suppressed, so it can be used for
harmless viewing of the latest PV.
 The behavior has also been made subject to a persistent Boolean option
-appendPV, configured to true, with a checkbox in General Options.
 A checkbox has also been added there for controlling the -variations option,
which is now also configured to true.

2013-02-18 10:25:40 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Remove checkbox for 'Move Sound'

This function is better controlled from the Sounds dialog.

2013-02-18 09:59:48 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Make switching between board windows absolute

Selecting the window in which to draw the board when the slave board for
-dualBoard was up was done by a toggle. For unknown reason this could cause
a situation where the drawing took place in the other window than the
intended one. The code is made more robust by explicitly specifying which
window to draw in, based on the partnerUp flag.

2013-02-18 09:50:48 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Warn about experimental nature of dual board

The message field, which for now is unused, now contains a warning message.

2013-02-17 10:49:21 -0800 Arun Persaud <arun@nubati.net>:
add close buttons to gtk windows

use windows instead of dialogs for toplevel elements

2013-02-16 19:02:52 -0800 Arun Persaud <arun@nubati.net>:
keyboard accelerators for both front ends.

moved them out of the string definition, so that gtk can add them automatically.
Also wrote a function that loads them for X11.

TODO: add them again when creating X11 menu names

2013-02-15 11:03:02 -0800 Arun Persaud <arun@nubati.net>:
added rotated shogi pieces for -flipback option and moved them to the themes directory


2013-02-14 17:27:27 -0800 Arun Persaud <arun@nubati.net>:
add test for pkg-config

exit configure in case we can't find it.

2013-02-14 13:02:08 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Only perform e.p. capture if there are rights

The heuristic for e.p. capture was such that any non-forward move with
a Pawn to an empty square would delete the piece behind its new location
(Xiangqi exempted). This is now limited to cases where EP_STATUS indicates
that the deleted piece was a Pawn that performed a double-push on the
previous move. (Or if there is unknown EP_STATUS.)
 This makes XBoard more generally suitable for use with variants that have
non-conventional Pawn moves (with legality testing off). To guarantee this
will not have any unexpected effects in normal Chess, this patch will only
be active if legality testing is off.

2013-02-11 16:25:09 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Unreserve tourney game on exit during engine load

While waiting for the feature timeout we are not yet in TwoMachines mode,
so quitting XBoard at that stage would not unreserve the upcoming game.
A global flag 'startingEngine' similar to 'waitingForGame' now signals
this condition.

2013-02-10 13:40:10 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Better fix of feature timeout

The previous fix of the 2nd engine's done=1 aborting the 1st engine's
feature timeout by starting the game caused problems with v1 engines,
or other engines that would never send done=1. It is now fixed by
explicitly testing for a pending feature timeout after resurrecting
the 1st engine, rather than having TwoMachinesEventIfReady wait for
initDone to be set by reception of done=1.
 Also refrain from freezing the UI between match games, as in TwoMachines
mode the UI is mostly disabled anyway.

2013-02-08 16:24:23 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Connect Ctrl key in WinBoard

The Control key is now checked during move entry, for excluding moves
during analysis, or for copying pieces when editing a positions. But the
WB front-end did not read it out yet!

2013-02-08 16:23:58 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix Makefile for install of Xiangqi pieces


2013-02-08 15:35:44 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add Xiangqi piece images to project

Let the images install in themes/xiangqi.

2013-02-08 15:14:43 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Revive -flipBlack option

The -flipBlack option in the cairo version is now implemented the same
way as it was in XBoard, swapping white and black pieces in filpView mode.
This requires a set black piece images that are the upside-down versions
of the white ones.

2013-02-07 13:55:59 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Remove empty-square SVG images from project


2013-02-06 19:00:35 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add desktop stuff for .xop MIME type.

Icons and desktop file were missing, as was the automake code to package them.

2013-02-06 16:48:33 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Update NEWS file


2013-02-06 14:48:27 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add boolean -autoBox option

This option controls the automatic popup of the ICS Input Box on typing
a printable character, which was annoing to some. The option can be
set from the ICS Options dialog.

2013-02-06 10:56:20 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Draw both coords in a1

The lower-left board square needed two inscriptions, but DrawOneSquare()
only allowed one, so that the file ID would not be written. Now two strings
can be passed, and when align > 1 to indicate the alignment of the top
string, it is still checked if a bottom string is present as well, which
then is drawn with alignment 1 (= bottom right).
 The rank coord or piece count can now also be double digit.

2013-02-05 23:03:42 -0800 Arun Persaud <arun@nubati.net>:
updated po files for new release (make distcheck)

2013-02-05 23:02:07 -0800 Arun Persaud <arun@nubati.net>:
updated Changelog, NEWS, etc.

2013-02-05 22:52:42 -0800 Arun Persaud <arun@nubati.net>:
new version number for release of 4.7.0

2013-02-05 11:56:10 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Let initial setting of Twice checkbox reflect current state

The Twice checkbox in the Tournament dialog of WinBoard was by default
starting unticked. It is more convenient to have it preserve the current
state, however, based on the -lpi or -lgi value. Like the auto-incremet
checkbox already did.

2013-02-04 22:07:05 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Put GTK warning in about-box


2013-02-04 21:59:05 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix warnings of build server


2013-02-02 15:58:04 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix GTK error auto-raising board

The variable mainwindow is not set in the gtk2 version. Unfortunately
it is still used to determine default board size...

2013-02-01 19:42:12 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Update Dutch WB translation


2013-02-01 19:37:06 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add some new strings to WB translation template


2013-02-01 19:34:32 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Assign shortcut char to WB menu item

The new Save games as Book item did not have a shortcut yet.

2013-02-01 19:31:18 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Let WinBoard start in its installation folder

The current directory in WinBoard was not automatically the same as the
installDir returned by SearchPath. Especially not if used without shortcut
by file association.

2013-02-01 19:29:26 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix warning in WinBoard


2013-01-30 14:25:57 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix expose of to-square with grid off

The code relied on drawHighlight to request an expose evet for the entire
square, but when lineGap = 0, drawHighlight is a no-op.

2013-01-29 08:39:30 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix library order

On OS X dialog windows could not get focus if -lXaw appeared after
-lX11 -lXt in the linker command. This was cured by moving
@FRONTEND_FLAGS@ (which expands to -lXaw) first in the LDADD parameter
of the automake file.

2013-01-28 17:38:26 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Preserve PGN tags when loading engine

Switching to EditGame mode would normally erase the tags, but when analyzing
a loaded game this would be undesirable. When the engine is playing, keeping
the tags is dubious, but it is not obvious what to put in a new name tag,
so we leave that for now.

2013-01-28 17:23:47 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Preserve mode on engine loading (sometimes)

That the mode always switched to EditGame after engine loading was
particularly annoying at the start of a game (where leaving BeginningOfGame
mode would disablethe Time Control dialog, and the engine would not
automatically reply if you start playing) and AnalyzeMode. These two
modes are now preserved, BeginningOfGame by refraining from calling
EditGameEvent() and sending 'force' to the new engine, and AnalyzeMode
be calling AnalyzeModeEvent() after the engine is loaded and repositioned.

2013-01-28 14:14:19 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Catch unknown engine in tourney games

When an XBoard instance is playing on a tourney it can happen that it is
using a settings file with an engine list not yet containing a certain
participant, because that participant was just installed in the instance
that created the tourney file, but it did not save those modifications yet.
This error is now caught, and leads to an error popup and stopping of the
tourney play. This error was made non-fatal to allow the user to decide
if he wants to quit with or without saving the settings (which, after all,
does not seem up-to-date).

2013-01-26 18:08:37 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Adjudicate pawn-drop mate as loss in Shogi

Mating with a Pawn drop is illegal in Shogi, but rather than burdoning
the legality-testing by this complex rule, we simply invert the result
for such a mate. After all, Shogi customs are such that everything illegal
leads to an immediate loss.

2013-01-25 10:05:44 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Adjudicate perpetual checks as loss also in Shogi

The Xiangqi code for recognizing perpetual checks is now also activated for
Shogi, but in this case we bail out before checking for perpetual chases.
The mini-Shogi case is then recognized as a win for gote.

2013-01-25 09:55:56 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Stalemate is a win in Shogi

In regular Shogi this never occurs, but in the mini variants it happens
often enough.

2013-01-22 20:49:34 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix engine timeout problem in match mode

Reload of a reuse=0 engine would cause a done=1 event which would be
abort the feature timeout of the other engine in the first game.

2013-01-22 20:32:33 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Enforce -popupMoveErrors

One move error was using DisplayError(), in stead of DisplayMoveError(),
and thus did not obey the -popupMoveErrors option.

2013-01-21 15:30:29 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Update texi file


2013-01-21 10:26:20 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix some warnings


2013-01-20 19:09:05 -0800 Arun Persaud <arun@nubati.net>:
Updated German translation


2013-01-19 17:54:59 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Do delayed board draw also with -stickyWindows false

Only CoDrag needs to be suppressed on a configure event with
-stickyWindows false, not the board redraw.

2013-01-18 11:11:48 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Let tellothers command add comment to PGN in local mode


2012-12-30 15:19:57 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Allow use of ~ in pieceToChar for shadow pieces in any variant

The use of shadow-pieces for promoted Pawns was limited to Crazyhouse and
Bughouse. But this is not necessary, and in fact undesirable, as it made
it impossible to implement Crazyhouse960, which combines shadow promotions
with Fischer castling (only available in variant fischerandom)s, and thus
has to be played as 8x8+5_fischerandom, using the Crazyhouse -pieceToChaTable.

2012-12-30 15:06:33 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Resend engine-defined options after reuse=0 reload

When the engine process is re-spawned for a new game under -xreuse setting,
any alteration of engine settings brought about through the Engine Settings
dialogs would be lost, and revert to the engine's default. This is now fixed
by re-sending all options (known from the previous load of the same engine)
with the current settings immediately after "protover 2", and ignoring the
engine's option features. Rather than clearing the option list, and
redefining it from the option features.

2012-12-30 14:58:16 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Implement aborting of games on engine request.

A new game-end command is added. For bakward compatibility it is disguised
as an old one, namely "1/2-1/2 {Engine Abort Request}". These resultDetails
are recognized as special, and rather than ending the game in a draw (which
GUIs not implementing this), the game will be completely 'expunged' from
the match or tourney it is part of, leaving no trace in the saveGameFile.

2012-12-30 14:54:43 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Store some more tourney params in tourney file

The hash and threads settings, as well as ponder and book options are now
also saved in the tourney file, so that tournaments can be played in
parallel where these parameters differ, and no mess up can occur on
restarting a touney after doing something else.

2012-12-23 11:26:31 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Do not automatically save aborted games in tourney PGN

When the game result is 'unfinished', we prompt for a filename rather
then using the specified saveGameFile if a tourney is in progress.

2013-01-19 08:35:52 -0800 Arun Persaud <arun@nubati.net>:
fix configure bug that showed up on OS X (couldn't find X11/Dialog.h)

X_CFLAGS (set by AC_PATH_XTRA) was not used in Makefile.am. It was used in tests for X11/Dialog.h, etc. so configure worked, but make didn't.

