Author: KarinsDad
Date: 17:44:09 01/06/99
Go up one level in this thread
On January 06, 1999 at 18:47:03, José de Jesús García Ruvalcaba wrote: >On January 06, 1999 at 17:36:35, KarinsDad wrote: > >>On January 06, 1999 at 16:11:16, José de Jesús García Ruvalcaba wrote: >> >>> I am not sure about what means that a program is stronger (or weaker) than >>>other program. >>> I see the program, its settings and the machine it is playing on as a single >>>playing entity (not to forget that also at different time controls we can get >>>very different results). >>> For me "Program X with such and such settings running on machine A is stronger >>>than program Y with some settings running on machine B at a certain time >>>control" is a meaningful statement; but "program X is stronger than program Y" >>>is not. >> >>If you use the phrase "program X is stronger than program Y", wouldn't you by >>default have to assume comparable settings and comparable environments (OS as >>well as hardware)? >> > >No. Some programs are designed specifically for some architecture. It would be >almost impossible to port them, and doing so would cripple them a lot. By comparable settings and environments, I did not mean specific ones. For example, Unix is comparable to Windows 95 as a multitasking 32 bit operating system. Yes there are 64 bit versions of Unix, but that would basically be comparing two noncomparable OSs. Comparable settings mean that if one program has the capacity to search during it's opponent's time, so would the other. Obviously, some programs have "features" that others do not. If one program is running on 450 Mhz Alpha chip, and the other is running on a 166 Mhz Pentium, then if you do not give a time advantage to the second program that corresponds to benchmarks between the two chips (and their motherboards), then you have an incompatability. The complexity of whether two systems are compatable is a difficult one. There is no way to be exact unless the two programs are running on identical hardware/software platforms which are configured identically, however, you can ballpark. Actually, I personally know of no programs that are intentionally designed to run only on a specific platform (intentionally implemented is a different matter). They probably exist. Granted, some programs have assembly code built in, but it wouldn't be a monumental task to convert one set of assembly code to another set unless a good majority of the program was built that way. With the state of today's compilers, your time would be better spent improving your algorithms than speeding up more than a small portion of oft used code in assembly language. And it seems real unlikely that porting a chess engine over to another platform would cripple it. Your basic premise is both correct and accurate: "Program X with such and such settings running on machine A is stronger than program Y with some settings running on machine B at a certain time control" is a more meaningful statement than "Program X is stronger than program Y". However, that does not mean that the statement "Program X is stronger than program Y" has no meaning. In the Crafty/Sargon example above, it is a relatively accurate and meaningful statement. KarinsDad > >>Also, some programs are so much "stronger" than others that unless you limit >>them in some way (such as turning off processing while the opponent moves, or >>limiting the depth of the search), they will still beat the "weaker" program in >>general while running on their minimal configuration while the weaker program >>runs on a much more powerful configuration. >> > >If two programs are intended for the same plattform, I agree that saying that >one is stronger than the other is to be interpreted as you propose. Still, if >the difference in strenght is small, the time control becomes an important >factor. This is obviously a factor. There are a lot of factors involved. However, I can guarantee you that Crafty 16.2 will defeat Sargon III (from the early 80s) every day of the week. They wouldn't even draw. Crafty 16.2 runs on Unix (it probably can also run in Windows). Sargon III runs in DOS. Can I state that Crafty 16.2 is stronger than Sargon III? You bet. > >>While technically correct in your phraseology, this is merely a matter of >>semantics. >> >>:) >> >>KarinsDad
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.