My commitment in Debian, dunc-tank, and related matters…

I've chosen debian because it fulfills an ideal. For me Debian is about to build an ”Universal OS”, and when decisions has to be taken, it's based:

• on the technical quality for the technical ones,
• the consensus of the community for the others.

To live decently, I do work also in a start-up company that also happen to write closed-source software. That gives me enough money to eat and live well, and beeing able to dedicate my free time on Debian and free software. That means I've done the choice on purpose not to mix my paid work with Debian, because I don't want to balance the decisions I take in Debian I against my well-beeing.

That said, I'm completely OK with people having paid works intimately related to their DD work. And if their work is intimately related to etch beeing released in time, then well, I wish them a lot of courrage. I wouldn't have the guts. I'm OK with this because we are talking about people that are not able to impose decisions to me that I judge irealist or bad. And still, I can dismiss such decision thanks to the CTTE if needed.

Though, when it comes to delegates, it's a full different story, because I do have to interact with those guys in my allday Debian work, and that undoing their decisions needs a GR as per constitution. It's way less easy than going in front of the CTTE. And if Delegates are paid whichever the entity that pays them is, they will have to balance technical choices against beeing able to eat tonight (or reimburse their house, or …).

Obviously, one will say that it's OK here, it's not a salary, and that the currently discussed^[1] contract is not about doing the RM releasing Debian in time, but helping toward that goal, or trying to release in time not sacrificing the release quality. I answer “bullshit!” to that. Our RM have — at least I happen to hope so — more higher standards than that. If they accept that mission it's to see it completed. I'm already fearing for them getting down if they can't achieve that.

I also know that at least one of the dunc-tank board member^[2] would like to see a structure able to fund DD's to work on Debian, and be able to live from that. I'm sorry, but that's an option I don't want the project to take. And if dunc-tank is a prelude to that, so please don't even think of starting it. When there is money in the game, the {donator,employer,boss,…} and the guy accepting that money have at least a moral contract — if not a legally binding one — of results. And I see no investor that would happen to be interessed into a contract where the due in line is undefined. Thinking otherwise looks like complete naivety to me. So either it won't work, either people will have to take into account other parameters than technical quality when the will have to make choices. I see in my paid work what such things mean, and I do consider me lucky, I'm in a company where people do know what quality means. Though the company also has to live, and that sometimes makes us take choices that I don't like at all. I never want have to even consider such choices in Debian.

I'm sorry, but putting such considerations in the project is jeopardizing what the project means and represent to me. The sole structure that would maybe fit would be some bounty board, with the following things done:

• any person (at least for DD's) can come with his project, with absolutely no censorship;
• the proposer that comes with his idea chooses the price for it, with again no censorship;
• anyone is free to do the job for free, obsoleting the bounty^[3], so that a bounty cannot become blocking.

And even with that some kind of checking that the job has been done correctly has to be done somewhere, and I don't see clearly who could do that, but I suppose that's doable to find some fair reviewers plus the donator(s) for that.

That said dunc-tank does not meet any of the previous requirements, is driven by delegates, and works around the thorny problem of finding a consensus among the DD's in the worst possible way.

Those are the reasons^[4] why I think dunc-tank is a bad idea.

And to the people offended that I seconded the recall, this is about the fact that I found the way dunc-tank has been imposed to us — after having asked for our opinion, and heard the acute opposition of some of us — hideous, and about the fact that beeing the DPL, or even a delegate and in dunc-tank at the same time is a conflict of interests. It's not a matter of beeing pro or against dunc-tank, matter against which there is no course of action for me anyway, as it has been on purpose and artificially kept outside of Debian.