2013-01-18 22:26:13 -0800 Arun Persaud <arun@nubati.net>:
Updated Ukrainian translations


2013-01-09 07:57:43 -0800 Arun Persaud <arun@nubati.net>:
removed trailing whitespace


2013-01-09 07:49:20 -0800 Arun Persaud <arun@nubati.net>:
Updated copyright notice to 2013


2012-12-22 22:46:21 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Do not move to forwadMostMove when unpausing AnalyzeMode


2012-12-22 22:44:50 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Use Ctrl key in AnalyzeMode to exclude entered move


2012-12-22 20:40:26 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Implement insufficient mating material for Knightmate

KBK is draw with Royal Knight. With exo-pieces also KNK and KFK.

2012-12-22 20:25:51 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix rep-draw detection in drop games

In games like Crazyhouse every move is reversible, as captured pieces
can eventually change sides again, and be dropped back on the board.
So we have to search back a full 100 ply for repetitions, passing captures
and Pawn moves.

2012-12-11 22:33:07 -0800 Arun Persaud <arun@nubati.net>:
updated po/pot files


2012-12-11 22:18:05 -0800 Arun Persaud <arun@nubati.net>:
new version number for developer release


2012-12-10 14:48:05 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Prefer pause mode on pondering engine over 'easy'

If the thinking engine is stopped after its move because it does not
support pause, is is still better to use pause on the pondering engine
than to switch of its pondering with 'easy', as the latter might not
be instantaneous. Also switch off pondering in the thinking engine before
it gives the move, just in case it does not respond immediately to
easy / hard.

2012-12-10 13:40:47 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Update WinBoard translation template


2012-12-10 13:35:41 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix min-Shogi promotion zone

In integer arithmetic N*2/3 is not the same as N - N/3, with as a result
that the white promotion zone was 2 ranks deep on 5x5 boards. This only
happened in LegalityTest, but because this would classify all moves to
4th rank an non-promotions, they would be printed with a faulty
deferral (=) sign as promochar in SAN.

2012-12-10 13:32:49 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Do not clear PGN tags on Analyze File


2012-12-10 13:28:37 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix -fSAN in AnalyzeFile mode

The exception that a PV relates to currentMove rather than forwardMostMove
applies not oly in AnalyzeMode, but also in AnalyzeFile. Furthermore,
PvToSan did not properly restore the pushed game tail in this case.

2012-12-08 16:02:09 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Improve scaling of border bitmap (WB)

The StretchBlt mode is set from dither to delete.

2012-12-05 16:34:58 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Implement use of pause / resume protocol commands

Engines that implement feature pause=1 will be paused during their move
on a pause event; pondering engines will be controlled either through
'pause' or (if they don't support that) 'easy', which both should work
instantaneously. If the thinking engine does not support 'pause', the pausing
will be deferred to when it produces its move, and is implemented by not
relaying the move, and switch off pondering for both.

2012-12-03 12:58:36 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix Loop-Chess promotions

Loop Chess is not a supported variant, but hould be playable as Crazyhouse
with an altered pieceToCharTable, not defining the shadow promotion pieces.
But with legality testing on promotion to shadow piece would take place
even if the shadow piece was not defined.

2012-12-02 22:44:55 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix move highlighting with animation off

In WB the moving of SetHighlights to after the animation had broken the
highlighting of engine moves, because it was also moved after DrawPosition.
While it is DrawPosition that actually draws the highlights. Now it has been
moved to just before DrawPosition.

2012-12-02 11:09:14 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Start browsing in currently-selected folder (WB)

The folder-browse dialog always started at My Computer, while file browsing
started in the globally last-used folder (which might not be related at all
to the file entry we are now browsing for).

2012-11-29 23:38:18 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Implement book-creation functions

A new menu item triggers conversion of the currently loaded PGN file
to a Polyglot book, saved with the GUI-book filename. A (volatile) option
-mcBookMode alters the probing algorithm to build a book from scratch by
playing games (and using a form of learning).

2012-11-29 22:23:44 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Use colors in Board-Options dialog also for font pieces (WB)

The piece color settings now control the pieceColor or fontPieceColor
settings, depending on the setting of -usePieceFont.

2012-11-29 22:14:05 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Implement auto-creation of ICS logon file

Under control of the new option -autoCreateLogon (which can be set from
the ICS Options dialog) the first two lines in response to the ICS "login:"
pompt will be saved on a newly created logon file (if logon from such
an existing file failed).

2012-11-29 21:56:02 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add Themes dialog (WB)

This dialog, done by the generic popup, contains controls for many settings
that so far could only be controlled by options. It also introduces the
concept of 'themes', sets of options stored in the multi-line option
-themeNames.
  Add some dummy routines to suppress compile errors in functions that
cannot be called yet in XBoard.

2012-11-29 21:32:03 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Allow external piece bitmaps and board border (WB)

The -pieceImageDirectory is now also used for WinBoard, to indicate
a directory of .bmp files, used to replace the built-ins. The outline
and solid bitmaps can be full-color. This allows winBoard to use the
WinBoard-zeta graphics. A new option -border can indicate a bitmap file
used to draw a half-square-wide border around the board.
  A white background has to be drawn underneath the black pieces, to avoid
color-interference with the board, just like happens for the white pieces.
These backgrounds now always use bright white, rather than -withePieceColor.

We still have to work on the scaling of the zeta (600x600) bitmaps to
the actual board size, as for some bitmaps the stratchblt is ugly.

2012-11-29 21:05:15 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Add 'Continue Later' button in Tournament dialog (WB)

This button makes it possible to accept the values specified in the
dialog without immediately starting the tournament (as OK does),
so you can re-open the dialog later, and OK it then.

2012-11-29 20:25:35 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Load opponent logo based on handle in ICS play (WB)

The logos for handles on FICS are sought in folder logos\freeches.org\*.bmp,
etc. If no logo for the handle is found there, the logo of the ICS is
displayed as usual.

