Karl Fogel wrote:
> Regarding these two:
> -r --read-only Make new working files read-only.
> -w --writable Make new working files read-write.
> Does anyone here actually use them?
> I've never had occasion or desire
> to check out working copies anything but read/write... But maybe this
> is important?
> If it turns out that none of us do it, I think it may not be worth
> supporting this.
I'd like to throw a small spanner in the works and suggest that
we need to support reserved checkouts and persistent locks on
the repo (the same thing, really), and --read-only should be the
default in such cases. Without going into details right now, I'll
just state a few reasons:
- Users want them. A lot of people around me are using CVS
(it beats the hell out of RCS), but would prefer to have
reserved checkouts instead of having to merge changes and
-- as has happened -- commit files full of conflict markers
- Repository backups. Lock the whole repo, do the backup, unlock
it. A backup of a repository for a large project can take a *long*
time. You don't want people to check in changes during that time,
but there's no reason to forbid checkouts.
- Release management. Lock the release branch and only let the
release manager check in changes.
- Usage tracking. Who's editing what? Just list the locks.
Most of that can probably be accomplished by modifying access rights,
but having persistent locks would be a good thing, IMO.
(Now, since asbesthos suits are politically unacceptable lately,
I'll have to see what's new in the ceramic heat shield business.)
home: <brane_at_xbc.nu> http://www.xbc.nu/brane/
ACM: <brane_at_acm.org> http://www.acm.org/
Received on Sat Oct 21 14:36:12 2006