Historical bug tracker for the Pacman package manager.
The pacman bug tracker has moved to gitlab:
https://gitlab.archlinux.org/pacman/pacman/-/issues
This tracker remains open for interaction with historical bugs during the transition period. Any new bugs reports will be closed without further action.
The pacman bug tracker has moved to gitlab:
https://gitlab.archlinux.org/pacman/pacman/-/issues
This tracker remains open for interaction with historical bugs during the transition period. Any new bugs reports will be closed without further action.
FS#8496 - Foreign finding (-Qm) ignore package versions
Attached to Project:
Pacman
Opened by Artyom (urandom) - Saturday, 03 November 2007, 17:13 GMT
Last edited by Dan McGee (toofishes) - Thursday, 24 July 2008, 03:23 GMT
Opened by Artyom (urandom) - Saturday, 03 November 2007, 17:13 GMT
Last edited by Dan McGee (toofishes) - Thursday, 24 July 2008, 03:23 GMT
|
DetailsExample: I have package kernel26 2.6.23.1-6
[ 07-11-03 20:04 ] [ artem@m16 ] [ /var/lib/pacman/local ] % pacman -Q kernel26 kernel26 2.6.23.1-6 All internet repositories with kernel26-2.6.23.1-6 was turned off: [ 07-11-03 20:04 ] [ artem@m16 ] [ /var/lib/pacman/local ] % pacman -Ss "^kernel26$" current-cd/kernel26 2.6.22.9-1 (base) The Linux Kernel and modules pacman-cached/kernel26 2.6.23.1-4 (base) The Linux Kernel and modules But foreign list don't contain kernel26-2.6.23.1-6 (and many other packages from inet repositories) [ 07-11-03 20:04 ] [ artem@m16 ] [ /var/lib/pacman/local ] % pacman -Qm flamerobin 0.8.1-1 gmail-notify 1.6.1-5 rapidsvn-svn 7686-1 sonata-svn 578-1 wxformbuilder-svn 3.x-1 |
This task depends upon
Closed by Dan McGee (toofishes)
Thursday, 24 July 2008, 03:23 GMT
Reason for closing: Won't implement
Additional comments about closing: See comments.
Thursday, 24 July 2008, 03:23 GMT
Reason for closing: Won't implement
Additional comments about closing: See comments.
What this feature request proposes is that we add version to the comparison operation.
Artyom's example may seem easy because it seems logical that when kernel26 2.6.23.1-6 is newer than any of other kernel26 packages - it is foreign (either compiled by user of installed from some other repo) - and such version check ("when local package is newer than any others - it's foreigh") could seem like a good idea.
(Artyom didn't say about this but I assume his idea was exactly this, because another variant when versioned checks are regardless of which package is newer is just plain stupid).
But there are not so easy cases:
1) how such modified -Qm would know that kernel26 2.6.23.1-6 was not installed from [current-cd] and then pacman26 in that repo was downgraded to kernel26 2.6.22.9-1?
2) confusing versioning issue comes here again (e.g. which is newer - 3.0b or 3.0? b could mean beta or update) - they are solved on -Su by specifying options=('force') in PKGVUILD, but how to handle them in -Qm?
So IMO changing -Qm behaviour will produce confusing results.