2012-11-19 20:35:05 -0800 Daniel Dugovic <dandydand@gmail.com>:
Fix configure script for --enable-zippy  (tiny change)

reported and patch by Daniel.

edited by Arun Persaud:
- also fixed #ifndef ZIPPY -> #if !ZIPPY
- didn't change #if -> #ifdef

2012-11-10 00:12:34 -0800 Arun Persaud <arun@nubati.net>:
don't define X_LIBS when using gtk-frontend


2012-11-07 22:16:09 -0800 Arun Persaud <arun@nubati.net>:
updated po/pot files; added new frontend files


2012-11-07 22:14:20 -0800 Arun Persaud <arun@nubati.net>:
fix configure script for --with-Xaw and --with-gtk

made Xaw the default and included xaw or gtk sources as extras
in case the other frontend is selected for make distcheck

2012-11-07 21:13:38 -0800 Arun Persaud <arun@nubati.net>:
new version number for developer release


2012-10-31 09:55:02 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Draw frames around memos and listboxes


2012-10-30 17:39:59 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Adapt lineGap during sizing


2012-10-30 15:04:32 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix clock highlighting


2012-10-30 10:59:45 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix Chat Box

The genric TypeInProc is re-instated, and replaces ICSKeyEvent as primary
handler for special keys in one-liner text edits.

2012-10-29 19:46:27 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Pay attention to NO_CANCEL dialog flag


2012-10-29 14:18:33 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix clock clicking with GtkEventBox

Like in WinBoard the button-3 case is now handled in the primary
event handler (and communicated to it in a kludgy way). It would be
better if this was absorbed in ClockClick().

2012-10-28 22:25:22 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix logo placement

Graphs were not recognized as possible SAME_ROW elements. A Graph requesting
FIXED_H (which only the logos did) is now packed into a (borderless) aspect
frame.

2012-10-25 23:04:42 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add window positioning

Sticky windows now sort of works, but needs unexplained fudge factors.
Reopening a persistent dialog centers it on the board, and disobeys
the positioning command. The WindowPlacement of the BoardWindow was
again added (it was commented out during development).

2012-10-22 22:41:14 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Make board sizing work through subtracting fixed height

There is now no attempt to give a size to the board drawing area, so it
can later be sized as small as we want. The frame width is determined
by subtracting the nominal board size from the actual window size,
(because the clocks will have that width, and the board will startas 1x1),
while the vertical margin is determined as the difference between the
bactual window height, the drawing-area height minus a fudge of 25 pixels
to compesate for the erroneously allocated blank space below the board.

2012-10-22 21:09:51 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix warnings

The highTextStart/End are now declared in engineoutput.h, shared between
nengineoutput.c and xengineoutput.c. Unused variables were removed.
Also fix warnings in xboard.c and xoptions.c.

2012-10-22 17:40:08 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Delete emptied front-end files, and move rest to gtk directory

Remove #ifdeffed-out Xt code that has GTK implementation. The xgamelist
and xhistory have been completely cleared off content, and were removed.

2012-10-27 00:12:42 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add copy-paste

The code from gtk-xt could be used with almost no modification.

2012-10-21 09:09:01 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Some experimenting with sizing

Does not work as intended yet.

2012-10-21 23:04:49 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add task-bar icon

Funny enough it was necessary to swap the two svg icons to make this work!

2012-10-20 10:29:37 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Make some tall dialogs multi-column

A dummy button was added to the last row of New Variant to make the
number even. A legality-test warning as put in the New Variant dialog
Fix width of tourney-file in Match dialog

2012-10-20 21:14:33 -0700 Arun Persaud <arun@nubati.net>:
updated some icons to SVG


2012-10-20 19:57:58 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add displaying of icons

A new Option type, 'Icon', was added for this as in GTK Icons and Labels
need different types of widget. The icons are now svg, so we could
remove the icon pixmaps.

2012-10-20 14:25:41 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix button bar


2012-10-20 14:03:48 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add BarBegin, BarEnd options

This makes the button bar display correctly. Also fix v-sizing of menu bar.

2012-10-20 10:01:35 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add listbox double-click callback

This function was formerly performed by WheelProc.

2012-10-19 17:29:40 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add hiding / showing second Engine Output pane

The handle of the GTK table for the second engine is stored in the
EndMark handle by GenericPopUp, so it can be used later to hise/show
that table.

2012-10-19 17:15:58 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Use different tables for different dialog columns

Making the same table wider (in steps of 3) to get multiple columns
does not work, because it forces options to align vertically. This
wrecks dialogs like Load Engine, where one column contains a very
high option (listbox). Now differnt tables are created for each column,
packed in a new itermediate hbox. In case of the Engine Output window,
a vbox is used instead to cause vertical stacking.
  Only multi-line text-edits and listboxes are now allowed to take up
vertical space changes, while only the middle colum of the table
can take up horizontal size changes. Within packing boxes space is
equally distributed.
  The labels above the particiant-selection widgets had not the same width
as the textview / listbox, leading to misplacement of the second label.
  For Engine Settings the spontaneous breaking into columns did not
involve a SAME_ROW flag, so the 'columns' were stacked vertically.
In addition the columns were too tall for GTK. They are now 20 options max.

2012-10-19 13:46:37 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Deselect first char in Move Type-in and ICS Input Box

After popup the Move Type-In had the typed character in it, but it was
selected, so that typing a second character would erase it again.
Same for the ICS Input Box. A prototype for CursorAtEnd is now also added.

2012-10-19 13:33:50 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Close Move Type-in on Enter

The code is squeezed into the ICSKeyEvent callback.

2012-10-18 21:57:50 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add access routines to checkboxes and FocusOnWidget

The <Enter> handler of the filter field needs it to return focus to
the listbox.

2012-10-18 21:42:49 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add game-list callbacks

The callbacks for listbox and filter text are treated by the same callback,
which does both key presses and button double-clicks.
Giving focus back to the listbox after filtering does not yet work!

