On Tue, 14 Feb 2017 21:08:56 +0100 Paul Gevers <elb...@debian.org> wrote: > On Sat, 2 May 2015 09:19:21 +0200 Raphael Hertzog <hert...@debian.org> > wrote: > > Given your answer about "embed-weakdep", I think that it would make sense > > to fix "replace" to behave like the documentation says and to introduce > > the other variant of embed. > > I would indeed like an embeb-weakdep version. Symlinking to the top > level doesn't work for me, because I need to embed multiple packages in > the same dir. And I really want to strip all the embedded files from my > tar ball because I needed to strip anyways and this way I can avoid the > need to write and keep up-to-date the copyright file.
Ping... I now have an RC bug against my package cacti because the depends got updated (and nothing except a manual bug told me what was going on). autopkgtests started failed, but didn't notified me that the versions added by dh_linktree were the problem. To me it really looks like the behavior of embed is often too strict (it is for me at least). Obviously I can request an binNMU, but really, is that needed just because of dh_linktree behavior? Paul
signature.asc
Description: OpenPGP digital signature
