Author: Don Dailey
Date: 11:20:37 05/26/98
Go up one level in this thread
On May 25, 1998 at 14:09:37, Fernando Villegas wrote: >Hi All: >No matter how much prejudiced or biased against Fritz 5 Thorsten is or >seem to be, no matter how much "emotional" his statements appears for >the scientifically minded people here, no matter his style, - a good >worded one, to be sure, I say as a pro- - etc, etc, it is clear he has a >point with respect to the way programs are measured. It is just not fair >running programs in different platforms, that's all. I'm not sure what your point is, but if you are advocating that we continue to move toward a unified platform type of competition, then I must very strongly disagree. I think this really encourages a very limited style of writing a chess program and closes our minds to other possibilites. I keep seeing proposals to FIX the hardware to some standard commonly available platform. But I'm sure this encourages mediocrity. In affect you are saying, do not write a great program if it uses innovative techniques not supported by this or that hardware. Examples of this are programs that thrive with extra memory (let's punish em' if they use memory), more innovative hardware like 64 bit stuff from SUNS, SGI or DEC (let's punish em if they write 64 bit code) and even designing your own hardware. An incredibly intelligent future program is likely to have larger memory needs, ask any AI researcher, but let's punish anyone who try's to innovate here. The solution is simple, just publish the results WITH all the detail. When you do the tests, make it completely clear what the exact configuration of each hardware/software entity is. Publish sublists to show how each one does on the same exact platform when possible. But let's not force this stiffling kind of conformity which instead of being fair, which it claims to be, will have exactly the opposite affect. Let's use our imaginations, and not limit our resources. In my opinion, this is the only way (or some variation): Provide a couple of standard platforms. Each entrant has the option to use this platform OR he must provide his own to some testing organization, for his programs sole use. I really believe this is the only "fair" way to do it. It's not the easiest I'm afraid, but give me a break, let's not pretend forcing these constraints is somehow fair. But more than being unfair, it's very limiting to those who want to innovate and experiment with different hardware platforms. I sure hope this is not what everyone wants to happen. - Don
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.