2012-10-18 16:27:57 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add optional callback to Label Options

For some reason this does not work. The callback isn't called!

2012-10-18 13:40:55 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add type-in event to board for popping up box


2012-10-18 13:12:48 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add Shift detection

The state of Shift and Ctrl was only probed on button clicks, so it
is now read out of the button-event struct in the relevant button handlers.
For backwad compatibility, they store it in a global, so the routine
ShiftKeys can take them from there.

2012-10-18 12:35:20 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove some unneeded low-level X11 code

Things like tabbing between controls, mouse-wheel scolling etc. are all
automatic in GTK, and thuse don't need to be ported.

2012-10-18 10:12:27 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add file browser

The gtk-xt code for browsing in response to menu items rather than dialog
browse buttons) was grafted. It seems desirable to merge the two browsing
functions, though. Pass extra params used by Xaw to FileNamePopUpGTK.
The code was altered to use DelayedLoad.

2012-10-18 09:54:41 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix ListBox, and add some support routines

The srolled-window for the listbox used a uninitialized variable for its
width, causing a crash. Now it pays attention to the Option.max and .value
size hints. The access routines were copied from gtk-xt. HighlightWithScroll
could just a wrapper for HighlightListBoxItem, as GTK auto-scrolls.

2012-10-17 20:01:37 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Connect CommentClick handler

This can make use of generic memo callback, so it could be entirely
backend. Therefore it was moved to dialogs.h. As it shared a global with
some other back-end popup routines to remember the current move, these
could now be moved out of xboard.c too. (Well, not so for the X11 front-end,
so it has now been put in a header. :-( )

2012-10-17 19:41:11 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix animation

Because FrameDelay does block the event loop, even though it releaes the
CPU, expose events start to lag behind the actual drawing. By combining
expose events for overlapping squares, rather than decompose them into
3 rectangles, this problem seems to go away. But only if there is an
extra exposure on the squares to be drawn also non-overlapping frames
will be properly drawn (?!).

2012-10-17 14:40:59 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Implement highlighting in engine output by through generic method

HighlightMove was made generally available for all text. It is now just a
wrapper for the generic HighlightText, and could be moved to the back-end.
(This requires the initialization to be widget specific, and the flag
indicating it has been done is now stored in the Option.min field.)

2012-10-17 14:33:28 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Let engine-output memos use new generic callback


2012-10-17 13:53:03 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add scrolling of Move History

To have access to the textview, its handle is now stored in the
Option.textValue field.

2012-10-17 13:11:23 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add highlighting in move list

For now it is assumed the normal background will be white (for clearing
the highlight).

2012-10-16 19:44:40 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Make generic memo-event handler, and connect history callback

Large text widgets can now define a user handler in their Option.choice
field. If they do, agneric callback is added tothem, which catches mouse
events. Button release and pointer motion pass their coordinate and type
to the specified user routine. Button presses in addition retrieve
the text from the widget, set the cursor in it to the clicked point,
and convert it to a character position, also passed to the user.
  The user is responsible for ignoring events he is not interested in.

2012-10-16 15:58:58 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add key-handler for ICS Input Box

In contrast to the X11 code, all intercepted keys are now treated by
the same callback. The ICS Input Box callback triggers on Up, Down and
Return key, and refers to these symbolically.

2012-10-16 15:05:45 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Highlight Pause button


2012-10-16 13:52:41 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Better cleansing of xboard.c from X11 types

The #includes for X11 files were still in there.

2012-10-16 13:28:32 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add text insertion in engine-output memos


2012-10-16 09:34:02 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Append recent engines to engine menu

This is now entirely moved to the back-end, by doing it in the table
from which the menus are later generated. The special callback for engine
items is abandoned, and the euProc for such items is left NULL to flag
that RecentEngineEvent should be invoked with the applicable engine number.
To calculate the latter a global variable firstEngineItem is dynamically
set to the length of the regular engine menu.

2012-10-15 22:17:50 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Transfer more gtk-xt code, and add some new

Add input-source handling
Graft color picker (<Enter> callbeack in the text field is still missing!)
Handle contents of label widgets (store handles in Option for access)
Add printing of clocks
Add printing of window title
Pay attention to BORDER attribute of Label Options (pack in frame)
Delete ReadBitmap code (also remove #including of bitmap files for icons)
Fix layout of textviews (adapt to newer Option format, incl. wrap, scroll)
Support vertical stacking in Break Option (no table widening when !SAME_ROW)
Make default horizontal scroll policy automatic (for engine-output window)
Fix disabling of unused spinners in TC dialog

2012-10-15 20:47:33 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Connect dialog Browse buttons to GTK browser


2012-10-15 19:52:17 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Implement menu checkmarking and enabling

odes in the (later over-written) handle field specify which items
are to be checkmarked, and by what style button.

2012-10-15 16:18:12 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Attach expose handler and connect to mouse events

The GraphExposeProc is conected to the Graph widgets. A gdk draw routine
is used to copy the buffer bitmap to the display. It is also connected
to button and motion-notify events.
  In this version the board can be fully operated with the mouse.

2012-10-14 22:00:14 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Transfer most available gtk-xt code to xoptions.c

GenericPopUp and PopDown are grafted from gtk-xt into gtk2.
The version compiles, but segfaults before doing anything, with error
messages about unexpected cases in GenericPopup for Graph and PopUp Options
of the main board. The spin options need special treatment in GetWidgtText.

2012-10-15 11:41:33 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Give LoadListBox two extra parameters

It often occurs that we just want to swap two elements, and reloading
the entire listbox would be a very inefficient way to do this in widget
sets that allow access to a single element.

2012-10-14 20:36:34 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove all X11 code by #ifdeffing it out

With #ifdef TODO_GTK we disable all X11 code, in order to get a compiling
X11-free version. The disabled sections can then be replaced by code
for another widget set, the X11 version still showing what it was supposed
to do.

