Development of WinBoard 4.3.14

[H.G.Muller]

[Note: this is a re-posting of a thread that got lost in a forum crash]

Now that WinBoard 4.3.13 is out, I already started working on WinBoard 4.3.14, and I cannot imagine a more natural place to discuss about future WinBoard developments than this forum.

Some of the improvements I plan to implement (perhaps not all in 4.3.14, but some of them already work!):
* add an autoKibitz option, which kibitzes the last PV info from the engine to the ICS
* allow entering of a ufor a Human player
* add the possibility to use node count or CPU time for engine timing and clock update. (WB protocol will be extended with a new command 'nps N' for this.)
& add a menu item to save the chessboard display on a bitmap file, for making diagrams
* add variant caparandom (the FRC equivalent on 10x8 board)
* as a spin-off: allow shuffling of the opening position of any variant
* add variant Janus (with two Archbishops on 10x8)
* add variant Berolina (which has differently moving Pawns)

If there are any remarks or comments on these features, they can be discussed below. Also if there are other ideas for improvements, please post them in this thread.

------------------------------------------------------

Note that several people have already expressed their wish that the development in WinBoard would be parallelled by a similar one in xboard. In a sense this is automatic: most of the changes I made are in the platform independent code of WinBoard/xboard. For the Linux version there exists already an xboard.c source file that supports the vartrious board geomettries, when linked with the WinBoard_F backend sources. The only thing that is needed is to add the options and menu items to switch on the vaious features. This would require someone capable of compiling and testing under Linux, preferably with experience in X-windows programming. (i.e. he should know how to make pull-down menus, dialog popups, and open new output windows).

[Zach Wegner]

OK, here goes, take 2:

I am interested in working on the Unix port. I don't have much X experience, but I don't think it should be too hard, given the code that is already in place. The problem is that I'm quite swamped with work these days, so it could be a while.

Once the port is done, we should try to get it merged into the official CVS repositories, so all the Unix/Windows stuff can proceed more or less in parallel.

Maybe there should also be an effort to clean up the code WRT machine dependent code. The last time I looked at winboard.c and xboard.c, it was quite scary, several thousand lines. Maybe we try to modularize this a bit, making generic functions... This might be hard, as IIRC the structure of Xboard is integrated into the X menu system...

[H.G.Muller]

Well, I will request the source code from Kong Sian, who is the last person known to have been working on this. I will then combine it with my newest winboard files, and send the entire source tree to you.

The source files are indeed titanic: winboard.c is over 10,000 lines, backend.c nearly 13,000.

[H.G.Muller]

Unfortunately, Kong Sian lost his sources. So we have to start from Holger's sources. He had a running version, though, capable f playing Xiangqi, and I am in possession of his enttire source tree.