Hi,
I have an EPD file with several hundreds positions that I use for testing engines. With windows, I use Arena which is very good for automatic analyis of several positions. Then, it gives a log with the number of positions solved, time used etc.
My question is : is there the same kind of software for Linux ?
Thanks.
Test epd for Linux ?
Moderators: hgm, Rebel, chrisw
-
- Posts: 338
- Joined: Tue Mar 13, 2012 9:59 pm
- Location: Germany
Re: Test epd for Linux ?
I know some Linux GUIs and I am not aware of any that can do that.
Have you considered using Arena in WINE? It usually works very well.
Have you considered using Arena in WINE? It usually works very well.
-
- Posts: 12662
- Joined: Wed Mar 08, 2006 8:57 pm
- Location: Redmond, WA USA
Re: Test epd for Linux ?
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
-
- Posts: 125
- Joined: Tue Jan 17, 2012 8:08 pm
Re: Test epd for Linux ?
There is no makefile. How do I compile it ?Dann Corbit wrote:The closest thing I can think of is epd2wb.
http://www.quarkchess.de/quark/epd2wb.zip
-
- Posts: 12662
- Joined: Wed Mar 08, 2006 8:57 pm
- Location: Redmond, WA USA
Re: Test epd for Linux ?
I remembered it was open source, but it is full of Windows specific function calls. It will be a problem to make it compile on a Posix system.
On WIndows, you can just compile epd2db.cpp and off you go, but on Posix systems, the windows specific calls would have to be replaced.
On WIndows, you can just compile epd2db.cpp and off you go, but on Posix systems, the windows specific calls would have to be replaced.
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
-
- Posts: 549
- Joined: Thu Mar 09, 2006 12:53 am
Re: Test epd for Linux ?
I also use Arena running via WINE. Seems to work just fine.styx wrote:I know some Linux GUIs and I am not aware of any that can do that.
Have you considered using Arena in WINE? It usually works very well.
-
- Posts: 125
- Joined: Tue Jan 17, 2012 8:08 pm
Re: Test epd for Linux ?
Wine is of no use for me because I have a dual boot windows/linux ...
-
- Posts: 125
- Joined: Tue Jan 17, 2012 8:08 pm
Re: Test epd for Linux ?
The archive alreaday contains the windows executable.Dann Corbit wrote:I remembered it was open source, but it is full of Windows specific function calls. It will be a problem to make it compile on a Posix system.
On WIndows, you can just compile epd2db.cpp and off you go, but on Posix systems, the windows specific calls would have to be replaced.
I need a 100% Linux software. I need all the power of Linux. The cpp file was obviously made for windoze.
I have found a funny way to run epd tests with Polyglot.
./polyglot14l epd-test -epd STS1-STS15.epd -max-time 10 -depth-delta 3 > test.txt
More infos here http://www.open-aurec.com/wbforum/viewt ... 61&start=0
The problem is that you must read the file and find by yourself what it missed... Not very practical when you run 1500 positions.
-
- Posts: 12662
- Joined: Wed Mar 08, 2006 8:57 pm
- Location: Redmond, WA USA
Re: Test epd for Linux ?
Here is a version by HGM that compiles on Linux:
http://hgm.nubati.net/cgi-bin/gitweb.cg ... EAD;sf=tgz
http://hgm.nubati.net/cgi-bin/gitweb.cg ... EAD;sf=tgz
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
-
- Posts: 28123
- Joined: Fri Mar 10, 2006 10:06 am
- Location: Amsterdam
- Full name: H G Muller
Re: Test epd for Linux ?
That is uci2wb, not epd2wb. And, as the name suggests, it does something entirely different: it is an adapter that allows you to run UCI engines under a WB GUI.
Unfortunately I have not equiped XBoard with any EPD-solving capability yet; I always figured there were enough programs to do that, and it is more a batch thing than a GUI thing anyway. I once made a start implementing it (IIRC XBoard already remembers the bm field when reading an EPD, and has an -epd option to cut terminate each 'game' after a single move in match mode, so that starting it as "xboard -matchGames 20 -epd -loadPositionFile foo.epd -loadPositionIndex -1" would auto-step through the first 20 positions of foo.epd, and let the engine play one move on each of them. The results are then accounted in the match result, counting good solutions as wins, and avoidance of an avoid moves as draws.
There is no code yet to report the indiviual results, however. Originally I had planned to display a report in the second pane of the Engine-Output window. But the problem was that if WinBoard encouters a start position where black is to move, it would put the second engine to think about it in match mode, which would then overwrite that second pane with thinking output. So it would require special treatment of match mode when the -epd option is on, not alternating the colors, but making the color assignment such that the the first enginewould always have the first move, so that the second engine never needs to be invoked at all.
Unfortunately I have not equiped XBoard with any EPD-solving capability yet; I always figured there were enough programs to do that, and it is more a batch thing than a GUI thing anyway. I once made a start implementing it (IIRC XBoard already remembers the bm field when reading an EPD, and has an -epd option to cut terminate each 'game' after a single move in match mode, so that starting it as "xboard -matchGames 20 -epd -loadPositionFile foo.epd -loadPositionIndex -1" would auto-step through the first 20 positions of foo.epd, and let the engine play one move on each of them. The results are then accounted in the match result, counting good solutions as wins, and avoidance of an avoid moves as draws.
There is no code yet to report the indiviual results, however. Originally I had planned to display a report in the second pane of the Engine-Output window. But the problem was that if WinBoard encouters a start position where black is to move, it would put the second engine to think about it in match mode, which would then overwrite that second pane with thinking output. So it would require special treatment of match mode when the -epd option is on, not alternating the colors, but making the color assignment such that the the first enginewould always have the first move, so that the second engine never needs to be invoked at all.