2012-10-14 17:37:18 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Move timer functions to new file xtimer.c

The timers for the clocks, periodic analysis events and ScheduleDelayedEvent
go into a separate file.

2012-10-14 15:31:56 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Make xevalgraph.c backend

The only front-end stuff left in xevalgraph.c was an unnecessary reading
of the window size (which could be obtained from the WindowPlaceent struct),
and drawing (all cairo now). The drawing was moved to draw.c, and the
remaining part renamed nevalgraph.c.

An attempt to list them in a less chaotic way.

2012-10-14 14:34:09 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Cleanse back-end code of all references to X11 types

When compiling for X11, common.h #included X11/Intrinsics.h so that
X11-specific data types could still be used throughout the back-end part
not shared with WinBoard. By temporarily disabling this, error messages
could be generated on all such occurrences, and were then solved. This
included the following:
  The InputSource struct is not uses in usystem.c at all, and its definition
could be moved to xboard.c. InputSourceRef, which logically should heve been
defined as (InputSource *), is in reality (void *), so the back-end can
pass it around.
  Some non-platform-dependent stuff of general interest had prototypes
(or external vars) in xboard.h. This is now moved to a new header xboard2.h,
to allow it to be shared without bringing in X11 dependence.
  The EngineOutputProc was defined inconsistently, but the header was not
sufficiently shared to flag this error. This has been repaired now, by moving
the template to menus.h.

2012-10-14 13:39:47 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove xedittags.c, .h from project

The tiny bit of code that was left was absorbed in dialogs.c, the one
prototype to menus.h.

2012-10-14 12:53:20 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove inclusion of frontend.h from backendz.h

This did not seem to serve any purpose at all.

2012-10-14 12:49:28 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Split xhistory.c in front-end and middle-end part


2012-10-14 12:21:09 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Move MarkMenuItem to xoptions.c

This to make xoptions.c completely independent from the inclusion of
frontend.h. The prototype is moved to menus.h.
  Loose end: DrawPosition in board.c does not see the prototype.
Very suspect that it wants to see it. Why would it have to correct the
flipView setting??? Better se to it that it is correct to start with!

2012-10-14 12:14:07 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
move testing for no options to back-end

The test if an engine has options and display a note instead of
the Engine Settings dialog when this is not the case is moved
to outside of GenericPopUp, to make the later independent of calling
other routines.

2012-10-14 11:20:16 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove some unnecessary header includes


2012-10-14 10:47:21 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Move ICS-engine analyze and AnalyzeGame code to shared back-end

This code was still duplicated between WB and XB in the front-ends.
Because of its incorporation in AnalyzeModeEvent the 'middle-end'
function AnalyzeModeProc could be eliminated. AnalyzeModeEvent was made
to return an int to flag its success, and keep the JAWS code based
on it out of XBoard.
 The AnalyzeGame code  is now incorporated in AnalyzeFileEvent().

2012-10-26 10:28:59 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add configure switches for Xaw vs GTK.

This patch is an adapted version of the one Arun made at the end of the
gtk2 branch. It uses the exact same version of his configure.ac, but the
Makefile.am is adapted to the situation in master at the beginning of it,
after duplication of the X11 front-end files in the xaw directory for
preserving them, while the versions that will be modified to GTK front-end
are still in the xboard directory.

2012-10-25 11:12:41 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Prepare xoptions.c for middle-end changes

Some changes that can be pre-empted, such as the splitting of Label
into Label and Icon Option type, and some dummy wrappers are already
implemented, so they can be used when needed by GTK.

2012-10-25 09:20:03 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Preserve copies of the X11 front-end in xboard directory

The X11 front-end has been moved to the xaw directory, but copies
are retained for modifying them to gtk versions.

2012-10-24 08:55:27 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Move X11 front-end to directory xaw


2012-10-31 18:18:02 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix initial enables in TC dialog


2012-10-30 14:47:24 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix image extension used for browsing to .pgn


2012-10-28 21:39:19 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix WinBoard compile error on enum PEN


2012-10-30 11:20:54 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix crash in promotion popup

There were not enough entries for Capablanca Chess with legality-testing off,
where King is also presented as a choice.

2012-10-28 18:19:23 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix message in New Variant dialog

Even in -ncp mode the message was referring to the 'current engine'.

2012-10-28 18:16:20 +0100 H.G. Muller <h.g.muller@hccnet.nl>:
Fix crash on animation after resizing

The animate buffers were still holding pointers to a discarded cairo surface.

2012-10-21 14:49:45 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix variant-dependent pieces

The code to select the actually used piece images from the available ones
(from InitDrawingSizes) is now made into a subroutine, and also called
after resizing the pieces.

2012-10-26 10:18:48 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Replace marble texture pixmaps by png


2012-10-26 10:13:27 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Replace xiangqi board pixmaps by png images


2012-10-26 10:08:44 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove the texture pixmaps from project

These were already replaced by png textures.

2012-10-23 15:02:56 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix re-rendering of svg on resize

The previously prepared bitmaps were retained if we resized, rather then
rendering the svg atthe new size.

2012-10-21 15:48:30 -0700 Arun Persaud <arun@nubati.net>:
update po/pot files


2012-10-21 15:47:12 -0700 Arun Persaud <arun@nubati.net>:
new version number for developer release


2012-10-21 11:10:36 -0700 Arun Persaud <arun@nubati.net>:
we still need a few bitmaps, so the directory needs to be included in Makefile.am

make distcheck complained about this

2012-10-21 15:56:43 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Rename svg shogi pieces, so they become usable

Only white pieces are present. For black they would have to be flipped.

2012-10-21 15:54:25 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix bug in resizing on variant switch

The fix of the garbage pixels had unfortunately broken this.

