Hi, Michael Biebl wrote: > > I somehow doubt that. As I see it, this either this broke only very > > recently (after the 0.8.3 upload) or it's only happening under some > > circumstances. > > Just curious, can you reproduce the issue with the steps I outlined?
Haven't tried it yet admittedly. > From what I can see, once /var/lib/aptitude/pkgstates exists, there is a > complete disconnect between the autobit state of aptitude and apt. Not in my case. A Sid amd64 desktop machine with close to 10000 packages installed show nearly no difference between apt's and aptitude's view on the autoinstalled flag: # colordiff -u <(aptitude search '~M' | awk '{print $3}') <(apt-mark showauto) --- /proc/self/fd/11 2016-10-20 02:17:28.622307073 +0200 +++ /proc/self/fd/12 2016-10-20 02:17:28.622307073 +0200 @@ -7254,7 +7254,7 @@ texlive-htmlxml texlive-lang-african texlive-lang-all -- +texlive-lang-arabic texlive-lang-chinese texlive-lang-cjk texlive-lang-cyrillic According to your description, I'd expect tons of differences. > The first time /var/lib/aptitude/pkgstates is created, it inherits the > autobit state from the apt db, but after that, any changes apt makes are > no longer applied to aptitude. > > Is this not happening for you? Seems not, no. But then again, I'm using aptitude most of the time and apt only in a few percent of all cases. Regards, Axel -- ,''`. | Axel Beckert <a...@debian.org>, http://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