Re: Censorship in Debian
On 1/7/19 9:12 PM, Russ Allbery wrote:
Miles Fidelman <mfidelman@meetinghouse.net> writes:
Well, first off, the process led to the resignation of the chair of the
Technical Committee - out of a feeling that the process had become too
"personalized."
Some decisions are just hard. I think nearly all of us involved in making
that decision burned out in various ways. I'm not saying we couldn't have
done a better job... well, hm. Actually, I am kind of saying that, if by
"couldn't" I include the people that we were at the time with the
emotional reserves that we had and the understanding that we had.
I could certainly do a better job *now* if I could rewind time, but that's
cheating, and humans don't get to do that.
We do get to learn from things, however.
I'm with Steve in that I'm pretty dubious that the process was the core of
why that decision was so hard. I think it was so hard because it spanned
the gamut from technical to social issues, involved some issues that were
relatively concrete and others that were quite nebulous (such as the
interactions between the goals of the systemd developers and the broader
community), and also involved deep social divisions in the project between
folks who want Debian to be a platform for all things and folks who want
Debian to be more tightly integrated and more technically excellent along
a single axis.
Well, I'd argue that part of it had to do with who had a voice, and who
didn't. (More below.)
This stuff is inherently very hard, particularly when friends end up on
opposite sides and believe passionately in how important their concerns
are.
I think we sometimes analyze process to death and refight the last
fourteen wars and dig up problems to argue about them some more. We're
human, this stuff is hard, some things are going to be brutal to get
through when we disagree, and it's okay to forgive ourselves for not being
perfect. Or even being pretty shitty at it.
That's not to say that we shouldn't look for opportunities to fix things
that we can. For example, we certainly uncovered some nasty edge cases in
the voting mechanism for the TC, which are now fixed. And many of us felt
that people serving for extended periods of time on the TC wasn't socially
healthy for either us or the project, so we fixed that too.
But I think there's a idealistic, utopian tendency among a lot of
technical people, myself included, to believe that any serious conflict or
(from our perspective) incorrect decision is a bug in a process somewhere,
and if we can just find the right process, we can fix the bugs. And it's
just not true. Humans are messy and humans disagree, and sometimes stuff
is just really hard, and is going to be really hard no matter how you do
it.
Beyond that, there are a rather large number of folks, impacted by the
decision, who did not have a seat at the table. Those of us who rely on
Debian in production, for example. Upstream developers for another.
Some of us knew about the issues & debates, without having a
"franchise," others found out after the fact. Seems to me that lack of
representation is, in itself, a rather big failure of governance.
Debian is *more* willing to try to take into account the needs of its
users than most free software projects, but Debian is still a volunteer
free software project, and the rule of just about every volunteer,
unfunded free software project is that the people who are doing the work
are the ones who are going to make the decisions.
Think of it this way: the people who are sufficiently invested in the
project to spend our time and energy on it over a long enough period of
time to become members are deeply invested in it and want to control where
it goes. Plus, we're all volunteers and don't have to work on anything we
don't want to work on, which means maintaining our engagement is
absolutely necessary for the project to survive.
I think you're minimizing the level of investment & commitment it takes
to either use Debian, particularly in production, and even more,
minimizing the efforts of upstream, and kernel, developers upon whom
Debian ultimately depends. There are also those who contribute by
providing support - e.g., answering user questions on Debian lists.
As far as I can tell, the only people who count, in Debian decision
making, are packagers - which strikes me as a rather bizarre case of the
tail wagging the dog. I remain amazed how much the impacts on users,
systems administrators, and upstream developers were dismissed as
irrelevant.
On a larger note, I point to the IETF as an example of a much larger
community, running huge infrastructure, where pretty much anyone who
shows up has a voice.
I understand your desire to have a say in something that's important to
you, but, well, if it's that important to you, the New Maintainer process
is right over there? We always need more help. Absent that, the people
who have put their blood, sweat, and tears into the project are the people
who are going to make the decisions, of course with an eye to our project
agreement to try to prioritize our users. But it's going to be our
interpretation of what's good for our users.
If people who weren't doing the work, who weren't part of that community
and taking part of the load, were calling the shots and the rest of us had
to obey them, well... we have a word for that. It's called a job, and the
entire relationship with the work is different, and I would expect to be
paid.
If it helps, think of having a voice in the decisions to be the pay that
we get for working on Debian.
It might, however, have led to the Technical Committee giving more
weight to the impacts of the decisions.
I always have to laugh at statements like that, since I think they come
from a well-meaning place of almost total lack of understanding of what it
was like to be on the TC during that decision.
I think I put more thought into all of the aspects of that decision,
including weight on the impacts of the decisions, than just about any
other decision I've made in my life. I have put less thought into where I
live than into systemd.
I think this is part of that all-too-human belief that one's own position
is so obviously correct that if anyone disagrees with you, it's just
because they've not thought about the problem hard enough. And it's just
not true.
I'm sorry to say this, but the only value that Debian provides to the
world, is packaging. And, personally, over time, I've found it more and
more necessary to download, build, and compile from source - reducing
the value of Debian.
Pretty soon, I expect I'll be migrating.
And, next time I do any serious developing, I expect the only init
scripts I'll provide are sysvinit based. That suggests that my platform
will be something other than Debian.
Miles Fidelman
--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra
Reply to: