On 22.08.19 00:38, Matthias Klumpp wrote: > Am Mi., 21. Aug. 2019 um 15:10 Uhr schrieb Matthias Klose <d...@debian.org>: >> >> On 21.08.19 12:54, Debian Bug Tracking System wrote: >>> Processing commands for cont...@bugs.debian.org: >> >> so it looks like you can work around these by binNMUing the affected >> packages. > > Not really... The dustmite bug (935275) was actually filed because a > binNMU failed. > I did just now build dustmite locally in a sid chroot and everything > went flawlessly without errors - in the container GDC 9 was default > though, which is not affected by this issue. > This highly likely means that we can close the dustmite part of this > issue, because that FTBFS only affected unstable. > > Out of curiosity I built dustmite in a buster chroot as well, and that > went fine (with GDC 8 from buster). > Then I built dub with debspawn and pbuilder in sid and buster chroots > and those are *also* fine now. > Before I could reliably repduce the linker error, and the Debian > buildds were failing at building the software as well. > Yet, I can't actually see anything that has changed between then and now. > > I am very confused now. > I'll attempt a few uploads of the effected packages this weekend (one > has to go to buster to fix the startup problem), and if those build on > the autobuilders, we can definitely close this issue. It's very > strange though that this magically fixed itself. > > Cheers and sorry for the noise,
I now tried to rebuild dustmite as well locally. Works here. Please tell me the dustmite and dub version numbers, then I'll add breaks to the libgphobos package. Matthias