2012-10-21 15:15:01 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix bug in fallback mechanism

After failure to load a piece from a specified -pieceImageDirectory
the default theme was not tried, because the cairo function to load
from a pgn file returns a valid handle even if the file does not exist.

2012-10-21 14:58:42 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Get svg error message


2012-10-21 14:49:45 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix variant-dependent pieces

The code to select the actually used piece images from the available ones
(from InitDrawingSizes) is now made into a subroutine, and also called
after resizing the pieces.

2012-10-16 12:08:25 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove caveat on available pieces fromNew Variant dialog

In the cairo version all pieces are available at all board sizes.

2012-10-21 11:41:11 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove some unused images from png directory

Currently only the three board texture files are used. But the piece
images are still left, because they might be useful if we ever want
to mae XBoard configurable for not using librsvg.

2012-10-21 10:40:37 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix bug in resizing

The nominal size of the window was not set in the correct Option struct.

2012-10-21 09:58:23 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Prevent odd-width line shift in length direction


2012-10-21 09:48:29 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Print missing-pieces error message to console

The GUI might not yet be up at this point.In addition, suppress
the rest of the code, because it could give segfaults.

2012-10-21 09:30:26 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix garbage pixels on the right of the board

The extra pixel that is requested in sizing the board window back to
an available size because we ofte get one less pixel than we ask for
can lead to a line of garbage pixels next to the board. This is combatted
by limiting expose events to the true board area, remembering that in the
Graph Option rather than the actual window size, and suppressing resizing
when the latter only exceeds the stored width by more than 1 pixel.

2012-10-15 10:27:14 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Trim board-window size

The board window will sometimes be a pixel larger than requested.
The ReSize routine will now reset the size values in the Graph Option
to what they were supposed to be, to prevent expose events will
draw nonsense in the extra space.

2012-10-13 23:26:10 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Adapt docs for svg/png in stead of bitmap/pixmap


2012-10-13 09:35:31 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Spontaeous changes in gettext stuff


2012-10-13 08:03:48 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Move Shogi svg pieces to own directory


2012-10-13 07:57:48 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove unnecessary Xt colors and call to MakeColors


2012-10-13 07:50:56 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Code cleanup: move expose redraw to draw.c


2012-10-13 07:37:45 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Suppress warning for InitDrawingHandle


2012-10-13 07:33:15 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Make fallbackPieceImageDirectory hardcoded


2012-10-13 07:18:23 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Increase drag delay too 200 msec

On my slow laptop it can take more than 100 msec to reallocate bitmaps
in the generic expose, with as a consequence that the DelayedDrag event
is not processed before the timeout expires. The timeout event then triggers
a redraw, which again takes so long that the timeout expires before the
event to reset the timer could be processed, and this just perpetuates
itself, leading to a large series of queued redraws, and a long waiting time
for an image to appear after the sizing stops.

2012-10-13 07:01:05 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix segfault on faulty command-line option


2012-10-12 14:42:26 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Suppress anti-aliasing in -monoMode

In the color-substitution code, always use bi-level alpha and color.

2012-10-12 13:34:04 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove piece pixmaps from project


2012-10-12 11:47:25 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Implement proper fallback cascade

Try in order cached svg, cached bitmap, user png, user svg, default svg.

2012-10-12 11:06:06 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Cache svg handles


2012-10-12 10:45:29 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Install svg pieces in themes/default

The svg pieces are installed with the XBoard data in the directory
.../xboard/themes/default . A new option -fallbackPieceImageDirectory
is added, and is configured in the master settings file to point to
these pieces. This options, which replaces -svgDirectory, is volatile,
and the menu control to set it is removed. The -pngDirectory option
is renamed to -pieceImageDirectory (-pid).

2012-10-12 10:02:50 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove bitmaps from project


2012-10-12 09:55:07 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Install the wood textures as png

The xboard.conf has to be edited to set these as default textures,
in stead of the .xpm files.

2012-10-12 09:50:31 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add hatched board texture

A png file with hatching is added, so it can be used for dark squares
in a monochrome theme. It is installed in a new data directory
.../xboard/themes/textures.

2012-10-11 22:21:06 -0700 Arun Persaud <arun@nubati.net>:
added a black and white theme to replace the mono option


2012-10-11 20:56:42 -0700 Arun Persaud <arun@nubati.net>:
added SVGs to dist files in automake


2012-10-11 19:44:22 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix exposure of atomic captures

The grid lines had to be repaired too, so damage marker 2 had to be used.

2012-10-11 19:13:24 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix animation with textures off

The blank square was not shifted to (0,0) when drawing the animation buffer.

2012-10-11 19:08:49 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix 1-pixel offset of grid lines on some cairo implementations


2012-10-11 09:51:07 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Solve odd lineGap problem

Not all cairo versions seem to round the same when an odd-width line
is to be drawn at integer coordinates. So now we explicitly ask it
to be drawn centered on half-odd-integer coordinates.

2012-10-09 18:52:43 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Add -trueColors option


2012-10-09 14:56:16 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Always render svg pieces anew on size change

The svg pieces are now unconditionally rendered to supply the pngPieceImage
master source otherwise read from png file. This is then fed into the
original png code (so that if for some reason it fails, we fall back on png
pieces, and finally on built-in pixmaps).

2012-10-09 14:06:47 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix rsvg version in configure.ac


2012-10-08 19:54:19 -0700 Arun Persaud <arun@nubati.net>:
initial svg rendering


2012-10-08 18:34:05 -0700 Arun Persaud <arun@nubati.net>:
added cairo and librsvg to configure process


2012-10-08 18:32:55 -0700 Arun Persaud <arun@nubati.net>:
added SVGs

converted from inkscape SVG to plain SVG before adding

2012-10-10 21:19:25 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Adapt Eval Graph code to new drawing system

The eval graph now triggers an expose after drawing, and gets its size
from the Option fields.

