Author: Shep
Date: 10:17:50 11/22/00
Go up one level in this thread
One of the things you must not forget is that this kind of "semi-public" beta-testing Ed and Christophe did is bound to be "incomplete". Complete beta-testing can only be done in-house; you need a precisely defined list of things to test etc. If you give the program to 20 people, there are 3 possibilities: 1) Everyone tests the way he wants to. Given the experience of the testers, this will likely catch a lot of bugs (in my experience, 99% of the important bugs occur during the first few days of intensive use by a skilled user). 2) Everyone tests according to a defined list. This will of course end up with 20 people testing the same things, 20 times testing the KBNK endgame, 20 times testing of "save a layout" etc. Thus it generates lots of overhead. 3) Everyone receives a well-defined part of the overall "to do" list. Of course 2) and especially 3) would require so much professionality from the testers that they would have to be paid for this. --- Shep
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.