Re: DFSG2: Why we need clear guidelines, not woolly ones
For the record I think we have to do something like Ian's new DFSG. The key
aspect I find compelling about the new DFSG is the method of disallowing *all*
restrictions, and then exempting specific types of restrictions. In the long
term this method is actually far more robust than trying to list every type of
restriction we don't like.
A few miscellaneous related opinions and responses (I haven't been reading my
mail every day and so this is a kind of batch reply to lots of posts in this
thread):
Wooly language yay: I don't believe any attempt to make language more precise
by making it wordier has actually yielded fewer arguments over its meaning.
The existing DFSG is direct and explicit, it's quite well-written, I just
think the new methodology is better. Ideally I think we need a DFSG about the
length and precision of the old one but in the format of the new one where it
prohibits everything and then lists exceptions allowed.
Patch releases boo: I would never undertake to manage a software derivative
that I had to distribute via a patch, and if I wouldn't want to do it then I'm
not comfortable saying it's ok for me to require others to. I don't want to
prohibit Qt et al completely though, I'm beginning to think we need a class of
software we consider free enough to include in the distribution but not free
enough to consider "core" Debian software.
Mutable standards yay: The claim that mutable standards are anathema (is that
even proper grammar? :) is poppycock. An exception to allow requirements to
change the name on such changes is plenty of protection for standards and as
much protection as any copyright can really provide in any case.
Argument by assertion; Argument by repetition; Personal attacks; etc: BOO!
Please everyone just state your opinion, response to questions and arguments
with factual argumentative responses, don't just say "no you're stil wrong,
nyah nyah!"
--
greg
Reply to: