Bunding WinBoard and suport programs

[Guenther Simon]

BTW because I edited my previous post very late, you probably missed
the point about my plea for the games of your test

Done


Marc

Thanks Marc, I have already received them

Guenther

[H.G.Muller]

My preference would go out to a class B book, as the purpose is not to immediately provide people with top-of-the-line books and engines, but just include examples, which will show them how things are done. This is why I proposed to include CPW and Fairy-Max, and not Fruit and Glaurung. (Another reason was that Fairy-Max plays some variants, so people could immediately try those too.) The inclusion of a book should be seen in a similar light.

If people say: "this book or engine is truly awful, I must get me another one", I would not think it a bad thing at all. We could add an html maual page that decribes in a few words how and where to download a book, including a list of links to publically available books, classified according to Marc's system.

[Thomas McBurney]

Hi Roger!

I haven't released the Game Analyser yet simply because it hasn't been tested to any significant degree. Maybe I should just upload it 'as is' to my web page and then let the Yahoo spam filter deal with the pursuing flood of bug reports.

Development of Kanguruh has stopped and I don't plan to do any more work on it. However, I am working on a new chess engine that is written in FreeBASIC, but unfortunately progress is painfully slow due to lack of time. My new chess engine is only half finished but can play without any problems. I will release it sometime in the future when I am satisfied with it and I plan to port it to Linux as well.

Cheers,
Tom.

[Thomas McBurney]

Is TLCS open source? I am not sure we can distribute software with WinBoard of which we are not prepared to hand over the source, or we might be violating the GPL on WinBoard.

About the engine manager: Is this part of WinBoard, or is it a separate program? From the screenshots it seems like you start it from the WinBoard menu. What exactly does it do? Is it an editor for the winboard.ini file, which upates the "/chessProgramNames" options in it?

Yes, it's a separate program that updates the winboard.ini file. But there is more to it. Winboard and Engman can communicate to each other via Windows Messaging. We had the following events set up...

' Send msg codes
Global Const WM_EngMan_Started = 1
Global Const WM_EngMan_Terminated = 2
Global Const WM_EngMan_Error = 3
Global Const WM_EngMan_Register = 4


' receive msg codes
Global Const WM_NotifyEngMan_Minimize = 1
Global Const WM_NotifyEngMan_Restore = 2
Global Const WM_NotifyEngMan_Quit = 3
Global Const WM_NotifyEngMan_Activate = 4
Global Const WM_NotifyEngMan_Registered = 5

When the user minimised the Winboard window, the Engman window would also be minimised. And if the user closed Winboard then Engman would also close automatically. We had plans to add more events, such as notification of a new Engine etc, but never got around to it.

Now that Thomas is here: One of the reasons I wanted to contact you before is this: some engines seem to crash TLCS by overloading the winboard.debug file with massive protocol-violating output. WinBoard ignores this, but it does end up in the debug file. Apparently it then overruns some input buffers in TLCS. Or perhaps they contain some keywords that confuse TLCS.

Ok, that's interesting. I definitely want to know about these issues. Can you tell me which engines you had this problem with? Or provide me with the debug file?

Would it be easy for you to change TLCS much that it ignores any line that starts with a '#' sign? I have included an option to WinBoard that would prefix every protocol-violating input line it ignores with '#' before writing it into the .debug file, for engines that do not do this themselves. I would be surprised if having TLCS simply ignore such lines would not solve almost all problems with such rogue engines.

No problem, I will do this today.

Cheers,
Tom.

[Roger Brown]

Hi Roger!

I haven't released the Game Analyser yet simply because it hasn't been tested to any significant degree. Maybe I should just upload it 'as is' to my web page and then let the Yahoo spam filter deal with the pursuing flood of bug reports.

Yes please! of course, since your spam filter works in reverse, it is entirely likely that the spam filter will now freely allow all of the unwanted mail (the bug reports) right through to your inbox.

Development of Kanguruh has stopped and I don't plan to do any more work on it.

Very well.

However, I am working on a new chess engine that is written in FreeBASIC, but unfortunately progress is painfully slow due to lack of time. My new chess engine is only half finished but can play without any problems. I will release it sometime in the future when I am satisfied with it and I plan to port it to Linux as well.

I will be here!

All the best Thomas.

[Denis P. Mendoza]

My preference would go out to a class B book, as the purpose is not to immediately provide people with top-of-the-line books and engines, but just include examples, which will show them how things are done. This is why I proposed to include CPW and Fairy-Max, and not Fruit and Glaurung. (Another reason was that Fairy-Max plays some variants, so people could immediately try those too.) The inclusion of a book should be seen in a similar light.

If people say: "this book or engine is truly awful, I must get me another one", I would not think it a bad thing at all. We could add an html maual page that decribes in a few words how and where to download a book, including a list of links to publically available books, classified according to Marc's system.

I'm into polyglot bookmaking for a while, thanks to Marc. I've done a few for the right categories of engine testing -class B, C and D. I rechecked them using SCID and ran a few tests, but didn't adjust the weights. These are just polyglot conversions of the common classic opening books we use. Pls. have a look and see if they may be of use. Thanks.

http://computerchessengines.mylivepage. ... ileid=3992

PG_Classic_10-moves.bin
PG_Neutral_12moves.bin
PG_Perfect.bin
PG_Random_10moves.bin
PG_Traps.bin

[H.G.Muller]

It is time to return to this project.

Now that we fixed the problem that prevented xboard to work under PSWBTM, I think we should do a release of this fixed version as soon as possible. And for the Windows and Linux executable packages, we should implement this idea of bundling.

