Hi all, I certainly interested in anything that interests ATrpms users. What is needed to build ARM packages? Can one cross-build on x86_64? qemu?
But Brian is also right, the current pressing matter is to alter the infrastructure in such a way to allow more contributors to have higher levels of access including root access to some resources. ATM ATrpms is hosted at the FU Berlin, and Jens Dreger, both a good friend of mine and one of the higher IT responsibles around there, wants to help upgrading/restructuring the hardware setup allowing for more degrees of freedom. But how can we help you ATM? BTW we are also not using koji, but self written scripts. But maybe for scaling reasons koji does make sense. So perhaps you'd even like to kojify ATrpms even outside of ARM? :) On Wed, May 02, 2012 at 02:00:15PM +0100, Gordan Bobic wrote: > On 02/05/2012 13:48, Brian Long wrote: > >Hi Gordan, > > > >I'm not sure how long you've been using ATrpms, > > Years. > > >but I believe someone > >would need to step up and handle the builds and any issues that crop up. > > Indeed, but most packages tend to build and work OK. So as a first > pass, having everything that builds as is from src.rpms would be > nice. I'd be more than happy to look into any build failures. I can > take all of atrpms src.rpm packages and rebuild them myself, but I > don't want to diminish the value of other repositories by > effectively usurping the work of the maintainers thus far. > > If there is no interest in supporthing ARM, then fine, I'll fork and > maintain the ARM repository myself (same way I keep up with upstream > - this wouldn't be any different). > > > We cannot just pile more work on Axel as I believe he's already > >overloaded. :-) > > I get that, see above. > > >A few of us have discussed helping Axel set up a new build system to > >automate more of the tasks (like koji for Fedora). It really hasn't > >taken off just yet. > > I'm in the process of putting together a koji for RedSleeve, but in > fairness I am not too sure what the advantage is relative to the > increase in complexity, compared to simply having a NFS path with > all the src.rpms and a 10 line shell script that puts down lock > files and fires up mock to build each package in turn (and copies > the built packages to a repository path, and saves the logs for any > build failures). Sure, koji is a bit more "enterprisey", but I just > don't see any fundamental benefit. > > >Would you be willing to dedicate time to help build the ARM version of > >packages? > > Absolutely, see above. The reason I ask is because I just had to > build mediawiki for el6-arm (from the atrpms src.rpm). This sort of > thing crops up all the time so I figured I can't be the only one > that would find this sort of thing useful. And of course it's nice > not having to do _everything_ myself. ;) > > Gordan > > _______________________________________________ > atrpms-users mailing list > atrpms-users@atrpms.net > http://lists.atrpms.net/mailman/listinfo/atrpms-users > -- Axel.Thimm at ATrpms.net _______________________________________________ atrpms-users mailing list atrpms-users@atrpms.net http://lists.atrpms.net/mailman/listinfo/atrpms-users
