On Fri, Apr 23, 1999 at 11:17:32PM +0200, Raphael Hertzog wrote: > > Any other objections to the text? > > Yes, the period of 30 days is the time we should wait before an NMU can > be done, but do we still need to ask the maintainer if we can do an NMU > after this period ?
>From what I understood the original text, the meaning is that the QA Group member must inform the maintainer about the bug and offer the fix (if he of course has one). If he gets no response in the specified period, then he has the authority to proceed and upload the fixed package. Of course, with notifying the mailing list etc. > I think that we should better say something like this : > Once a a patch has been submitted to the BTS, the QA member may announce > his intent to NMU the package to the maintainer, and if he gets no > response within 15 days (this does not count if we know that the > maintainer is absent, if so then refer to the instructions he may have > sent to debian-private before leaving) then the QA member is authorized > to do the NMU. Good. I'll rephrase it to: +++ When QA Group members find problems in maintained packages (package that have an active maintainer), they will send fixes to their maintainers through the bug tracking system (BTS). The QA Group member should offer himself for doing the NMU, but he must not instantly do it. See below for exact delays. If in that period the maintainer does not do the upload himself, the QA Group member who submitted the bug report will try to find a reason why the maintainer didn't do it. If the maintainer really appears not able to do the upload himself, the submitter (or for that matter, any other QA Group member) will proceed and do the bugfix upload. +++ I left out the details about the reasons - we can't start naming specific reasons, because it would become a really long list ;> Generally, QA members are interested in a simple answer - will someone do the upload in some reasonable delay, or not. Is this written clear enough now? We obviously need a native English speaker to read the text and find any wrong/awkward/unusual expressions/words. > I personnaly think 15 days is enough. But if the maintainer responds, then > there won't be any problems usually. I'm not really sure... people can get in situations where they can be totally unreachable for a week or so, and if we do the NMU too soon, they may not get the chance to do it themselves because they didn't even get to read the message (when you don't read debian-devel for 10 days you surely won't notice one message among others). We must remember, that this is basically stepping on people's toes as careful as possible :) I say, lets stick to 30 days. Of course, rules for pre-freeze and freeze stay shorter, since these periods are really crucial, and there can be no excuses at that time not to fix bugs. What do you others think? -- enJoy -*/\*- http://jagor.srce.hr/~jrodin/