The WinBoard in this release (which will then be 4.3.15) can contain the few additions I made since the last release. (These concern mainly the auto-increment of the game or position index in match mode, the addition of a new variant Superchess, and the fixing of the time in pgnExtendedInfo, which now uses GUI time.) During my holiday I had some time to learn about and play with the X-windows front-end of xboard, and am currently trying to get the engine-output window working in xboard. I hope I get this running fast enough to include it as well.

When studying the xboard code, I noticed that the Winboard_x game-history window already had xboard code for it in Tim Mann's 4.2.7 version, but it seems the code was never called. (I could find no menu items or command-line options to pop this window up.) The code seems defective, though, as now that I added a menu item to pop it up, using that menu item crashes xboard. I am not sure if this is due to an error I introduced. Can someone used to working with xboard confirm that the game-history window (called 'move list' in 4.2.7) was indeed dead code, or had problems?

[Michel]

And for the Windows and Linux executable packages, we should implement this idea of bundling.

Bundling goes against the Linux philosophy. Software under Linux is distributed through apt-get or yum. Bundling is achieved through dependencies between packages.
E.g. PSWBTM would have wxWidgets and xboard as a dependency. The new xboard would have polyglot as a dependency. This is a very convenient system which makes installing software under Linux trivial.

I would prefer to get PSWBTM and xboard into the Debian/Ubuntu repositories (like I did for Toga 1.4.1SE).

For xboard there would not be a problem except that I suspect that perhaps a new name should be used (maybe xboard3 ?). I will write to the current xboard Debian maintainer to see what he thinks.

PSWBTM may have a non GPL compatible license. However it is still open source in the OSI sense. So unless I am mistaken there should be no problems there either (if the unicode issue can be resolved).

[H.G.Muller]

Well, for Windows the bundling is most certainly needed, as most Windows users consider WinBoard too user-unfriendly to use, for no other reason that it is not bundled with its support programs.

As long as xboard 4.3 is not in any official Linux distro, I think it would still be useful to bundle it in an executable package, although such executable packages might be contrary to the Linux spirit in the first place. I don't know if a Linux execuatble can be distributed like a Windows executable an, with the guarantee that it will run on any Linux. I would like to distribute something that does not draw on external run-time libraries, but perhaps this is not possible for Linux.

As to getting to a stage where xboard 4.3 will find its place in the apt-get circuit: I don't really know how this works. I noticed that most packages come with version numbers. What if people want to install xboard under Linux, how would they know what version to request? Should they do "apt-get intall xboard.4.2.7" if they want Tim Mann's version? What if they want to have xboard 4.2.5 or 4.1.7? How can they know what the most recent version is?

[Michel]

What if they want to have xboard 4.2.5 or 4.1.7? How can they know what the most recent version is?

It you do apt-get install <package> you get the most recent version which is usually
what you want. If you want older versions you could try to install the .deb
package directly using dpkg but this might well give depency failures.

As to xboard: either 4.3 is accepted as an xboard version in which case
apt-get install xboard will yield 4.3. Otherwise there will be a different xboard package
(e.g. xboard3). Then users can do either "apt-get install xboard" to get
Tim Mann's version or "apt-get install xboard3" to get the new version.

Most users use Synaptic to install software on Linux. So they will see they have
the choice between two versions, and the attached comments can explain what is going on.

As said we need to ask the Debian maintainer of xboard.

I don't know if a Linux executable can be distributed like a Windows executable an, with the guarantee that it will run on any Linux.

You could try to statically compile xboard. But I am not sure if you can make one binary which works both on 32 bit and 64 bit systems.

[H.G.Muller]

OK, if that is how it works, I think it would be best if people that use "apt-get xboard" would get 4.3.15. Giving them 4.2.7 in that case would just lead to confusion, as they expect the latest version, and not a 7-year obsolete one. We would have to dress op the release in a way similar to 4.2.7, though, and make sure nothing is left out, in order to qualify.

[Michel]

I did a test compile of the xboard sources available at your website (not the private version you sent me) for Ubuntu 8.10.

There were some minor issues:

- configure was not executable after unzipping. I guess zip does not preserve file permissions. .tar.gz does.

- missing xt headers were generically reported as "missing X headers or libraries".

- the Xaw dependency was not detected by configure.

The last two things are autoconf issues. I do not know autoconf very well, but I assume
it is just a question of making configure.ac up to date. I assume that xt and Xaw were
once standard parts of X11 but this is no longer the case with the appearance of gtk.

After satisfying the dependencies xboard compiled fine. When started it complained
about Fairy-Max not being found .

Regards,
Michel

PS. Somebody should really unify winboard/xboard using Qt. Qt is GPL (you can
also get an expensive closed source license). Qt is for example used by
Acrobat Reader both on Windows and Linux.

[Michel]

Here is a configure.in which I think properly detects the dependency on Xt and Xaw

http://alpha.uhasselt.be/Research/Algeb ... nfigure.in

You should replace the current configure.in and then run autoconf. This should generate a new configure.

Michel

[Miguel A. Ballicora]

OK, if that is how it works, I think it would be best if people that use "apt-get xboard" would get 4.3.15. Giving them 4.2.7 in that case would just lead to confusion, as they expect the latest version, and not a 7-year obsolete one. We would have to dress op the release in a way similar to 4.2.7, though, and make sure nothing is left out, in order to qualify.

Xboard 4.2.7 is already in an ubuntu package (that is how I installed it), as well as polyglot, crafty, sjeng, gnuchess etc. Installing (and uninstalling!) is really extremely easy for any user. Bundling is not necessary and it fact, it could be a nuisance for a regular user. The libraries and dependencies are already worked out in this packages.

Miguel