AW: Semantics of locking a "revision"

[Julian Ruhe <jr...@axway.com>]

Hello Stephan,

back from vacation...

You're right. My set-up was not correct. I thought I could lock a wc revision of a file that was moved in its forward history. In fact, it was copied not moved  so the rev I successfully locked was still existing on HEAD. Please forget about the issue. Thank you for your reply

Greetings,
Julian

-----Ursprüngliche Nachricht-----
Von: Stefan Sperling [mailto:stsp@elego.de] 
Gesendet: Dienstag, 18. Dezember 2012 18:27
An: Julian Ruhe
Cc: users@subversion.apache.org
Betreff: Re: Semantics of locking a "revision"

On Tue, Dec 18, 2012 at 05:14:18PM +0000, Julian Ruhe wrote:
> Hello,
> 
> if I co a working copy with -rREV with REV older than HEAD, I am able to get a lock on any file. But it seems to me, that the lock is never published to the server as no "svn info URL" shows an existing lock on the file on any working revision, while the lock is existing in my wc (displayed by svn status / info).
> 
> So what is  the purpose of having an isolated lock in my wc? Did I miss something?
> 
> Best regards,
> J. Ruhe
> 
> p.s. Using svn 1.7.6 (win32 cygwin client), svn 1.7.7 (debian server)

With a 1.7.x build of Subversion, in a simple test over ra_local (i.e. a file:// URL) I get this:

Updated to revision 1.
$ svn lock alpha                                                   
svn: warning: W160042: Lock failed: newer version of '/trunk/alpha' exists $ svn up -r2 Updating '.':
At revision 2.
$ svn lock alpha
svn: warning: W160042: Lock failed: newer version of '/trunk/alpha' exists $ svn up -rHEAD Updating '.':
U    alpha
Updated to revision 3.
$ svn lock alpha
'alpha' locked by user 'stsp'.

So obviously I cannot reproduce the issue.
Can you please provide more details?

Are you sure the file you're trying to lock is out-of-date?

Which repository access method are you using (http://, svn://, ...)?