Am Montag, den 05.06.2006, 21:16 -0500 schrieb Josh Boyer: > On Mon, 2006-06-05 at 20:27 -0400, Max Spevack wrote: > > On Mon, 5 Jun 2006, Josh Boyer wrote: > > >> * Extras sponsorship > > >> - Is this process smooth? I heard some grumblings at the Summit > > >> - What do we need to fix? > > > For those of use that weren't at the Summit, could you expand on what > > > these grumblings were please? > > Just that people can't get their packages sponsored, and things sit for a > > long time without any sort of review (either acceptance or rejection). > Ok. I've heard those grumblings. Well, reviewers are limited and sponsors, too. [...] > > FESCO folks on this list: are there issues here? what, if anything needs > > to be done? > There are various ideas kicking around within FESCO. Review days, > writing a tool to check for most of the "Must"s in the review list, etc. Yes, these are ideas. "More sponsors" is also an idea, but we can't make everyone a sponsor. "Better Documentation and Education" is also on the list -- warren and tibbs are working on that currently. > Not that anyone would mind fresh ideas/insight on the problem, but what > is the board going to do here that FESCO can't? In other words... is > the board meeting the right place to be having this discussion? There is one thing I have in mind where the Board maybe can help a bit because it probably needs a lot of help from the infrastructure side. I got the impression that at least some Sponsors fear to sponsor people because those can do a lot of damage and it is hard to watch and control there doings. That's why I'd like to propose roughly something like this (not completely worked out yet, just ideas ATM): - Enable ACLs in CVS so that new contributors only have access to their package(s) - Create a new group in the accounts-system named "cvsextras-limited" (or something like that). Members of that group only have access to some packages in CVS but no permissions to requests builds in the build system. This would allow new contributors to start with co-maintaining existing packages. The real maintainer still can check the contributions easily and only has to request the build when everything is fine. The real maintainer gets rid of some of his maintainer-work this way and might have time to invest this in reviewing and sponsoring. And if the new co-maintainer did its job fine for some months he's sponsored to be a full contributor. (Note: without ACLs in the Buildsystem this is IMHO to risky, especially due to the CTRL-C problem where people might suppress the changelog-mails to commits-list) - Send direct mails to the Sponsors and the main-package Maintainer when a new contributor did something (uploaded a new revision to cvs, request a build). Yes, we have extras-commits-list, but stuff gets lost in the noise there easily -- sending direct mails to those that are responsible for the new contributor and the package is IMHO a lot better. CU thl