Author: Uri Blass
Date: 04:47:24 01/25/06
Go up one level in this thread
On January 25, 2006 at 07:21:41, Kolss wrote: >On January 25, 2006 at 06:01:18, Uri Blass wrote: > >>I wonder if people tried to use arena to test rybka with reduced strength and if >>arena works without bugs with that feature. >> >>testing it with movei suggest that arena1.1 can steal significant time from >>engines and it may be the reason for the strange results in 1+1 under arena that >>I read here(I found that more than 0.5 second was stolen from movei after >>playing the first move and similiar situation for other moves). >> >>I wonder if there is a good interface that cause no problems for testing with >>unequal time control because it seems that arena may be good only for slow time >>control in the best case and testing movei with 40 minutes/40 moves against >>rybka with 4 minutes/40 moves can take a long time. >> >>Uri > > >Hi, > >I have never tried to test with time handicap. >We never use Arena for testing though - with good reason: > >Perhaps it is my incapability, but the Arena GUI is *painfully* slow. The loss >of 0.5 seconds per move is absolutely normal. Although I tried to switch off >most of the features with which the GUI is totally overloaded, it is still >extremely slow. Especially the display of principal variations seems to take up >enormous amounts of time (I have already switched off the move list window, as >that is even almost worse...). Something seems to go seriously wrong there (I >have this problem on all of my computers, some of which are *not* slow). > >You can say whatever you want about the Chessbase GUI, but it is certainly >amazingly fast. The loss of time for communication and display must probably be >measured in microseconds, at most milliseconds, while for Arena, it is in >centiseconds or even seconds, and some of this time is deduced from the engine'e >clock. > >For me, that renders the Arena GUI useless for matches with short time control. >You have to make sure that the time left on the clock never drops below about 1 >minute in a *normal* game (meaning within about 150 moves) to avoid that time >scramble (and time forfeits) becomes an issue. That excludes all blitz games. > >Again, this is just my personal observation and possibly my inability to work >with the Arena GUI properly; also, it holds for Arena 1.1 - I don't know whether >this problem has been tackled in succeeding versions... > >Best regards - Munjong. I prefer to use winboard for that reason but unfortunately winboard does not support unequal time control(latest version support unequal time control when white get more time but I want unequal time control when one engine get more time. I can play 2 matches one with white get more time and one with black get more time but it means that I have limited computer time for testing and I do not like to spend time for irrelevant testing. Uri
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.