On 1/23/13 10:40 AM, "Carol Frampton" <cfram...@adobe.com> wrote:
>> Sure, Flex 4 is one tine of the fork, the other is Flex 3. But we already
>> donated the Flex 4 'tine', and trunk at Adobe had Flex 4 stuff, so I am
>> calling the 3.x branch the 'fork'. If you look, there haven't been that
>> many checkins to it since the revision that became Flex 4. I think we
>> only
>> need to either commit a patch for each of those revision, or create one
>> patch that represents the difference between that revision and the current
>> state of 3.x. Or we could branch from the head of the Apache Flex trunk
>> and
>> reverse out everything needed to result in the head of 3.x. We'd get the
>> headers right that way. It sort of depends on what we want SVN history to
>> show.
>
> Theoretically I suppose this works but it sounds awfully complicated to
> get right especially without the benefit of the Adobe svn version history
> to trace the changes/branching of the files.
>
> Say I were to get 3.x from Open@Adobe. I pick a file and the version in
> 3.x is different from the version at Apache. How do I easily figure out
> if I can back out enough changes from the version we have at Apache to
> match the 3.x version?
>From which version at Apache? Just like we did for 4.y branch the first
goal would be to make a parity release with 3.x. Someone could write a
smart diff tool that ignored headers and proves that each file otherwise
matches.
Maybe I don't understand the scenario.
--
Alex Harui
Flex SDK Team
Adobe Systems, Inc.
http://blogs.adobe.com/aharui
