Le Tue, Aug 27, 2002 at 09:55:45AM +0200, Stefano Zacchiroli écrivait: > IMO a BTS based approach is a nice idea expecially for the availability > of history (usefull for the AM) and (even if less important) because > it's more 'official' than the current CGI approach to my eyes. > > Anyway pollute the BTS with sponsorship request/response is a bit bad, > BTS is a bug tracking system, and I hard see request/response cycles fit > well in it: it's an hack (I know this can be a good point to use the BTS > for us :-)
Why is it bad ? It does fullfull our needs ... you recognize it yourself. That's enough to start using it. It doesn't really pollute the BTS since all bugs only appears in the "sponsorship" package. Concerning RC bug list, well we'd have no release critical bug. And even if we have, the scripts already have list of exclusion to ignore some packages (like wnpp) and we can easily add the sponsorship package to the list of exclusions ... > Anyway we have the BTS sources and engine up and running is pretty easy > to set up another one with almost no pain [1] that we can use only for > sponsorship. Just put it somewhere like sponsor.debian.org. This just means that you never installed a BTS yourself ... :-) It certainly isn't that easy. And a full new BTS just for a single pseudo-package entry is just ridiculous. We already have pseudo-packages used for things that are no bugs. I don't see how a sponsor-request in the "sponsorship" package is worst than a list-request in the "lists.debian.org" package. > [1] I know few of the BTS but I suppose sources are in the debian cvs > and that the handling is cron based. I volunteer to set up > sponsor.debian.org if we reach such a consensus, even if I suppose I > don't old enough unix permissions to do that ... It's kind to volunteer. But you can really use your time more efficiently for debian... Cheers, -- Raphaël Hertzog -+- http://strasbourg.linuxfr.org/~raphael/ Formation Linux et logiciel libre : http://www.logidee.com
