On Wed, 2001-12-12 at 15:24, Ben Collins-Sussman wrote:
> Greg Hudson <ghudson@MIT.EDU> writes:
>
> > Incidentally, I've had a little revelation: we need to do a newline
> > translation after commit as well as a keyword substitution. Otherwise,
> > you *could* wind up with a local mod after a checkin, which is not
> > really a state we should allow.
>
> How would this happen?
>
> In our current system, when a committed file is 'bumped' after a
> successful commit, this doesn't just mean increasing its working
> revision number. The working file actually *becomes* the new
> text-base. So your model fits in just fine here.
Ah, but in order to be considered unmodified, the working file isn't
supposed to be the same as the text-base; it's supposed to be the same
as the text-base with any applicable line ending translation and keyword
substitutions applied. So if we don't do either of these steps on the
working copy after commit, the file will appear locally modified (if the
contents are examined, at least).
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Oct 21 14:36:52 2006