On Wed, Feb 11, 2009 at 10:45 AM, Mikkel L. Ellertson <mikkel@xxxxxxxxxxxxxxxx> wrote: > It sounds like your example is the exception to the rule. I'm pretty sure we went through the same sort of thing early on in the d-bus process in previous Fedora releases before D-bus announced API stability. And we are going to go through it again as new libraries are introduced and application start trying to use them before they are API stable. These maybe exceptions..but you can't reasonable predict which ones will be exceptions and which ones will not be because we don't have a way to track which projects adhere to API stability best practises. My point is this.. you can't know whether an individual library claims API stability or not without checking with the upstream project as to whether its unstable. I very much doubt right now at this very moment tell me which accuracy which libraries projects claim to have a stable stable and which libraries do not claim api stability. There isn't anything like a coherent SDK specification that would even attempt to summarize which libraries try to provide a stable API for application developers and which do not. You can throw darts, and hope. But you can't really know. There is no enforcement, nor guarantee, or any summary information at all which can be pointed to to tell you that particular library project is expected to be following the best practise soname rule. The only expectation you have is past performance..and that's a poor expectation generator. If a library project hasn't made API stability a communicated feature, your banking on your projected assumptions about the developer's interest in API stability. -jef -- fedora-list mailing list fedora-list@xxxxxxxxxx To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
