[maemo-developers] [RFC] Maemo package guidelines: mandatory categories
From: Marius Vollmer marius.vollmer at nokia.comDate: Thu Apr 17 15:36:15 EEST 2008
- Previous message: [RFC] Maemo package guidelines: mandatory categories
- Next message: [RFC] Maemo package guidelines: mandatory categories
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
"ext Andrew Flegg" <andrew at bleb.org> writes: > On Thu, Apr 17, 2008 at 12:33 PM, Marius Vollmer > >> I am sure you notice the conflict here: whatever list you come up with >> will be unsuitable for someone. You want strict policy enforcement, >> based on community 'feelings'. How can that work? > > Can it be any worse than the mess we're in now? (Yes, I actually think it could be worse, but I think we agree on the important points, so let's not get distracted about this.) > Having said that, perhaps the MOTU-style proposal of gatekeepers doing > QA checks could help here. Deviation is permitted, if it gets through > a gatekeeper: Yes, I agree. I was proposing this, in a more fine fashion: in addition to being able to say who goes in and who doesn't, the gatekeeper could also say: You go in but I am going to change your category to something sensible whether you want to or not. (I.e., the category of a package can be overwritten by the repository.) >> One approach in a situation where consensus is clearly beneficial is to >> make a first shot at a concrete policy that everybody is supposed to >> follow, but make it possible to deviate from that policy in practice. > > That's what we've got now! Yeah, but you cut the important part: At the same time, make it possible for people to improve policy compliance by doing concrete work (i.e., enable them to fix non-compliant things). The gatekeepers would be the ones fixing non-compliance (by rejecting violations, or by correcting them directly).
- Previous message: [RFC] Maemo package guidelines: mandatory categories
- Next message: [RFC] Maemo package guidelines: mandatory categories
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]