Author: Robert Hyatt
Date: 05:11:30 04/26/00
Go up one level in this thread
On April 26, 2000 at 00:53:37, Bruce Moreland wrote: >On April 25, 2000 at 18:23:59, Robert Hyatt wrote: > >>I don't think it is "horribly crippled". Just "crippled". In ways I >>understand, but don't want to waste the time fixing. Too much depends on >>saving some time with pondering. Turning it off would be doable, but a lot >>of testing and modification would be needed to fix all the holes doing that >>creates. I barely have enough time to work on the program as is, and I would >>much rather work on the Beowulf search than spend time tuning something I don't >>and won't ever use... >> >>If you are rich, money is no object. In this case, time is money and I am >>nearly broke. I watch _every_ penny (minute) of time I spend since I have so >>little of it nowadays.. > >Perhaps you could default pondering to on, and comment out the "ponder off" >command. If someone wanted to turn it back on for testing, they could, but if >someone wants to use a stock version of your thing to do a "I'm going to prove >that Bob is stupid" test, they'd have to modify your code to do it, which would >make the test more than a bit suspect. > >bruce I had thought about this. But then _I_ get into the hot seat all of a sudden as when I am testing I always use ponder=off so I can reproduce the same node counts over and over, when playing a few moves in a game. By doing this, I would be continually either (a) editing the source to enable this for every test run or (b) releasing versions with it accidentally still enabled. I am doing my best to follow the KISS principle. But it is sometimes difficult. :)
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.