Hi, Le dimanche 21 janvier 2024 à 21:38 +0100, Paul Gevers a écrit : > Hi, > > On 21-01-2024 21:06, julien.pu...@gmail.com wrote: > > Would kicking the mathcomp-analysis package out of testing allow > > the > > migration of the rest of the Coq-related packages and at least give > > a > > coherent set there? > > It seems src:mathcomp-analysis is a leaf package that can easily be > removed indeed. >
It is. > > That would still leave mathcomp-analysis broken in > > unstable, of course, but that's a lesser evil. > > Indeed. . > > If the issue isn't cleared for testing by february first, I'll just > > ask for a full RM of mathcomp-analysis : a brutal fix, but an > > efficient one. > > Well, I think that if the package (once fixed) is still useful to > have in Debian, temporary removal from testing is a reasonable trick > to solve situations like this, no need for a full RM. Once fixed it > can migrate again. Obviously if you think the situation for this > package in Debian is bad, than full RM makes sense of course. Well, I have some code depending on it as a user... broken since that long, so yes it would be useful. But if upstream doesn't follow the ecosystem timely, I might have to port my code around it to stay current. As a DD... I'm quite annoyed to see the upstream side of the ecosystem move and the Debian side get frozen for a single package. > I have added a removal hint (for removal from testing). Thanks, that will be a relief to at least save testing from this mess. Thanks again, J.Puydt