> > I've been waiting for version 1.2 for a while now, and tried it for the
> > first time over the weekend. It looks very good - I've just got a
question:
> >
> > It seems that if I have a file which has the svn:needs-lock property
set,
> > but currently noone has taken the lock, if I alter my working copy of
that
> > file, and then just do a svn commit, the file will be submitted into the
> > repository. Is this correct? I've tried when another user does have the
> > lock, and again the commit tries to upload the changed file, but fails
with
> > the fact the other user has the lock.
> >
> You can use a pre-lock hook script to disallow that. The design is
> deliberate.
Ok, I'll take a look at how to do that. It does seem to be the easiest way
around the issue for us.
Just out of interest, why is the system designed like this? Its purely for
interest so I have a better understanding of how
subversion expects people to work - in no way challenging its way of working
;-)
> > The reason I ask is that currently we use a system where you have to
lock
> > then modify, and submit changes. We've got into the habit of being able
to
> > make temporary changes to files with no risk of them being uploaded to
the
> > latest version.
>
> So, ideally, you'd want it to ignore modified files with svn:needs-lock
> and that are not locked? If so, no, that's not possible. But you can
> reject the commit and remind people to undo their local changes.
That sounds to be the ideal way forwards. Many thanks for the feedback!
- Dave.
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Mon May 16 14:05:53 2005