Mattia Rizzolo <mat...@debian.org> writes: > On Wed, Dec 27, 2017 at 01:35:50AM +0100, Philip Rinn wrote: >> Is it true (and intended) that -guest users can't create projects >> within teams/groups they are member of? Or am I missing something? [I >> was not able to create a project within the Debian Science Team] > > As Anton noted, you need to be a "master" in that group. The difference > between "master" and "developer" in groups is only that masters can > create projects.
There is another important difference: masters can add new team members. Which means that we actually loose any control on the team members if we grant "master" to all. For collab-maint, since all DDs are masters there, anyone can actually decide to grant "master" to any non-DD on the base of *personal* trust (right?). > So yes, you need to get a DD to do the one-off work of creating a > "collab-maint" project and grant you "master" powers in that projects > (so that you can add future members yourself). Now, is that really so > burdersome? For collab-maint, I think this is a problem: we actually want to get people into collab-maint, and so the threshold to do so should be as low as possible. If, as a non-DD, I first have to request the repository creation from someone else, I even may decide to create the repository in my personal space instead. From my pre-DD experiences I remember, that it is quite boring to need to wait for others on every step, and I personally would just skip those steps which are not needed to complete my package. For specialized teams (f.e. Debian Astro, and maybe Debian Science), I however don't think this is a problem: There is usually someone who anyway observes the ITPs for packages that may fit (f.e. I do that for Astro; Andreas is known to answer on potential Science packages). Then one could also just offer to create the repository. Best regards Ole