On Mon, May 12, 2008 at 4:38 PM, Bill Nottingham <notting@xxxxxxxxxx> wrote: > > Sure, but using the *actual* data they want (which repository > it came from) versus an approximation of that data (using > a 'jpp' release tag, a specific 'Group' tag, or a specific > 'Vendor' tag) seems to be the appropriate answer. This makes sense to me too. > That being said, wanting to swap out large portions of the > releases' stack willy-nilly is still kind of nuts. Well, largely our "core" stack (and by this I mean comps) does not (yet) have dependencies on many Java libraries. So there's not a big deal swapping out say our objectweb-asm for JPackage's. However, this is most definitely not a sustainable system in the long term. I sort of view JPackage as just a specific instance of the cross-distribution collaboration problem. Most distributions tend to focus primarily on dependencies driven by desktop apps or their own internal needs, but the general free software world is far larger than that, even before OpenJDK opened up a lot of Java software to us. -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
