Author: blass uri
Date: 21:52:06 12/03/98
Go up one level in this thread
On December 03, 1998 at 22:01:25, Detlef Pordzik wrote: >On December 03, 1998 at 07:51:28, blass uri wrote: > >> >>On December 03, 1998 at 05:49:37, Amir Ban wrote: >> >>>On December 03, 1998 at 05:21:19, blass uri wrote: >>> >>>> >>>>On December 03, 1998 at 05:04:52, Harald Faber wrote: >>>> >>>>>On December 03, 1998 at 02:31:09, Hans Christian Lykke wrote: >>>>> >>>>>>Junior lost on time >>>>> >>>>>>[White "Junior 5 P200MMX 47MB"] >>>>>>[Black "Fritz 3 P90 15MB"] >>>>>>[Result "0-1"] >>>>> >>>>>I didn't think it is possible and programmes have implemented kind of panic time >>>>>or so. Stil wondering.... >>>> >>>>It is possible >>>>I discovered this bug before some weeks and told Amir Ban about it. >>>> >>>>Junior always feels a duty to finish all the moves in the iteration it does and >>>>this is the reason for it. >>>> >>>>Usually it does not happen but it can happen 1 time out of many games. >>>> >>>> >>> >>>This is not supposed to happen. This is an engine interface issue. The engine >>>tells the shell when it wants a wake-up call, and the shell sends it. This was >>>not working properly for J46, but was fixed for J5. Looks like we didn't test it >>>thouroughly enough. >>> >>>I suppose that what actually happened is that J5 declared itself lost ? There is >>>no other way to lose on time in auto232. It's quite possible to overstep the >>>time limit without anything happening. >>> >>>Amir >>I looked in the times and the if I compute the total time of Junior5 then I see >>that the time is only 6892 seconds so I do not understand why Junior lost in >>time in this game. >> >>The total time of Firz3 was 7243 seconds after 37 moves and 2708 seconds for the >>38 move. >> >>If there is no mistakes in the times that Junior5 printed then Fritz3 lost on >>time >> >>Uri > >I really don't dare to trust my eyes, collegue Uri ! >What should such a game setup teach ? >A 48 megs 200MMX based 1998 prog comparing to a 1995 prog under a platform of 15 >Megs P 90. >THIS missleads, and THIS affects - clearly, the rating - noone to wonder. This does not misdlead because the rating is of fritz3(pentium90) Fritz5(pentium200) also played Fritz3(pentium90) and noone played Fritz3(pentium200) >Tell me please - do you vary 200 MMX vs 200 MMX - or do you play, whatever >combination you can gat a hold on ? I do not play between computers. The ssdf volunteers do it. The ssdf volunteers always give new programs better hardware. It is not misleading because in the rating list there is no fritz3(pentium200) but only Fritz3(pentium90). 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.