about 2.5.10 // Resolved 0002294: DB gets corrupted after schema change

[Robert Brenstein]

>  > Remember our discussion about branching? It is exactly to avoid such
>>  dilemmas. If you have proper setup for branching, fixing the bug in
>>  2.5.x branch requires a single CVS command to have it merged into
>>  3.0.x branch, where this bug also exists most likely. You make
>>  current release more stable and don't sacrify time spent working on
>>  the next release.
>
>All visa versa although.
>
>* We have now TAG 2.5.9
>* we have go forward with 3.0 in the HEAD/MAIN branch of CVS
>
>* YES I can spend 1-2 days
>     --> Extract totally all on date of 2.5.9
>     --> create new 2.5.10 branch at this state
>     --> merge fixes of few bugs
>     --> produce build
>     --> trash all
>     --> return to main branch
>
>This is steps of bug-fix release.
>
>And, it is not so single command as reality shows :-)
>After merge you never have believe you have not catch problems.
>
>So we keep in each branch TESTS that corresponds to STATE of that branch.
>And after merge we need yet make sure that tests are not broken.
>So everything is not so simple as theory says :-)
>
>
>* AND yes, it is possible to build to 2.5.10 even after 3.0 ship.
>Although not sure that Lynn will agree that.
>

If that is how you work with branching, then yes, but it is not how 
branching is properly used. It seems that you use branching but do 
not fully understand it yet to use it to your benefit beyond 
archiving specific releases.

And yes, being able to release 2.5.10 after 3.0 should be not just 
possible but easy if you do things right :-)

Robert