[ List Earliest Comments Only For Pages | Games | Rated Pages | Rated Games | Subjects of Discussion ]
Comments/Ratings for a Single Item
ChessV
. Program for playing numerous Chess variants against your PC.[All Comments] [Add Comment or Rating]Yes, welcome back.
📝Greg Strong wrote on Mon, Aug 28, 2006 11:28 PM UTC:Thanks, Sam. ChessV, thankfully, is not dead. It had just been so long since the last release, that I wanted to take the time to make sure that the next release was a major step forward. Yeah, the 0.8x versions had a nasty bug that resulted is the occasional really bad move. That has been fixed, along with a dozen or so other miscellaneous bugs. In addition to supporting the rules for Schoolbook, if you look in the openings directory, you will see a small opening book. Although it only has a few lines of play in it, each move was calculated extremely deeply using several very powerful computers I have at work. A few days worth of CPU time went into it, although, that being said, it is just a start ... (In addition to Schoolbook, a similar effort has gone into making opening books for other 10x8 variants, primarily Grotesque, Ladorean, and Univers.) A note about opening books: The opening books are just text files, and while the format should be pretty self-explanatory, it occurs to me that there is one thing that hasn't been documented and is worth noting. Some moves in the opening books are preceeded with a question-mark. The question-mark denotes an inferior move, and the computer will not make that move. The point of including it is so that if the human makes that move, the computer will know the correct response. Regarding porting to other operating systems: I am very interested in seeing ChessV ported to other platforms, primarily Linux. This has always interested me, but I have very little knowledge about other platforms. Still, I made a few design decisions specifically for the purpose of making such a port easier. For one thing, I did not use Microsoft Foundation Classes, or any other Windows-centric class library. ChessV uses only standard Win32 calls, and I called as few different functions as possible. Sound, animation, and other bells-and-wistles were deliberately left out to make the code more portable. Also, all calls to Windows functions are preceeded with :: (which is unnecessary because these functions are global anyway, but I did it so that if you someone searches for ' ::' it will turn up all calls to Windows functions in one shot.) Finally, I wrote some classes to be a thin wrapper encapsulating some of the Windows graphics objects, like pens and brushes, so that the code would only need to be modified in one place, but I must admit that I didn't do nearly as much of this as I would like. So, yes, I would be happy for any help you could provide to make ChessV more portable. I do not know much (if anything) about MingW32, though... Does it only allow one to port console applications, or does it emulate some of the GUI functions as well? Thanks again for the positive feedback! Greg
Joe Joyce wrote on Tue, Aug 29, 2006 01:51 AM UTC:Excellent! I've enjoyed playing this; that it plays opening pawn moves alone puts it ahead of Zillions. The only part I don't like is that it keeps beating me. Great to see this out. Thanks. Joe ps: love the shatranj section!
hey greg, help me out. i downloaded the whole zip thing. then i click on the exe. file and choose a game... the board comes up. there are no pieces! i tried a few different games...
📝Greg Strong wrote on Tue, Aug 29, 2006 11:59 AM UTC:The files are in folders within the zip file. When you unzipped it, it should preserve the folder names, but it is possible to turn that off. Look at the directory you unzipped it to. There should be a subdirectory called 'images' and then subdirectories under that for the different piece sets. If all your images are out in the main directory then that's your problem...
There should be a subdirectory called 'images' and then subdirectories under that for the different piece sets. Ok, I checked that and that's right. the images are in subds under images, they're not out in the main directory. maybe i still don't get it. is there something else that i can try and fix?
Yes when I unzipped it, the same thing happened to me, all the image files were being replaced, then nothing came up when I launched the program. Here is what to do to fix it. 1. Double click the chessV.zip file. 2. If using Winzip, you have to hit the 'extract' button. 3. This brings up a window asking you where you want the files to go. Make sure you check the 'Use folder names' box on the bottom right! It seems to me this shouldn't even be an option it should always do this. Anyway that will extract ChessV files into all the right places.
As for extratracting ChessV, it's a lot better to extract it using Windows XP's built-in .zip extractor. Basically, since I don't have Winzip, I just double click on the ChessV zip file, which will open up a window with a folder called 'ChessV'. I copy this folder (right click, copy) and paste it in the C:\ folder. This gives me a ChessV folder that I can play ChessV from.
Winzip has a way of destroying all of the subfolders that a .zip file has; this will break ChessV. The solution is simple: Don't use Winzip. WinXP's .zip extractor is simpler, better, and cheaper.
- Sam
📝Greg Strong wrote on Tue, Aug 29, 2006 10:31 PM UTC:Carlos carlos: Hmmm... Your folder structure is appearing, but you still aren't seeing any images. What version of windows are you running? One thing that you could try, is creating a shortcut to ChessV.exe. Then edit that shortcut by right-clicking and selecting 'properties.' Then, in the text box labeled 'start in' make sure it has the proper path to ChessV.exe. Then run it with the shortcut. The idea being that you are trying to make sure that the current directory for the ChessV process is the correct one... It looks for all images (and include files, and opening books, etc) in subdirectories relative to the current directory. Failing that, more radical experimentation will be required... But we'll get it working.
ok! the first suggestion was right after all... it looks cool. i'm gonna try a few games out.
