Pacman

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.
Tasklist

FS#7355 - [PATCH] Add an ignorearch option to makepkg

Attached to Project: Pacman
Opened by Andreas Hauser (buggs) - Monday, 04 June 2007, 07:56 GMT
Last edited by Dan McGee (toofishes) - Wednesday, 13 June 2007, 05:04 GMT
Task Type Feature Request
Category General
Status Closed
Assigned To Aaron Griffin (phrakture)
Dan McGee (toofishes)
Architecture All
Severity Low
Priority Normal
Reported Version 3.0.4
Due in Version 3.1.0
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Since many PKGBUILD especially on aur miss a complete arch= array, many people have edit the PKGBUILDs and add their arch. Currently this affects mostly x86_64 and ppc users, but as x86_64 gains popularity i686 are increasingly affected.

An untested patch is appended as incentive.
   makepkg-ignorearch.patch (1.3 KiB)
This task depends upon

Closed by  Dan McGee (toofishes)
Wednesday, 13 June 2007, 05:04 GMT
Reason for closing:  Implemented
Additional comments about closing:  Added in GIT, pull it from there if you want it now, otherwise it will be in the 3.1 release.
Comment by Dan McGee (toofishes) - Monday, 04 June 2007, 17:23 GMT
We didn't put in this error message just for kicks, so I doubt this is going to get put in makepkg. Sorry.

How hard is it to add an arch=() line anyway?
Comment by Andreas Hauser (buggs) - Monday, 04 June 2007, 17:57 GMT
There is no reason to make it a fatal one for everyone.

Let me guess you don't use x86_64 with a lot of aur?
I tell you it gets tiresome after 50 packages or so.

If i wanted that kind of bureaucracy i'd went for debian.

The way you do it, the default is "No i don't work on you arch"
while in reality most packages work without a problem.
So the default is not the common case, that's not good!

If you at least used a !$arch (like !x86_64) the work would be less.
Or change it to a warning not aborting the build.
Comment by Aaron Griffin (phrakture) - Monday, 04 June 2007, 19:32 GMT
The default is not "I don't work on you" the default is "this is how it was built, and I know it works" - it is a way of saying "tested, it works"

The problem is that most people only use one architecture at a time.

I see your point though, but I don't know if making it non-fatal is a good idea. However, I do think a command line option may be acceptable, as you can't accidentally type "--ignorearch" - it is fairly explicit.

Side note though, your patch is backwards (your added lines are marked as removals with a minus). Also, the 'if' line would probably be better with:

if [ "$IGNORE_ARCH" = "0" -a ! in_array $CARCH ${arch[@]} ]; then

Could you test that out?
Comment by Andrew Fyfe (space-m0nkey) - Monday, 04 June 2007, 21:10 GMT
I've pushed this patch to my git tree ready for commit (http://repo.or.cz/w/pacman.git?a=commitdiff;h=ef6dab0c2b6e591a74298c8731c33d71a4655859)
Comment by Andreas Hauser (buggs) - Friday, 08 June 2007, 16:24 GMT

Attached is a tested (not reverted ;) patch.
It also changes the behaviour to still emit a warning even with ignore on.
Or in other words, the ignore arch flag changes the error from fatal to warning.

@space-m0nkey i get a 403 forbidden when accessing your git repo via the URL you gave.
   makepkg-ignore-missing-arch.p... (1.7 KiB)
Comment by Andrew Fyfe (space-m0nkey) - Monday, 11 June 2007, 09:00 GMT
@buggs the links break when branches are updated, here's a link to the branch http://repo.or.cz/w/pacman.git?a=shortlog;h=FS7355

I've merged your new patch into the branch.
Comment by Dan McGee (toofishes) - Monday, 11 June 2007, 11:59 GMT
Patch pushed to GIT.

Loading...