Re: PREV() (Was: Adding changeset-like functionality to subversion)

[Noel Yap <ya...@yahoo.com>]

--- Noel Yap <ya...@yahoo.com> wrote:
> > If a change happened at revision 53, on a branch -
> > like it did here -
> > then the change is returned by a diff of revision
> 52
> > and 53. You do
> > not have to go back and search the previous
> revision
> > the same file (or
> > directory) actually changed last.
>
> Thanks for clearing this up.  It's all crystal clear
> now.  The more I learn about Subversion, the more I
> like it :-)

Like I said in my last post, I like this feature of
Subversion.

OTOH, unless I'm mistaken, a PREV(rev) functionality
would still be helpful in implementing "svn blame" (or
"svn etiology" ;-).  I'm thinking the command can be
implemented by going backwards through the physical
revision history (eg the revisions in which the file
really changed) of the file and annotating the file
along the way.  PREV(rev) would go a long way in
helping with this unless there's a better way to
implement it.

OTOOH, maybe PREV isn't the way to go since "svn
blame" is used on a per-file basis rather than on an
entire tree.  Anyway, the functionality would still be
helpful (unless I'm still missing something) even if
it's not called PREV.

Noel

__________________________________________________
Do you Yahoo!?
Faith Hill - Exclusive Performances, Videos & More
http://faith.yahoo.com

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: PREV() (Was: Adding changeset-like functionality to subversion)

[Nuutti Kotivuori <na...@iki.fi>]

Noel Yap wrote:
> Anyway, the functionality would still be helpful (unless I'm still
> missing something) even if it's not called PREV.

Well yes, the functionality would be helpful. Nowadays if I know that
this file was last modified in revision 1234 and I want to know what
revision before that was the last modification, I'll have to either do
'svn update -r 1233' on the file in a working copy and then 'svn info'
- or 'svn log -r1233:1' on the file and then Ctrl-C it away when I've
seen what the last change was.

Doable, but more than a little tedious.

-- Naked


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: PREV() (Was: Adding changeset-like functionality to subversion)

[Karl Fogel <kf...@newton.ch.collab.net>]

Noel Yap <ya...@yahoo.com> writes:
> OTOH, unless I'm mistaken, a PREV(rev) functionality
> would still be helpful in implementing "svn blame" (or
> "svn etiology" ;-).  I'm thinking the command can be
> implemented by going backwards through the physical
> revision history (eg the revisions in which the file
> really changed) of the file and annotating the file
> along the way.  PREV(rev) would go a long way in
> helping with this unless there's a better way to
> implement it.

Yes, something like this would be useful, because you may not know the
exact revision where something changed, you just know its current
revision and want to see "the most recent change that produced this
content".  (Of course, you can find out the last-changed-rev, but
right now it's inconvenient.)

But if you run 'svn log' on a file, you'll see all the revisions where
it changed.

-K

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org