[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Question to all candidates: financing of development

On Sun, Mar 14, 2010 at 05:01, Stefano Zacchiroli <zack@debian.org> wrote:
> Some DDs are able to pursue specific Debian projects due to bounties
> they put on the projects (both AJ and Raphael have similar initiatives
> on their homepages, even though I don't know how much they are
> successful in terms of "customers").

Unless you mean some other AJ, that's not right. I mean, technically I
guess I haven't removed the page for that I had back in 2005, but I
don't think it's linked from anywhere anymore except maybe old blog
posts; Google doesn't see any links anyway. I think I got about $100
out of bounties all up.

For anyone who cares, I blogged about the concept at:


And the stuff that got done with that was:

    (BTS usertags, usercategories)

    (apt pdiffs)

    (debootstrap miscellania)

    (testing scripts memory management)

It was fun and educational, but covered about a month's worth of
broadband. OTOH, I was lucky enough to be able to get a couple of
ideas directly funded at a more useful level (about $2000 AUD from
Andrew Pollock and about $3000 USD from HP iirc).

      -- support for "unembargoed" uploads

      -- support for the mirror split and inclusion of amd64

I did get flamed for accepting money to work on the latter project
while at the same time being an ftpmaster. But then I also got flamed
for the AJ market thing.

> A bit of history. IIRC, the Dunc-Tank affair has gone through two
> consequent problems. The first one was the proposal to use Debian money
> to pay DDs. That proposal was taken back, since it was obvious that most
> DDs were against.

For the record, it was taken back because /some/ DDs were /strongly/
against it. At that point there hadn't been a formal poll, and far
more people had posted in support than against. It's been the
appropriate number of years, so the thread could even be declassified
now if someone wanted to go to the effort...

The eventual dunc-tank implementation had a few ballots on it; there
was the "we recall the DPL (in order to disassociate ourselves with
it)" one [0], which failed by 277 votes to 48; there was the
"reaffirms support for the DPL; dunc-tank isn't a Debian project; wish
success to projects funding Debian or helping the release of etch" one
[1] which succeeded by 227 votes to 93 (and was preferred to the
proposed amendment "reaffirm support for the DPL; but not endorse or
support his other projects" by 177 to 128 voters). Another set of
resolutions were proposed to explicitly endorse dunc-tank or to tell
the RMs to decline payment and donors not to donate [2] didn't receive
sufficient seconds to be voted on.

Personally, I guess I'm more surprised people are still inclined to
raise the issue -- there never used to be that many people looking
into funding DDs, and unlike back then, it's now obvious that there's
not an insignificant amount of opposition to deal with if you are
interested in trying something out.

> (For full disclosure and as an additional note: back then in Dunc-Tank I
>  was not against external founding. However, the fact that it was
>  _still_ that much controversial and flame-prone is enough of a reason,
>  for me as potential DPL, to discourage any DDs/DMs for attempting it
>  again. The benefits of the founding can be totally overtaken by the
>  disadvantages of troubles created in the community.)

One of the challenges of being DPL is working out when to let a few
violently opposed people block projects and ideas being worked on, and
when (and how) to put up with the flack, deal with their concerns and
objections and continue anyway. Anything the leader tries to do will
fall into one of two camps: no one will care, or someone will be
opposed and try to make life painful for the people trying to make it

Kicking Jonathan/Ted Walther out of the project, adding GSoC, letting
Java into non-free, getting a decision process for debconf 7's
location that wasn't just a cabal picking their favourite [3], and
polling users regarding the firmware GR, amongst other things all
required going through a varying degrees of pain for issues that are
now relatively comfortable and settled.

It's entirely possible to avoid flames and angst and still get things
done by simply disempowering everyone who disagrees with you -- that's
not necessarily even bad: they can always fork the software, or go to
Ubuntu or whatever afterall -- but, personally, I think there's a lot
of value you can pull out of passionate criticism, and that that's one
of the things that made Debian as technically excellent as it is.

Obviously YMMV.


[0] http://www.debian.org/vote/2006/vote_005

[1] http://www.debian.org/vote/2006/vote_006

[2] http://lists.debian.org/debian-vote/2006/09/msg00446.html

[3] And from what I can tell, the process we ended up using for
DebConf 7 that resulted in the Bosnia team losing, this year resulted
in a clear win for an improved Bosnian bid for DebConf 11, even
against what seemed like a pretty fantastic bid from the German cabal.
I think this demonstrates that sometimes putting up with a bunch of
crap in the short term really does result in a way better eventual
outcome. YMMV of course.

Anthony Towns <aj@erisian.com.au>

Reply to: