On Thu, 2002-08-29 at 13:59, Raphael Hertzog wrote: > You don't know if the current system works well enough (neither do I). > Do you know if all sponsoree find a sponsor ? They don't. I recently though about packaging a program called metacity-setup. There was an open RFP, so I thought no developers were working on it. After a little digging, I found that someone had created packages in July, but nobody responded to his sponsor requests so he just left Debian devel on hold. (He will soon request a sponsor again, as soon as he updates the packages for a new release) > BTW, using the BTS wouldn't change much, we'd still exchange mails in > debian-mentors ... I vote for the BTS too. I think this may also put more discussion into debian-mentors that would otherwise have been done privately between NM and sponsor. Reading through these would be useful for prospective packagers - it may even reduce the workload of developers. One point I must make is that whatever system is chosen, it must be made absolutely clear what the NM should do. Currently, it is too hard to decide whether to mail the list or use the cgi, especially for new developers who are learning lots of new things at that time. One advantage of a cgi approach is that you could have a page telling the users what they should do before requesting a sponsor (e.g. read docs, check package) and also have entries for all the required info (maintainer name, url, description, package url, nm status). This would then be a simpler system for the nm. This could of course also be done within reportbug. -- +----------------------------------------------+ | Mark Howard cam.ac.uk mh344@ | | http://www.tildemh.com tildemh.com mh@ | +----------------------------------------------+
Attachment:
signature.asc
Description: This is a digitally signed message part