Re: Possible data corruption issue?

On Sun, Jun 1, 2008 at 2:40 AM, David Glasser <glasser_at_davidglasser.net> wrote:
> I don't have time to think in detail right now, but couldn't the
> opposite situation happen: where the user has the text of r1 but the
> fake metadata of r2, and commits happily not ever seeing r2? (ie, if
> the update is not a back-update.)

OK, using a DAV server this time, I first recreated Karl's scenario
and got the same behavior.

I then used a "normal" svn up to set my entire working copy to an
older revision of the repository. I then did svn up with
--depth=empty on a single file and specified the latest revision. I
confirmed that the contents of the file were still not correct, but
svn info thought it had the right revision. In other words, I was
seeing all of the "expected" problems.

I then made some modifications to the file and attempted to commit.
This is what happened:

$ svn ci -m "This should fail"
Sending snippets/HtmlSnippet1.html
subversion/libsvn_client/commit.c:919: (apr_err=160024)
svn: Commit failed (details follow):
subversion/libsvn_ra_neon/commit.c:495: (apr_err=160024)
svn: File or directory 'HtmlSnippet1.html' is out of date; try updating
subversion/libsvn_ra_neon/util.c:711: (apr_err=160024)
svn: The version resource does not correspond to the resource within
the transaction. Either the requested version resource is out of date
(needs to be updated), or the requested version resource is newer than
the transaction root (restart the commit).

So it still appears that corruption cannot enter the repository via
this problem.

-- 
Thanks
Mark Phippard
http://markphip.blogspot.com/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: dev-help_at_subversion.tigris.org