SoylentNews is people
SoylentNews is powered by your submissions, so send in your scoop. Only 18 submissions in the queue.
SoylentNews
https://lists.debian.org/debian-devel/2014/11/msg00174.html
Joey Hess has apparently left Debian after 18 years, stating that the Debian Constitution is leading Debian in "very unhealthy directions".
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
If you find a solution and become attached to it, the solution may become
your next problem.
(Score: 3, Insightful) by Marand on Saturday November 08 2014, @11:50PM
My summary would be that as a developer he felt the devs & technical committee had put a lot of time and effort into fully evaluating and coming to a technical decision, only to have a general resolution / policy proposed that would override the technical decision (on the basis of the opinion of people who hadn't participated in the original technical discussion) a matter of weeks before a release / freeze, then to see the whole thing descend into arguments over legalese, and then escalate to threats.
Some devs and techies are very intolerant of legalese - I can understand that, if I have to spend more time reading the contract than doing the work, I conclude you probably don't want me to do the work, and if you threaten me then it's definite. If I wanted to spend all my time on legal aspects of corporate constitution I would have been a lawyer.
After reading the links in your comment and the ones in fritsd's post, the impression I get is that, at a basic level, this is essentially "the developers have spoken, and you are not developers, so you should do as we say. How dare you question our judgment?!" Same sort of thing that has been happening in GNOME since forever, except that Debian (perhaps unintentionally) provided a way to fight the decision and now people are pissed about that and claiming it isn't what it was intended to be used for.
Thing is, everything I remember of that TC decision looked like some kind of surreal joke. The TC itself was abruptly proposed to bypass a more tempered (and more Debian-like) approach to choosing which init and when to support that was already in the process of being set up, for one. Then, because of how the vote itself was set up, the split vote ended up resulting in a massive change instead of a decision to wait. One would think that the init should be important enough to not be changed over a split decision broken by a single vote. The fact that the TC itself couldn't reach anything even close to consensus should have been a clear indication that the decision needed to be postponed until the next release, because none of the proposed replacements were a clear enough improvement.
Users of distributions need to think carefully - if your distribution takes a technical direction you don't like, then there are others, you can find a new distro. Or you could decide to call your developers names, and collectively try and override their technical decisions. You could show less respect than I've seen from any commercial software boss who is actually writing paychecks for the devs. Then you don't have to find a new distro, instead you have to find a new set of devs, to work for you, for free. Good luck.
I wonder if the entire tone and reaction would have been milder if the abruptly called vote hadn't happened, and the original approach had instead been taken. There still would have been vitriol on all sides, but the GR itself probably never would have happened, since there wouldn't be a TC vote to need overriding.
Personally, I appreciate the work they do and normally respect their decisions even when I don't agree with their changes, but the way this one has been handled on both sides just stinks. It was a pissing match between the systemd and upstart proponents, and anybody that suggested anything else basically got told to butt out. I don't love systemd, but my problem wasn't that it won, it was how it won. I would have felt the same about Upstart or even OpenRC (which I have no issues with) winning.
As for losing devs over it, it sucks, but maybe this is the kind of shake-up Debian needs. This whole thing has brought to light various types of agenda-based policy abuse that never really happened before. Perhaps policy will change/improve and Debian will end up stronger for it.
(Score: 2) by Marand on Sunday November 09 2014, @12:27AM
Addition to comment based on something I found:
This post [debian.org] by him seems to indicate he was bothered by the TC voting farce, too. So, I'm not sure if the combination finally pushed him over, or if he's backing the decision even though he disagrees with it because he feels it shouldn't be contradicted, and thus angered by the GR fighting it.
(Score: 2) by choose another one on Sunday November 09 2014, @09:20PM
With respect, I think you are reading it wrong. His previous posts indicate that he backed and respected the previous TC decision process and felt it was inappropriate for someone to try and overrule all the effort that went into that by appealing to a different forum, and doing it 2 weeks before release freeze. He also believed it was an abuse of process:
https://lists.debian.org/debian-vote/2014/10/msg00246.html [debian.org]
I don't know what his position is on systemd itself, and I'm not sure it matters. I have had stand up rows with PHBs for trying to micromanage and override technical decisions at incredibly stupid points in the development cycle, after they had not wanted to be involved in all the technical discussions leading to the decision. It wouldn't have mattered which side of the decision I was on, if I even took a side - I would back my technical team's right to make their technical decisions at the right point in the development cycle and have them respected. In exactly the same way, I would _not_ (and would _not_ back a developer who wanted to) go to the PHB a day before quarter accounts deadline to suggest that the revenue recognition model is wrong (whether I thought it was or not).
The comment you have linked is not at all about the TC default-systemd vote, it is about more recent discussions on how the init system should be selected on upgrade. His point seems to be that this was all already resolved by consensus with all the relevant init package maintainers, but then someone stepped in and made it political, together with statements that he (see https://lists.debian.org/debian-ctte/2014/11/msg00047.html) [debian.org] interpreted as overriding the current work on upgrade path. I don't think it is a coincidence that the same person was also behind the GR, and consider it entirely possible that breakdown in relations between those two people has lead to this resignation. Interestingly it appears to be not the only recent TC resignation...
(Score: 2) by Marand on Sunday November 09 2014, @11:56PM
You're right, I missed that it was a follow-up TC vote instead of discussion about the original one. That puts things back where my longer post left them: the whole thing still stinks.
I can understand him (and others) being pissed that people want to undermine the TC decision, but I also still think that the way the vote was handled in the first place fueled that fire. This was an odd thing for Debian: a distro that cautiously approaches sweeping changes has jumped head first into this one, despite problems, lack of maturity in the solutions, and a lack of consensus in the TC.
It's also looking like either Debian's policies or its people aren't holding up well when faced with contentious decisions. Hopefully the end result is that things end up better, with a stronger Debian. If not, though, there will still be other options. Not that I wish ill on the project or its members. Not at all, in fact. I've generally had good interactions with Debian and KDE folk (which is one of many reasons I still use Debian+KDE), so it's kind of a shame that this crap is happening. Still, maybe a shake-up will be good in the long-term, either for Debian or for something new that may come from it.