[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Branch philosophy

From: Greg Goodrich <ggoodrich_at_medinotes.com>
Date: 2004-11-04 22:03:52 CET

Peter Valdemar M°rch wrote:

> Duncan Murdoch murdoch-at-stats.uwo.ca |Lists| wrote:
>> If it's a stable branch containing patches for a release, then this
>> wouldn't make sense. Even if you have no plans to continue patches
>> for that release, someone might, someday in the future.
> This is almost a FAQ isn't it? "How do I make a branch/tag read-only".
> (It isn't in the FAQ.) Well you can't unless you fiddle with apache
> configuration. (I'm using svn+ssh so I'm out of luck anyway).
> I have no idea how difficult it would be to implement, but form a
> users point of view it would make sense to be able to
> svn propset -m "Locking the branch" svn:nocommit $url

This is an AWESOME idea. I've been trying to figure out a clean way of
handling this issue for a few days now for tags. I think deleting a
branch will work for the branch situation. I don't want someone to do a
checkout on a tag, and then later forget that the WC is from a tag and
check something in. I was thinking that the properties would be the way
to go on this. Now if we could just get subversion to implement this

> or something equivalent and have that affect the entire subtree below
> it, preventing commits to it.
> But today we can't. (Something for a pre-commit hook? I've never
> written one..)

The precommit hook is nice, but seems a bit underpowered for locking
things down in this way. For example, I would like my tags area to
allow tags to be created in there, but never modified. Well, when you
do an svn copy, the pre-commit hook gets called, so now you have to be
able to differentiate between a copy and a normal commit, which it
doesn't seem is an easy thing to do. I investigated the svnlook
command, but it doesn't appear that there is really a good way to use
that command to be able to tell the difference between a copy and a
commit. The only mechanisms I can come up with for this are somewhat
hacks (running grep on the output for "^A" versus "^U" for example).

> -------------
> Actually, I just noticed that subversion doesn't store the read-only
> "bit" for directories or files either.. Hmm. My mistake? Regardless,
> that (preventing writing in the checked-out working copy) would not be
> the same as the the svn:nocommit property I suggest above (disabling
> commits).
> Peter

*Greg Goodrich*
Development Manager
*MediNotes Corporation*
1025 Ashworth Road, Suite 222
West Des Moines, IA 50265
Phone: 515.327.8850 ext. 251/
/Fax: 515.327.8856
*Charting Plus - "The Best EMR Value on the Market!"
**www.medinotes.com* <http://www.medinotes.com/>
Received on Thu Nov 4 22:04:05 2004

This is an archived mail posted to the Subversion Users mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.