Author: STEPHEN A. BOAK
Date: 09:11:00 02/28/99
Go up one level in this thread
On February 28, 1999 at 02:06:10, Timothy J. Frohlick wrote: >Just before I took my dogs for a walk up in the hills of Albuquerque, I decided >to download the Rebel 10c upgrade. I was amazed that the Benchtest program >rated my machine at 2596. This was a 60 point increase. I was ecstatic and >played a couple of quick games and was quickly dispatched. Energized by this >new chess terror, I embarked on a short four mile hike among the coyotes. > >Going up the hill a guy in a pickup truck races by and shouts, "Have a nice >day!" He turns around and I holler back "Happy New Year!!!" They get out of >the truck and go to a nearby house. I say, " I thought you guys won the >lottery!" Everybody laughs. In the distance I hear the yipping of the coyote >pack. > >I get home and take a close look at the benchtest and find as did Francois >Bertin that the program only analyzes the middle chess problem and gives a -1 >second time for the first and the last. What? Could this be a bug? I have the same experience with the benchtest of Rebel10c. Check Rebel site and Schroder says the benchtests are not comparable, between Rebel10c and prior versions. He said he'd ask users to use the benchtest found on the original Rebel10 CDROM, when sending in benchtest scores. I play >another game and take it to a KPQ vs K ending and predict a mate in 14. The >program cogitated and made good moves and found a mate in 7, seven moves later. > >So what we have here is a very good program but it is not a quantum jump. It is >more of a yip and not a howl.
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.