Re: changing the head to another branch

[Robert Clark]

On Friday March 31, 2006 15:45, Larry Jones <address@hidden> 
wrote:
> > This means, when you work on the 1.3.2.x branch, for every
> > operation the server has to compute the current revision by
> > starting at 1.10 then applying the diffs down to 1.3 and the
> > applying the diffs to the most recent revision of the 1.3.2.x
> > branch.  I imagine this may take considerable time if you have a
> > large project with a couple hundred or even more files.
>
> That's why you should keep your main development on the trunk.

I've sometimes run into the case where the HEAD branch has become 
stagnant and all future development should occur from code currently 
in a side branch.

What we do in this case it merge the side branch down to HEAD, 
unconditionally replacing, rather than merging, all the code on HEAD. 
From that point on, all development occurs on the HEAD branch. make 
sure you tag HEAD before you do the merge in case you change your 
mind later.

I've found this to be a good reference when making this kind of 
change.

  <http://helpdesk.wisc.edu/middleware/page.php?id=4087>

Look for "Replacing one Branch with Another" on that page. 

- Rob