Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
In addition to pkgver and pkgrel it is checked if pkgname is defined in the PKGBUILD.
|
|
Don't use globing as this will match a package of the same version but another arch.
We can be more specific here as there are no longer packages without $arch extension.
|
|
Using the hostname to decide which repos to use is not releiable and hard to test.
Instead use config.local to configure these.
config files for sigurd and gerolde were added which can be copied or symlinked to config.local
on the specific host.
|
|
|
|
|
|
|
|
This also checks if the sam package exists within the old package
layout (without package pool)
|
|
See FS#20745
|
|
* When writing the db file ensure that it has write permission of the group which owns the parent directory.
* This should make the adjust-permissions cron job obsolete.
|
|
Don't duplicate the user:group pair all over the place; put all of the
associated repos and directories in the same case statement for a given set
of credentials.
Signed-off-by: Dan McGee <dan@archlinux.org>
Signed-off-by: Pierre Schmitz <pierre@archlinux.de>
|
|
* Repositories can now be defined in the config file for each host
* added community-staging, gnome-unstable and kde-unstable
* Exception is the adjust-permission cron-job; but we might want to use acls in future anyway
Signed-off-by: Pierre Schmitz <pierre@archlinux.de>
|
|
This will really make sure [core] packages flow to the top of the list,
regardless of architecture. They should anyway, but anything that was only
updated in the first architecture updated could get artificially pushed
down before.
Signed-off-by: Dan McGee <dan@archlinux.org>
|
|
Signed-off-by: Dan McGee <dan@archlinux.org>
|
|
Signed-off-by: Dan McGee <dan@archlinux.org>
|
|
This has been untracked by version control for a while on gerolde in both
the cron-jobs directory and in /etc/cron.hourly/. Add it here so we can make
changes and know what is going on.
This is an improved script over what we currently have. It is one script
instead of two, and it does things a little smarter with the logging.
Signed-off-by: Dan McGee <dan@archlinux.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* runtime reduced to < 1 second
* uses file lists and the comm command
* searches for missing packages
* removes empty legacy directories like extra/os/any
|
|
* This should fix the cleanup of packages that were not within the pacakge pool before being updated.
* a test case for this was added
|
|
If reading from a PKGBUILD fails stop immediatly. Also put out more usefull error messages.
|
|
Signed-off-by: Allan McRae <allan@archlinux.org>
Signed-off-by: Pierre Schmitz <pierre@archlinux.de>
|
|
Check if there are packages to update for given architecure. Previously
db-update quit when only one arch of a package was available.
|
|
|
|
|
|
|
|
|
|
|
|
db-move does no longer need a specific architecture. It will move all architecures of a given package
at once.
testing2x has been rewritten to respect these changes and testing2x64 is no longer needed.
|
|
|
|
|
|
|
|
|
|
|
|
* db-update now updates all repos with packages in its staging dirs
* sanity checks are performed before any repo is touched
* improved performance
* less code; easier to maintain
|
|
|
|
|
|
|
|
Don't try to be smart and remove packages from the staging dir without asking.
|
|
These functions are copied from makepkg
|
|
Added a function to check if user has permission to alter the repos
and db files.
|
|
There are no longer architecture-specific subdirs and the structure was switch to this:
ftp
└── pool
├── community
└── packages
packages contains all packages from core, extra and testing; this naming is in sync with the svn repo naming:
svn-packages and svn-community
|
|
|
|
|
|
This patch sources a PKGBUILD in a subshells instead of directly.
This way we don't polute our scope or overwrite our vars which might lead to unexpected behavior.
|
|
* fix indention
* use consistent output
* set vars local
|