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

Re: Mnemomic names for revisions

From: Adam Gibson <agibson_at_ptm.com>
Date: 2005-05-04 22:43:57 CEST

H. S. Teoh wrote:
> On Wed, May 04, 2005 at 11:18:35AM -0500, Steve Greenland wrote:
>
>>On Wed, May 04, 2005 at 02:33:19PM +0200, Gerco Ballintijn wrote:
>>
>>>Scott Palmer wrote:
>>>
>>>
>>>>It's not really a BIG deal, but I would rather avoid having to treat
>>>>some parts of the tree as serving that sort of special purpose and
>>>>not as part of the organizational structure of my data.
>>>
>>>
>>>This seems to be the main reason people want to have revision labels.
>>
>>No, I want them so can type 'svn diff -r Release_05' in my WC rather
>>than dicking around with URLs.
>
> [...]
>
> Exactly my point. Revision labels are milestones, or landmarks, in the
> history of the filesystem. They should not be conflated with a
> sub-part of the filesystem *at the current time* (i.e., at HEAD).
>
>
> T
>

I wonder if this could somehow help the problem of revisions changing
when you dump and reload and remove some revisions because of a major
commit problem/mistake or if you merge two repositories. The revision
numbers change and now your log messages are incorrect(merged r200 to
r210, etc). You would just have to make sure any log comments would
refer to labels which would move to the newly shifted revisions if a
repo revision is removed from the middle.

One way that I can think of that would help a lot to make this more
functional would be for all commits to create a label to report to the
user based on username. Something like 23johndoe. That would represent
the 23rd commit from user johndoe. The real revision number can still
be determined for those doing some real low level changing to the
repository but the reported revision to the user would be an auto
created label like 23johndoe (maybe report the repository revision too
since so many people use them currently).

Later on if you remove some revisions from the repository it wouldn't
matter unless you remove the revision that 23johndoe points to which
would of course be ok. 23johndoe would always point to the correct
revision even if there are only 22 commits from johndoe because the
revision rNNN was removed which might have had a label of 21johndoe.

Using the username is just an example but this seems like it has some
potential.

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Wed May 4 22:46:07 2005

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.