On Wed, Oct 17, 2007 at 09:37:21AM +0200, Bernhard R. Link wrote: > > If you want to help reducing this number, a good point would helping > getting more documentation about the different versions (and their > differences) available, and (at least for me) make some backports of the > suggested version people should use in your opinion available for etch > (via backports or whatever). At last I don't have the heart to change > to another version without giving it some more robust try > privately (as libdb has not the nicest history of ugly quirks and > shortcomings in the different versions) before that. And robust try > means real work test means within a (at least mostly) stable environment. >
A deep analysis of release notes, discussions with upstream and accurate analysis of APIs are mandatory in this case. I think it is simply unbelievable that the library is not backward compatible at every new release. Issues with specific versions should be managed and solved: having a new source for every release is simply not acceptable. BTW, this is not the only case of such a mess around for libraries (maybe some upstreams should be spanked...). -- Francesco P. Lovergine -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
