Anders Engström <aengstrom@gnejs.net> writes:
> Now - looking at subversion - how would this issue be handled by
> subversion? Is there some way to "push" the modified file back to the
> client as a result of a successful commit?
No, we are just like CVS in this respect.
> As a parallell - suppose you have a simple text file with an $Id$ line, and
> you modify the local copy, and the issue a commit, I suppose that the
> keyword substitution for the $Id$ field is performed both on the server
> and on the client (by the "server subversion process" and by the "client
> subversion process")?
No, only the client performs the substitution.
> IMO, this could be handled by the server only - on
> a successful commit the server would send back a diff to the client
> containing the modified keyword substitution line ($Id$). *If* the
> commit worked this way, our ugly "Keyword substitution hack" would work
> without the mentioned drawback...
>
> Any comments on this?
Yes -- "Abandon all hope, ye who enter here."
:-)
-Karl
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Mon Sep 29 17:34:56 2003