Tinkering with Winboard

[Thomas Mayer]

Hi Alessandro,

I just want to inform you that all my mentioned problems are gone finally. NOW I really have fun with your winboard_x... Great ! The analysis-problem was the only thing which did prevent me so far to fully switch to your version...

Greets, Thomas

[Vladimir Xern]

Hi, Alessandro, your Winboard X is a thing of beauty. It appears that most of the engine testing kinks have been worked out and that you've returned to development of Kiwi, but are you still open to feature requests for whenever you may get back to Winboard?

Nevertheless, keep up the good work.

[Alessandro Scotti]

Hi, Alessandro, your Winboard X is a thing of beauty. It appears that most of the engine testing kinks have been worked out and that you've returned to development of Kiwi, but are you still open to feature requests for whenever you may get back to Winboard?

Hi Vladimir, thanks a lot!
Development on Winboard has been slowed down a lot but not stopped, and requests or comments are always welcome!
BTW I'm currently working on a new engine, which is still in the very early stage but already able to play. So far, it's no match even for Kiwi though.

[Vladimir Xern]

Thank you for your reply, Alessandro. A few of the features on my mind stem from quirks of Winboard that slowly erode my patience.

First is the excessively cumbersome edit position mode. I come across a middlegame position somewhere that I want to set up and think to myself, "Aw, geez, not again." I have a few ideas in my mind of how it should be, but I digress...

Second is the minor annoyance of pasting a game fragment and not being directly in edit game mode, having to manually switch to it to fool around with the position.

Third is the typing in of moves. I favor keyboard input for most things and find myself typing in moves most of the time. The thing that annoyed me with that at first was how I had to capitalize for Bishop in order to disambiguate when it was completely unnecessary. That got me thinking about how cumbersome the whole process was, such as having to press enter every time. I thought that there should be some "smart typing" capability similar to "smart move" entry. I.e., minimalize typing input by making the move as soon as the input is unambiguous. For example, only typing one "o" for castles, or "6" for Bh6 when that is the only piece that can go to the sixth rank. Of course, this whole smart typing concept is superfluous, and I'd be happy with just the bishop disambiguation.

Fourth is an extension of edit game mode that is similar to Scid's trial mode. For example, I paste a game and start playing through it. I get to move 20 and wonder what would happen after Black plays such-and-such. I hit "t" for trial (or whatever), play a few moves, hit "t" again to zoom back to where I left off on the main game and continue along my merry way. Normally if I made a different move, the whole game would be supplanted by that one move.

Lastly, window controls (minimize, close) on board sizes smaller than "slim."

That's all I can think of for now. If anything else comes to me, I'll be sure to holler.

P.S. Good luck on your current brainchild.

[Marc Lacrosse]

Hi, Alessandro, your Winboard X is a thing of beauty.

Hi Vladimir, thanks a lot!
Development on Winboard has been slowed down a lot but not stopped, and requests or comments are always welcome!

Hi Alessandro,

One feature comes immediately to my mind : the possibility to have one or several engine(s) analysing one or several game(s) that one is observing on FICS or ICC.
Equivalent or improved form of the "icsEngineAnalyze" that has already been included in the CVS version of Winboard at:

http://cvs.savannah.gnu.org/viewcvs/xbo ... cvs-markup

This was cruely missing while observing CCT yesterday...



... with all my thanks to all these wonders that you already added to our cherished brave old WB!

Marc

P.S. an older desired feature (I already told you) : the possibility to save these "first evaluation after leaving the book" as an epd file with only these positions and their corresponding evaluations as opcodes.

[Vladimir Xern]

I think I may have found another visual glitch concerning the arrows. At smaller board sizes, when you make a diagonal move (I think that's about it) the arrow's thickness varies half-way inexplicably. It's somewhat hard to explain, but that's the gist of it.

[Cristian Tala]

Hi Allessandro,

Do You know a format named PGC (Called PGN Compressed format)?. This format is apparently fine and fast.
Is a good feature for some version of Winboard or this is most problematic?.
Whats is your feeling about this?.

reference web page:
http://pgn.freeservers.com/

[Guenther Simon]

Hi Allessandro,

Do You know a format named PGC (Called PGN Compressed format)?. This format is apparently fine and fast.
Is a good feature for some version of Winboard or this is most problematic?.
Whats is your feeling about this?.

reference web page:
http://pgn.freeservers.com/

Hello,

I don't see a need for WB to include compression algorithms for pgn
files. It is just one click to use a compression tool. BTW I bet George
Lyapkos pgn compression is even better than PGC.
http://www.geocities.com/lyapko/lgpgnc.htm
However, as I still don't know more than one person currently
using it, when all other just use ZIP or RAR or transfer to SCID format
or CBH/CBZ it seems a bit anachronistic in times of larger and
larger HDs and faster and faster internet connections.
I guess that are also the reasons no one uses the 5 years old PGC?

Regards,
Guenther

Guenther

[George Lyapko]

There are some features I've always missed in winboard. It would be much easier to perform automatic tournaments and I think some of them are easy to implement:

1. Termination of game if there is mate/stalemate on board (many engines doen't send "1-0","0-1" or 1/2-1/2" and game ends on a win by time). It will save much time in automated tournaments.
2. Optional termination of game on 50-moves rule(the same reason).
3. "Forfeit due to illegal move" doesn't work for illegal castlings.Here is an example:

[Event "Computer Chess Game"]
[Site "LYAPKO"]
[Date "2006.03.24"]
[Round "-"]
[White "lyapkyur"]
[Black "Bestia_090"]
[Result "1-0"]
[TimeControl "60"]

1. g3 d5 2. Bg2 c6 3. Nf3 Bg4 4. Kf1 e6 5. Ke1 Nf6 6. O-O
{White wins on time} 1-0

4.Termination of game if program rejects a legal move.

Thanks in advance,
George

[Alessandro Scotti]

Hi George,
Winboard doesn't know much about castling, and may have somewhat relaxed rules in general. IIRC this was a design choice to accomodate the many variants it supports.
I think it makes sense to add some better handling of "classic" chess, but have no idea how difficult that would be.

[Christos Gitsis]

Hi Alessandro,

Your modified version of WinBoard seems very nice. You mention on your site that some of the changes you made will also work in XBoard. How would I compile such a version of XBoard? I dowloaded the source code but there is no makefile in it.

[Uri Blass]

I think that it may be a good idea if you let the user to add increasement in milliseconds and the level command should also accept time that is not in seconds but in milliseconds

from the uci protocol

winc <x>
white increment per move in mseconds if x > 0
* binc <x>
black increment per move in mseconds if x > 0

One of my problem to implement the uci protocol is that
movei does not know to accept increasement in milliseconds.

My increasement is in seconds and I multiply it by 1000 to get the time in milliseconds.

I plan to change it to tell movei to translate everything it gets to milliseconds and to work only with milliseconds.

The engine also should tell winboard if the engine supports milliseconds and in case that the engine does not support it winboard should tell the user that the engine does not support it.

Uri