2012-10-10 20:37:58 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix expose requests seek graph


2012-10-10 20:07:53 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Make draw handle for board globally available


2012-10-10 19:45:39 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Let expose requests pay proper attenton to widget

So far, expose requests were always assumed to be for the board widget,
preventing display of the labels.

2012-10-10 18:45:59 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix initial display of logos

The logos were selected and printed during the initialization,
before the main event loop was started. So the expose events during
widget creation, which created the drawing surfaces, were only processed
after the initial drawing of logos, so that there was nothing to draw on!

2012-10-10 17:25:25 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix alignment of highlight expose


2012-10-10 17:19:52 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
remove NewSurfaces


2012-10-09 13:57:40 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Make expose handler generic

The Graph Option now returns two handles in hitherto unused fields of its
Option struct: Option.choice is a cairo handle of a memory buffer the user
must draw to, while Option.textValue is the handle to the widget.
Expose events copy the memory buffer to the display for the exposed area.
A routine DrawExpose() is supplied to let the user generate expose events.

2012-10-06 21:30:59 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Reload piece images when pngDirectory is changed

The OK function of the board options dialog now tests for a change
in pngDirectory and passes a parameter to InitDrawingParams to tell
it to destroy the old images. This cause a reload on scaling the pieces.

2012-10-06 20:16:54 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Suppress redraw during sizing


2012-10-06 20:10:18 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix bug in resize trigger

Even if the square size does not change, the window has to be shrunk
back to fit the board.

2012-10-06 19:54:50 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Make Piececolor options work on png pieces

The scaled, antialiased bitmaps are now colored according to the option
settings. This is done by decomposing every pixel in a white, black and
transparency component, which is possible because the opacity is
directly give in the alpha channel, and black originally was a hard zero.
So what you se in the red channel must be white (as the original white
was #FFFFCC), and the rest of the opacity must be due to black.
  The color to be replaced (white for the white pieces, black for the black
pieces) is then replaced by the original weight of this color times the
color from the option.

2012-10-06 17:53:25 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Plug resource leak on rezising with pixmaps


2012-10-06 16:19:41 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fall back on built-in pixmaps if png pieces unreadable


2012-10-06 14:21:11 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Do coordinate text alignment with cairo


2012-10-06 13:40:12 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Some cleanup


2012-10-06 12:14:16 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Clean up drawing code

Now all routines only draw through cairo, the destination surface can
be passed to them, making it possible to do the duplicate drawing
to the memory board image in a cleaner way.

2012-10-06 11:37:20 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Check in draw.c, draw.h


2012-10-06 11:26:24 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove all bitmap & pixmap drawing

All Xt code for drawing the board is now removed, as are the
-pixmapDirectory and -bitmapdirectory option. This version is now
completely dependent on cairo for handling of all graphics anywhere.
(The build-in pixmaps are still left, though, for later use as an
alterative cairo source when no png files are found.)

2012-10-06 10:29:06 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Separate off drawing routines from xboard.c

A new source file draw.c is created to hold all code associated with
board drawing. The routines for loading bitmaps and creating colors
were left in xboard.c, though, as they are also used for handling
icons and colors of widgets.

2012-10-05 19:53:04 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix erasing dots in seek graph

The cairo dots were just a little bit bigger than the old ones, so they
were not properly erased, but left a remnant.

2012-10-05 19:25:58 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Cleanup CairoOverlayPiece


2012-10-05 19:22:34 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Draw arrow also on backup image


2012-10-05 19:16:30 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix highlight clearing


2012-10-05 18:59:30 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove acceleration trick


2012-10-05 18:54:01 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Implement variant-dependent png piece symbols


2012-10-05 18:42:51 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix clearing of markers dots with promo popup


2012-10-05 18:32:24 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Also render coordinates to backup board


2012-10-05 18:30:35 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Remove debug print


2012-10-05 18:29:13 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Let cairo also do evenly colored squares.


2012-10-05 17:56:43 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Make dragged piece for excluding moves transparent


2012-10-05 17:42:18 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Fix DrawSeekText


2012-10-05 17:40:55 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Redo coordinate / piece-count printing ith cairo


2012-10-05 15:40:14 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Also use cairo on slave board

With dual-board observing the drawing should be redirected to the
slave board, so SwitchWindow should also switch the xlib-surface to
one for the slave window.

2012-10-05 15:12:41 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Cure flashing of piece on from-square

In a drag-drop move the piece was flashed on the from square after the
user dropped it on the to-square, because highlights were set before the
move had updated the board. And highlighting with arrow can cause a redraw
of the last board. Same with setting or clearing marker dots.

2012-10-05 14:20:12 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Specify proper condition for using cairo animation


2012-10-05 14:07:53 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Allow resizing of board window

In full png mode the size of the board window is no longer locked.
After resizing, the square size is calculated as the largest one that
would fit the window, and the window is then shrunk around it. To prevent
excessive redrawing during sizing, the redraw is delayed by putting it
in DragProc.

2012-10-05 10:41:02 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Switch to using 64x64 png images

The bitmap scaling works better when the needed factors are not too large,
as the svg rendering's anti-aliasing did not blur more than 1 pixel.
So after shrinking too much the images are not effectively anti-aliased
anymore.

2012-10-05 10:16:47 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Maintain in-memory copy of the board image

As reading back from the xlib-surface proves incredibly slow (cairo probably
converts the entire image from Xt format to cairo format when we set it
as a source, even though we only want to read a few dozen pixels), we now
keep an exact copy of the board image as a cairo bitmap in memory. This
restores the speed of the animation, but unfortunately leads to awful code.

2012-10-04 22:15:05 +0200 H.G. Muller <h.g.muller@hccnet.nl>:
Do animation with cairo

Unfortunately this is incredibly slow.