On 19/12/2012 10:03 p.m., Michał Górny wrote:
Doesn't this prove that the recruitment process fails to work?
If I were to throw random ideas, I'd think about letting new recruits
did all commits through a proxy (mentor?). Of course, it all would be
easier if we used git.
I know this side question of "git" migration is one we want to avoid
discussing, I know its in progress.
But I am literally waiting for it to happen, because for whatever
reason, the present barriers to contribution are too high for me without it.
I can't put an exact finger on it, but devs seem to think the quiz
methodology is "easy", but it ( oh, and CVS ) are a high barrier to
entry for me.
I don't have the time/motivation/focus required to commit to even
completing the quizzes, and I don't have the time/motivation/focus
really required to be a "full dev", and I don't even want to be a "Full
dev" really.
But I basically have found every time I've done the quiz, its eventually
boiled down to a cycle of
1. Read quiz
2. Find it hard to find documentation on
3. Search for
4. Get lost
5. Find the resulting information I eventually find is vague and
confusing with regard to the question.
6. Eventually get distracted and do something other than the rest of the
quiz.
I know, it should be easy, and I'm probably making excuses, but it boils
down to
1. People in Gentoo have asked me to/encouraged me to do the quizzes
2. I've tried several times
3. Still not there.
4. This problem is not so prevalent in the dozens of other projects I've
contributed to.
As soon as Git migration is done, then I can just
1. Fork
2. Hack
3. Somebody can watch/review/cherry-pick commits I make if they like
them, if not, I'm not worried.
But the git part aside, back to the quiz.
Surely, I'm not the /only/ person to get roadblocked by the quiz.
The only thing really keeping me around as a half-assed dev is the fact
we have overlays and the fact that the overlays are git based, and I get
/some/ notion of contributions being of value there.
Can we short cut the whole quiz process and have some "Inbound"
repository until we're full git, which people can fork/commit/pull and
trusted people can review submitted branches and apply them to CVS?
Because I feel its quite possible partly that CVS is due to blame ( due
to requiring of trusted commit, which requires the questions ) that
there is difficulty getting devs, and the longer we're stuck with it,
the more it will be a problem.
It could actually be just the Proxy Maintainer workflow is not clear
enough, or simple enough, and that we need more push towards a more
heavy proxy-maintainer based system ( I don't know, I'm ignorant to too
much of proxy-maintainer-ship stuff, to discern /why/ that is might be
difficult, but I'd imagine my ignorance is part of the problem )