I will split my thoughts into two parts: this is part 1

The power of anarchy. It seems that the Gentoo project has no goals. Developers within Gentoo have goals; I have goals, particular projects of Gentoo have goals. But the project itself has no goals. It is a project born out of fun, of pulling new packages and building a system to meet ones needs. It is a project born out of a boon of development, where the few became many, and now the many become few (or soon anyway).

Gentoo is bleeding experienced devs; that is to be expected after a while. People can’t work on this stuff forever, other interests show up, paths change, people change. But I myself am frustrated. Because Gentoo is a project that grew too fast, and the people that came in essentially forced out (for lack of a better phrase) the only leader Gentoo ever had. The political landscape of Gentoo is grim, in my eyes. People compete for power (perceived or otherwise). Now I’m generalizing here, when I say ‘people’ I mean the promenant members of the community.

There are probably 70-80 devs that just maintain their stuff and don’t care about all this bullcrap, and they don’t read my blog, and they will probably enjoy Gentoo no more or no less, as long as they can maintain their stuff, or their port, or whatever.

I’m referring to those with an agenda. Just to be clear; I too have an agenda, so some of this article is hypocritical.

Part of the problem is that Gentoo doesn’t have goals. It used to, and it had a set of people who enforced them (more or less). They were not well defined but they existed and to some extent they were met. Gentoo grew up, Gentoo got more devs (for better or worse) and Gentoo still lacks goals. The double bonus being the more specific goals you have the less chance you have of everyone agreeing with you (since there is a good chance not everyone will agree with you). There is also the annoying fact that any complicated or costly goal is shot down.

Take for example Userrel. Gentoo User Relations, which at their last meeting came up with a plan to try and alert users of problem areas (gcc update, glibc, xmms removal, expat see below post on upgrades). This requires a team of a few developers to keep on top of what is going on *before it hits stable*. For things like eclasses and package maskings, this isn’t even possible (as they aren’t announced a priori, not even in treecleaners).

Now you can say ‘but they should try anyway, hit the cases they can. Trying is worth it even if you know you will fail (aka can’t cover everything).’ And I’d agree with that statement. But at the same time it is sad to have a goal they can never reach. Why can’t they reach it? Policy, and enforcement, training, Gentoo Goals.

Gentoo is not about users for most people (myself included, oh the irony), Gentoo is about themselves. So how do you enact a Distribution wide goal without the support of all parties? This is the core problem for many other projects as well.

QA? Some devs don’t give a rats ass, some devs don’t even use repoman. Some teams *cough games team cough* break QA policy fairly often. It creates a double standard where everyone has to spend time following QA except the games team because they are just ‘that good’. Now to be fair, they are pretty good at their job, but they are human too. They screw up just as every other dev does. Thats why the tools and policy exist.
Gentoo was not formed with goals of QA or the Users in mind, so it is difficult to add these things as distro-wide goals.