Re: Patch for fields of `struct buffer'

[Tom Tromey]

RMS> So why merge the macro calls into the trunk?
RMS> You could put them in a branch.  If you ever get preemptive threads
RMS> working such that we want to install it, we could install that change then.
[...]
RMS> That goal seems unobjectionable, but why do you need to change these
RMS> calls in the trunk to be able to work on your branch?

We did have a branch, but it quickly got stale.  Because the needed
infrastructure changes are pervasive, it was very hard to do merges --
there were a lot of conflicts.

This is not a problem in itself, since we could simply not do merges.
But, then that also would greatly reduce the prospects for ever merging
back.  Putting these changes on the trunk makes it possible to keep the
branch relatively up-to-date, making it simpler to merge into trunk when
the time comes.

Tom