I notice that your documentation accompanying the ChessV 0.9 program is a Microsoft Word file (*.doc). If you, as an author, prefer an Adobe Acrobat file (*.pdf) which preserves the exact graphic representation of the document as you saw it and wrote it (without any possible font substitutions), then I would be glad to send you my Adobe Acrobat 4.0 (Standard) Full on 1 CD which I have recently replaced with Adobe Acrobat 6.0 Standard Full. Most readers (myself included) prefer *.pdf files because they open faster and are very flexible upon display. I think it is probably compatible with whatever version of Microsoft Windows NT you are using. If you issue a mailing address for you to me privately via my E-mail address, then I can send it to you immediately USPS. Rest assured, it is authentic software, still in perfect condition.
Great program, I don't know why I didn't download it sooner.
📝Greg Strong wrote on Thu, Aug 31, 2006 05:13 PM UTC:Derek: Thanks for the offer. Actually, I have Acrobat, I just didn't get around to converting it. I will post it in both formats with the next version (the documentation isn't quite finished anyway...)
Hi, Ed.
Do we know it's the ChessV program that's being implemented in these games? Couldn't that just be someone's handle? I mean, 'ChessV' might not even know about the ChessV program.
If we don't know, it could be proven one way or the other easily enough, I imagine, by going through some of the games and finding out whether the moves accord with that which the ChessV program would make, right?
One feature I really like about ChessV is the 'test this position' feature. This is a veryuseful feature for helping build up an opening book for my Schoolbook variant; I've been spending all of last weeking using ChessV to help me come up with some opening lines in Schoolbook Chess.
I'll post some basic opening theory for Schoolbook later on this week. As a start, I like the way white's position looks after 1. f4 Nd6 2. g4
- Sam
📝Greg Strong wrote on Mon, Sep 4, 2006 04:32 PM UTC:Sam: You may have figured this out already, but in case you didn't (or in case anyone else is interested) ... After you run the 'test this position,' it ranks the moves. Then, if you run it again to a greater depth, but tell it to restrict to the first x moves, it uses the first x moves from the previous test. So, you can, for example, test all moves to a depth of 5, then test the best 12 to a depth of 7, then the best 6 of those to a depth of 9, etc. Glad that you are making use of it; calculating opening books is exactly what I wrote that feature for.
Hi, Can someone explain that opening book testing feature to me? It sounds like a great idea! I want to do an opening book for another variant and I'd love to figure out how to do this. I also don't understand why you'd hafta restrict moves, cann someone explain this to? BTW I am most interested in Janus Chess.
📝Greg Strong wrote on Mon, Sep 4, 2006 08:26 PM UTC:Dan: You don't need to restrict moves; it is just a way to save time. You don't need to calculate every move to some really deep depth if you know that move of them are definitely bad (like pushing edge pawns.) So, it goes something like this... select 'test this position' - set I-Depth to something like 5, and set the maximum number of moves to a big number (like 100) so that everything will be tested. Click ok, and then be patient ... What it is doing is making each move and then running an I-Depth 5 search on the resulting position to determine a value for this move. Then it undoes that move and tries the next. When it is done it will present a list of all moves tested, and the resulting evaluation. A depth-5 search isn't deep enough to determine which moves are the best, but it is good enough to rule out really bad moves. Then, to test the 'good' moves to a deeper depth, select 'test this position' again. Turn up the depth, limit the moves to the top 12 or so, and click ok. Then repeat it again, testing the top 6 positions or so to a depth of 9 or greater. When it completes, any moves which are within 50 points or so of the best move should be added to the opening book. Of course, you could just test all moves to a depth of 9 from the start, but that wastes a LOT of time. The 'test this position' search uses a more intensive search in several ways than the normal search to ensure the best results. But this means that it is quite slow. If your computer has a lot of ram, before you start testing, change the computer settings, and allocate more memory for the transposition table and evaluation cache. If you allocate more memory than you can spare, though, the operating system will start swapping to disk, and that will slow things down to a crawl, so use the task manager to check your free physical memory first. Any more than 512MB for the transposition table is totally pointless, and any more than 256MB for the evaluation cache is probably pointless. Also, when setting the table sizes, use powers of two. The size of the tables must be a power of two, so if you enter 100MB, ChessV is just going to round it down to 64.
There are a few bugs in the program. Some of the 'Chess with Different Armies' armies don't work, like the Andromidan Invaders and the Almost Chess army, instead of them you get the standard army. And in Chess with Augmented Knights the knights are not able to move because they're pinned (!), and the Standard Knight doesn't have a picture of his own. and sorry for trouble.
📝Greg Strong wrote on Thu, Sep 7, 2006 10:50 PM UTC:Abdul-Rahman: Thank you for the bug report. Very good observations! These were minor problems that have been corrected. I am also adding a couple of tiny optimizations, and a new game or two, and perhaps a couple more texture options for the squares, and then I will release 0.9.1. You can expect it to be released within the week. NOTE: The problem with the Andromedan Invaders has been resolved by removing that army. I created that army only for the purposes of testing strange pieces used in other games, but they were never balanced in strength with the FIDE army. I might turn them into a real army and reactivate them again someday. Please continue to report any bugs that you discover.
Oh, i posted a few weeks ago, but it seems when all those dumb 'my chess program is better than yours' posts were deleted, my post went too. Ok, so, what a brilliant program! And it's free! And best of all, for me anyway, it plays Great Shatranj, and it plays very good with pawns too. Hopefully, this program will just get better and better, congrats Greg!!
Yes, Dan Kelly is a good chess player in my estimation, he beat me yesterday, and I am challenging him to another game today
If you have a copy of the saved games where you beat ChessV, I know Greg would love to see them to see where ChessV went wrong.
- Sam
25 comments displayed
Permalink to the exact comments currently displayed.