On Sat, Jan 03, 2004 at 10:07:40AM +0000, MJ Ray wrote:
> >Almost all the support for non-free in Debian is a free result of
> >our support for free software. [...]
> >none of them would be significantly simpler
> >or even different if we didn't support non-free. [...]
> I disagree with your choice of "significantly".
That's nice. My comment is a result of my experience working on the
BTS, on testing and on the archive. Do you have any experience that
would back up any opinion you might have on this? Any repeatable
measurements? Anything beyond a sincere wish that it's true?
> >Basically, the issue is why spend 100 man hours on maintaining Debian,
> >then another 90 man hours on maintaining a separate non-free
> >repository,
> >when you can spend 101 man hours maintaining Debian and it's existing
> >support for non-free?
> Further, I disagree with your time estimates. The time spent on
> non-free is not directly related to time spent on Debian, either now
> or in the future. People are already spending time on non-free rather
> than on Debian and I think it likely that DD time is spent on non-free
> when it could be done by non-DDs.
Uh, there's nothing special about DDs compared to non-DDs. All it takes
is sitting through the n-m process, and given things like sponsorship,
it doesn't necessarily even take that.
> Basically, the issue is why waste
> any DD hours and project facilities on maintaining non-Debian things
> when you can spend all that on Debian?
You don't get to choose how developers spend their time; they do. If
they want to work on non-free software, that's what they'll do. If
that's all they want to do, presumably they'll quit Debian. If it's
not all they want to do, they'll have to devote less hours to Debian,
because they'll have a new non-Debian project to take care of.
I mean, sure, you might go from 99% of Debian development being on free
software to 100%, but if that's 100% of 50 man hours rather than 99%
of 100 man hours, that's a loss.
And if you're assuming that non-free is still important enough to Debian
users that someone will maintain infrastructure for it, then that effort's
got to come from someone, and that someone has to be familiar with Debian,
has to have a fair chunk of free time to donate to the Debian community,
and has to be reasonably skillful at admining and setup work; all of
which is basically the job description for Debian developers.
I'm not sure what's so hard to understand about this, or why people who
aren't willing to try setting up some replacement infrastructure are
nevertheless dismissive of how much effort that is.
> I also think that there will be benefits of new developers, supporters
> and collaborators if we let non-free go.
*shrug* I think you're dreaming.
Cheers,
aj
--
Anthony Towns <aj@humbug.org.au> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.
Linux.conf.au 2004 -- Because we can.
http://conf.linux.org.au/ -- Jan 12-17, 2004
Attachment:
signature.asc
Description: Digital signature