[Opensim-dev] to branch or to break?

[Sean Dague]

On Fri, Nov 02, 2007 at 02:10:43PM +0100, Anders Arnholm wrote:
> On Fri, Nov 02, 2007 at 07:35:54AM -0400, Sean Dague wrote:
> 
> > I used to think this as well.  Then I started using hg a lot, and
> > realized that I actually had a chance of merging correctly with hg
> > (mercurial).  More importantly, you get to try the merge in your tree
> 
> That depens on if you are ok with it didn't conflict and something got
> out that compiles, or you seen the need to understand the changes, what
> they do, why thet do it that way and that the limitations of that
> solution. In the end after a merge you have a diff, a change to
> understand, if this are a big one it's much harder, this is the big
> problem with all merging.

Yes, but the problem is made better in distributed SCM by understanding
the parent of change.

A ---*---> B ------> C ------> D
     |
     +---> E ------> F

If I know that E is really a change from A instead of a change from D, I
can do a much better job on recognizing what I was fundamentally
changing, and have a better job resolving F & D.  

Again, nothing complete solves 3 way merge, but you've got a 10x chance
of doing it right (with at least 10x less brain power expended) on an
SCM tool that actually understands how you got there, than one one that
assumes source code is a linear thing.

	-Sean

-- 
__________________________________________________________________

Sean Dague                                       Mid-Hudson Valley
sean at dague dot net                            Linux Users Group
http://dague.net                                 http://mhvlug.org

There is no silver bullet.  Plus, werewolves make better neighbors
than zombies, and they tend to keep the vampire population down.
__________________________________________________________________
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://opensimulator.org/pipermail/opensim-dev/attachments/20071102/04224f1a/attachment-0001.pgp>