OpenChess

this http://talkchess.com/forum/viewtopic.ph ... 48&t=46586

I think is interesting. Ugly hacks have been introduced in a polyglot-variant to allow the user to set the number of cores of UCI-engines within the Win-/Xboard-GUI 'easily' (the very reason for the hacks?). Luckily there's still a fork (?) out there from Fonzy Bluemers: http://www.geenvis.net/polyglot1.4w29.zip which is working different but as indendet (certainly as it has ever been: xboard for winboard-only engines; for uci there's polyglot to take care ALL of the options).

My question: How do users in Windows (with Winboard) environment achieve multithreaded-IvanHoe for example? Do you use another (as opposed to built-in) polyglot or is there some other workaround?

alfons wrote:this http://talkchess.com/forum/viewtopic.ph ... 48&t=46586

I think is interesting. Ugly hacks have been introduced in a polyglot-variant to allow the user to set the number of cores of UCI-engines within the Win-/Xboard-GUI 'easily' (the very reason for the hacks?). Luckily there's still a fork (?) out there from Fonzy Bluemers: http://www.geenvis.net/polyglot1.4w29.zip which is working different but as indendet (certainly as it has ever been: xboard for winboard-only engines; for uci there's polyglot to take care ALL of the options).

My question: How do users in Windows (with Winboard) environment achieve multithreaded-IvanHoe for example? Do you use another (as opposed to built-in) polyglot or is there some other workaround?

The concept of setting the number of CPUs in Winboard is to act as a global option for any/all engines loaded. H.G.M does make a good point regarding the confusions from the various SMP [UCI] names & options employed by various chess engine authors. This forces each GUI developers to adapt to yet another "new" SMP names/options - when in fact, the other-way round should solve this engine-GUI conflicts, as suggested by H.G.M: make all authors comply with one universal SMP names/options. This would not only make things/programming easier for GUI authors, but will eliminate most of engine-GUI compliance disparities we see with various engines/GUIs.

On your question regarding IvanHoe's [SMP] Threads, those compiling the newer or latest versions of IvanHoe from its source followed one of the common and accepted SMP name/options (at least, compared to its earlier TitanicCPUsSizing options) and it appears that "Threads" is one of the SMP names/options already predefined under Winboard's Max nr of CPUs by H.G.M. Bottom line is this, newer IvanHoe's SMP name/option has no conflict with Winboard's Max nr of CPUs and obeys whatever number of CPUs is set in Winboard's Global UCI option menu. If I remember correctly (I'm too lazy to test it out now), the older/earlier IvanHoe versions' SMP options was visible in engine parameter (after being loaded in Winboard) had to be set in it engine parameter-dialog. Not so with the newer, SMP name/option-compliant IvanHoes, as its SMP option is no longer visible in its engine parameters - an indication that it & Winboard are in-sync in this aspect.

Hope this answers your concern.

PS: I do have my concern regarding IvanHoe's RobboBases which I would like to ask H.G.M: Are Robbobases now incorporated into Winboard's egtFormat like Nalimov, Gaviota, and Scorpios EGBB have been? Can Winboard now recognize and honor (honour) Robbobases with respect to the IvanHoe?

PS: I do have my concern regarding IvanHoe's RobboBases which I would like to ask H.G.M: Are Robbobases now incorporated into Winboard's egtFormat like Nalimov, Gaviota, and Scorpios EGBB have been? Can Winboard now recognize and honor (honour) Robbobases with respect to the IvanHoe?

Ha! My point...(a bit) - again, no problem here with polyglot from Fonzy (the default built-in wouldn't recognize the Bases for whatever reason; perhaps because they're not: nalimov, gaviota and so on). This has droven me mad too (and the DEFAULT 4 mb Nalimov cache-nonsense).

nb:
The startup on Linux is shitty* ('fairymax excited unexpectedly')...; selecting an Engine (Stockfish z.B.) leads to nothing (it excitesbla) - I did never get so far to see such a nice configuration-window like your attached file shows. Version 4.6.2 sits here in my bin folder; running tournaments through scripts is a joy...; that cannot be said about the ability to change things by graphical menu (with its mindbogglingancient own-file-chooser/manager).

I do like of course the new features overall, but I'm certainly annoyed of some regressions those changes introduced.


*the maintainer from Debian is to blame

alfons wrote:

PS: I do have my concern regarding IvanHoe's RobboBases which I would like to ask H.G.M: Are Robbobases now incorporated into Winboard's egtFormat like Nalimov, Gaviota, and Scorpios EGBB have been? Can Winboard now recognize and honor (honour) Robbobases with respect to the IvanHoe?

Ha! My point...(a bit) - again, no problem here with polyglot from Fonzy. This has droven me mad too (and the DEFAULT 4 mb Nalimov cache-nonsense).

nb:
The startup on Linux is shitty* ('fairymax excited unexpectedly')...; selecting an Engine (Stockfish z.B.) leads to nothing (it excitesbla) - I did never get so far to see such a nice configuration-window like your attached file shows. Version 4.6.2 sits here in my bin folder; running tournaments through scripts is a joy...; that cannot be said about the ability to change things by graphical menu (with its mindbogglingancient own-file-chooser/manager).

I do like of course the new features overall, but I'm certainly annoyed of some regressions those changes introduced.


*the maintainer from Debian is to blame

I'm not exactly sure what's happening on your end. Perhaps misconfiguration somewhere? I intended to address the PolyGlot I use - when I noticed you already answered my post. I have been using Michael Van den Berg's PolyGlot, for the fact that Michael's PolyGlots allow a user to set options such as: Book depth, Book treshold, enable/disable Book Randomness etc. Fonzy's PolyGlot, if I recall correctly, doesn't provide these options to users. H.G.M's PoklyGklot (vwersion 2.0.1) also incorporates same options as Michael's plus few more, which I also use.

BTW, I'm using Winboard 4.7 betas (the last beta vwersion is 4.7.2012.1223). It's got many GUI features added and improved on. Bugs fixed. Graphical User Interface improved and can be alternated seamlessly without manually editing the Winboard.ini like you would with previous Winboard versions. For all practicall purposes, Winboard 4.6.2, though the last stable released version, is somewhat obsolete in certain functions and graphics capabilities when compared to Winboard 4.7 betas.

Thanks to H.G.M for all that he has done in forwarding Winboard/Winboard X and immense help/contributions to the chess community. Winboard keeps getting more and more pleasant. PolyGlot now seems to be the bottleneck to us Winboard users for the fact that PolyGlot's book algorithm and its book-learn system is broken.

Attached below is just a side show of Winboard 4.7 beta GUI seamless alternation without editing. Of course, the chess fonts and textures etc require the user to preset them in Winboard 4.7 beta's.ini:

allow a user to set options such as: Book depth, Book treshold

I think not. Setting the depth of a book while playing is most useful indeed.

enable/disable Book Randomness

Yes.

Thanks to H.G.M for all that he has done in forwarding Winboard/Winboard X and immense help/contributions to the chess community.

I second that: The GUI is very fast with a small footprint on the system while running, it never crashed on my system and is nicely scriptable. Anyway, the development is aimed at switching to Gtk entirely (not tried it yet though; yesterday I tried to create a chessfont (ttf) of hgm's chesspieces bitmaps with fontforge for fun (which is wasn't); so perhaps later...

bye