Hi Michael, Michael Meskes wrote: > I'm with Daniel on this one as I cannot reproduce it either: > > Preparing to unpack .../webext-browserpass_3.7.2-1+b1_amd64.deb ... > Unpacking webext-browserpass (3.7.2-1+b1) over (2.0.22-2) ...
Indeed. Using a clean Sid chroot, installing webext-browserpass from Buster and then upgrading does not exhibit this issue. Nevertheless IIRC I ran into it when upgrading a (production) Thinkpad from Buster to Bullseye, i.e. no other versions of webext-browserpass than those from Buster and Bullseye were involved. There were though a quite few other webext-* packages involved on that dist-upgrade. Those are now installed now on bullseye and contain the common directories involved in that symlink/directory switch: # dpkg -S /usr/share/chromium/extensions /usr/share/mozilla/extensions webext-browserpass, webext-ublock-origin-chromium, webext-bulk-media-downloader, webext-privacy-badger, webext-https-everywhere: /usr/share/chromium/extensions webext-browserpass, webext-ublock-origin-firefox, webext-treestyletab, webext-bulk-media-downloader, webext-form-history-control, webext-privacy-badger, webext-https-everywhere, webext-noscript, webext-debianbuttons, firefox-esr, webext-umatrix: /usr/share/mozilla/extensions > Something fishy is going on here. I'm not sure how to find out what > though if I cannot reproduce it. Also I wonder if removing the > package from testing is helpful or even correct in such a case. Good question. My gut feeling at least says that the RC severity is justified as quite some people ran into it and it actually causes apt to abort in a quite nasty way. > Anyway, any idea how to find out what's going on and what is > different on your systems? Currently not, unfortunately, as I don't have further ideas where to look for. I currently suspect a relation to respectively overlap with a similar symlink/directory switch of maybe one of the directories mentioned above. > For instance I tried on a sid system where I install the old > browserpass package. (It also seems important to not just remove but really purge the current package in case it was installed befotrehand. But I assume you either did that or used a fresh install.) > Did everyone with the error see it on a dist-upgrade only? In my case yes. I have it installed on multiple sid boxes and it didn't occur on any of them (which are upgraded in a rolling release fashion). Will soonish upgrade another productive Buster desktop to Bullseye where webext-browserpass is installed. Will have a close eye on the moment when upgrading webext-browserpass respectively will upgrade that package in a separate package upgrade from the remainder. Regards, Axel -- ,''`. | Axel Beckert <a...@debian.org>, https://people.debian.org/~abe/ : :' : | Debian Developer, ftp.ch.debian.org Admin `. `' | 4096R: 2517 B724 C5F6 CA99 5329 6E61 2FF9 CD59 6126 16B5 `- | 1024D: F067 EA27 26B9 C3FC 1486 202E C09E 1D89 9593 0EDE