Re: [Issue 1523] New - 'svn revert' doesn't reset entry-timestamp

On Thu, Sep 18, 2003 at 05:31:13PM -0000, sussman@tigris.org wrote:
> http://subversion.tigris.org/issues/show_bug.cgi?id=1523
> Issue #:|1523
> Summary:|'svn revert' doesn't reset entry-timestamp
[snip]
> ------- Additional comments from sussman@tigris.org Thu Sep 18 10:31:13 -0700 2003 -------
> I discovered this bug while trying to reproduce issue 1431.
>
> 1. checkout a working copy from http://localhost
> 2. 'svn add; svn commit' a 200M file of random binary data.
> 3. 'svn up'
> 4. echo "blah" >> bigfile
> 5. 'svn status' ----> returns an instant "M bigfile"
> 6. 'svn revert bigfile'
> 7. 'svn status' ----> takes many minutes to return an answer.
>
> AFAICT, 'svn revert' isn't resetting the entry-timestamp when it
> copies the text-base back out. That means that 'svn status' is forced
> to do a brute-force comparison of 200M.
[snip]

I've noticed this before. Sorry I didn't mention it :)

The workaround (on Unix. Haven't tried Windows) is, of course:

[...]
5. svn status
5.5. rm bigfile
6. svn revert bigfile
7. svn status

I noticed this because I have a nasty hack that prints warnings when the
the brute force comparison is done, but the checksums match.

-- 
Michael Wood <mwood@its.uct.ac.za>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org