> Wadsworth, Eric (Contractor) wrote:
>
> > So he does
> > his work in a non-working-copy directory, and then he has
> to copy any files
> > he modified into the actual working copy directory, so they
> can be checked
> > in. This, of course, overwrites any changes anyone else
> made... *grumble*)
>
> What???? Subversion (like every other Version Control
> program I have every
> used) doesn't work well with shared sandboxes. Can I go out
> on a limb and
> assume you are using a shared sandbox on a web server for
> online development?
> That is not a good way to handle this particular use. Sure,
> have a sandbox for
> the test web server, but it should only be updated from the
> repository, not by
> dumping files in randomly. Give this guy his own sandbox
> (and a copy of
> TortoiseSVN if, as I suspect, you are a Windows shop), and
> make him check his
> own files in. Run a commit trigger to keep the web server up to date.
>
> Obviously, your users haven't ever used CVS, since that
> requires exactly the
> same kind of control directory. Except in the case of .svn,
> it is relatively
> easy to make the directory itself invisible for almost all purposes.
>
> My 2 cents
>
> John
I should not discuss my co-worker's refusal to use any version-control
tools. It's not subversion-related, it's just a personal problem (or should
it be 'personnel' problem?). The only bearing it has on this discussion is
that I think that more widespread adoptation will occur if users perceive it
to be less invasive, which is a point in favor of external bookkeeping
instead of .svn/ subditrectories in the source tree.
--- Eric
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Aug 25 21:06:46 2003