John S. J. Anderson:
>I think that's fine -- my position isn't based on any kind of
>'information should be free' moral position, but rather on the
>logic that all software has bugs, and without the source, it's very
>difficult to know that those bugs _aren't_ in the routines that
>calculate the answer that you're going to publish.
Thanks for the clarification. My beliefs are also much closer to
your position than the 'information should be free one.
>what I'd really like to see is some sort of certification by a trusted
>third party.
>>Since that's likely to be unworkable for a number of reasons, source
>code access is a workable intermediate step.
Ohh, that's an interesting thought. Problems I see are who will
pay for it, how to you come up with a uniform rating system, how do
users of the review balance usefulness vs. errors.
Still, I'm imagining a sticker for the program like "The XYZ seal of
approval" or the startup message "Inspected by ZYX". :)
Andrew
dalke at acm.org
