Ryan Schmidt wrote:
> He didn't say he was sharing a working copy. :) If he is sharing a
> working copy, then yes, he should stop that. But for example at the web
> design shop where I used to work, the developers also used Windows
> machines, and used TortoiseSVN to check out working copies onto
> Samba-mounted shares from a central Linux development server. These were
> private working copies for each developer, in their home directory on
> the Linux server. It worked. Something could be strange about Alexxoid's
> Samba setup that's preventing it from working for him.
It's very possible that you are correct that the Alexxoid's Samba configuration
is busted. But I still maintain that if the *consumer* of the working copy
resides on the Linux box (i.e. a web server), then it should have it's own
*local* working copy.
The links I provided, especially SVN::Notify::Config make it trivial to create
as many post-commit synced working copies as you should desire (i.e. one per
developer). The sole drawback is that this means commits can be made that will
not be consistent (i.e. "break the build"), if the only testing done is after
the commit itself has been made (not uncommon in a dynamic development
environment like PHP). There are ways around that as well, by having feature
branches for each developer (which has it's own set of complications).
John
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: users-help_at_subversion.tigris.org
Received on 2008-01-29 13:16:45 CET