Re: GR: Removal of non-free
On Sat, Jan 03, 2004 at 02:10:28PM +0000, MJ Ray wrote:
> >> I have no measurements to give right now, but you don't either,
On 2004-01-03 16:46:34 +0000 Raul Miller <moth@debian.org> wrote:
> > He was asking for any of a wide variety of things, including
> > measurements. You have elected to provide none of those things --
> > and to focus purely on the measurement aspect of the question.
On Mon, Jan 05, 2004 at 12:22:38PM +0000, MJ Ray wrote:
> This accusation seems odd to me. Regardless, I thought the answer to
> the first questions was both obvious and public knowledge. For the
> record: I have run BTS and archives for projects of my employers only;
> I have not run any of the debian infrastructure.
I don't think it's fair to even assume that anyone has read every message
posted to this list. How is it fair to expect list readers to know what
you've done for your employer?
> The questioner clearly knew the last part of that already and I think
> that was why the question was asked.
I didn't see anything in Anthony's message to make me believe that he
knew about your employment history.
But maybe you're right -- I just don't know.
> I considered the last question ("Anything beyond a sincere wish
> [...]") not possible to answer beyond what was written in other
> messages, without sparking a long semi-OT thread.
I think a summarization of the relevant points would not be
off topic.
> > But the fundamental points here are:
>
> (Always a nice way to avoid the other questions.)
What other questions are being avoided?
> > You've claimed that non-free, as it currently exists "hinders
> > debian", but most of your claims seem to be based on false ideas
> > about why things are in non free, and what people spend their
> > time on.
>
> What have you been reading? On this list, I said what my current
> opinion is and why (I am puzzled why you quote "hinders debian".) I
> think I've never claimed to know all the reasons for DDs wanting to
> put things in non-free. Indeed, if you can give any evidence for my
> (unstated AFAICT) ideas being false, I'd love to read it and
> reconsider those ideas. If you have compelling reasons to keep
> non-free, or even some more insights, please post them.
I was thinking of stuff like:
. Are the people using the Debian infrastructure to support
. non-free helping to prevent the problems from being
. solved? Already, someone has mentioned some Java packages
. that I think could be in Debian but aren't. Is that because
. contrib is an easy enough home for them? If so, then removing
. non-free and contrib from our infrastructure would probably
. encourage them into Debian, solving one problem.
Here, you've implied that the presence of non-free hinders the progress
of debian.
But if you want to restate your reasons for getting rid of non-free,
that's fine with me: you're the expert on that.
> > Anthony has claimed that stripping non-free out of debian would
> > likely result in duplicated effort [and, thus, less productive time
> > available for debian]. He's offered his own experience maintaining
> > BTS and so on as his reason for thinking this.
>
> Anthony is probably one of the best qualified to show or describe
> interesting data about this, yet has preferred to make things up. That
> vexes me.
How is he to measure man-hours spent by debian developers on non-free,
vs. man hours spent by debian-developers on main?
Those are the measurements you're asking about, correct?
> >> There is the n-m process. I think that DDs have to know something
> >> to get through it, as well as spend the time sitting through it.
> > But drive (motivation and persisntence) has a lot more to do with it
> > than knowledge.
>
> Do you think that n-m is too easy and allows through people who do not
> agree with the philosophy, procedures, tasks and skills?
I think that "drive and motivation" vs. "knowledge" is nearly orthogonal
to issues of philosophy, procedures, tasks and skills.
> >> It should not surprise anyone that apparently fictional numbers
> >> support the arguments.
> > Ok, here's some numbers for the other side of the argument:
>
> Making up silly numbers is no use to anyone! That was the point I was
> trying to make. Of course one finds silly numbers that don't agree
> with a prior belief "less convincing" than ones that do. If anyone has
> more ideas about how to collect interesting data, please share them.
I think that approximating "hours spent" with "bytes uploaded to archive"
is about the closest one could ever get -- and even that is a gross
approximation.
> >> No. I say let the "bazaar" decide.
> > You mean, instead of voting on it?
>
> No.
Then that's an unfair statement -- because you're advocating replacing
"letting the bazaar decide" with "let's vote on it."
> >> I'm not sure why it's so hard to understand that it's
> >> unreasonable to ask people who disagree with the infrastructure
> >> to set it up.
>
> > How is that more unreasonable than asking people who agree with
> > the infrastructure to dismantle it?
>
> Possibly only marginally, because it is asked for after the decision,
> rather than before. It looks like it is less work than the creation
> request, too, and has possible benefits for non-DDs.
I don't understand what you're saying, here.
--
Raul
Reply to: