Re: [gentoo-dev] Bugzilla handling for maintainer-wanted things

[Nathan L. Adams]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ciaran McCreesh wrote:
> Oh come on, haven't you heard my rants about the state of the tree and
> the number of monkeys who have commit access?

Yes I've read those rants, among others.. :)

But with everyone screaming 'not enough manpower' the number of devs
with commit access is just bound to increase. So why not focus on how to
increase quality by default?

> Problem is, getting decent
> QA done once things hit the tree is in many cases very difficult

So why not build peer review into the process/policy? Require that the
team leads (who could deligate as they see fit) perform verification
(peer review) before closing out bugs.

> -- the
> kind of people who won't accept QA feedback are usually the kind who
> are making the worst mistakes. The maintainer-wanted list is simply an
> easier target...

True; being a premadonna isn't pretty or helpful to the project, but I
bet alot of it is due to bad expectations. There seems to be a vocal
minority of devs who equate being a dev with a God-like status. "How
dare you question me or my work?!?" And it would make sense that the new
devs would pick up on that as a way to 'fit in'. So how can we set
better expectations for new devs up front? Update the dev policy docs?:

- - Expect to have your work peer reviewed at all times
- - Realise that peer reviews are intended to improve the code not
evaluate dev performance

Ideas?

Nathan



-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQFDBe572QTTR4CNEQARAtYpAJ0aZ4gnfyE4lTUrbYr/DcWmIUX67ACghyvl
TTCM9mWVTkuUWm33WnSeE9A=
=gE1q
-----END PGP SIGNATURE-----
-- 
gentoo-dev@g.o mailing list