The previous revisions worked on /src/freeboogie. That is why checking out the whole repository allowed diffs to be executed on that folder (outside of /trunk/)).
StatSVN doesn't care what version numbers are there. It basically is supposed to do one diff per revision for which we can find changed files. We don't need to do a diff when there is only a delete in the revision ( as with below ).
<logentry
revision="492">
<author>rgrig</author>
<date>2009-06-23T21:03:02.135053Z</date>
<paths>
<path
action="D">/src/freeboogie/trunk/FreeBoogie/fbrun.bat</path>
</paths>
<msg>stop supporting bat on windows (use jar instead)</msg>
</logentry>
There are also a few unsupported scenarios listed in StatSVN's limitations that you might be encountering. These are known an documented, but no one has yet tackled these issues.
Thanks,
---
Jason Kealey -
[hidden email]
LavaBlast Franchise Software -
http://www.lavablast.comSimplifying day-to-day franchise operations
On Mon, Jul 27, 2009 at 9:59 AM, Radu Grigore <
[hidden email]> wrote:
>
> On Mon, Jul 27, 2009 at 2:41 PM, Jason Kealey<
[hidden email]> wrote:
> > Can you post an extract of your log file that includes an svnsync from
> > another repository? I don`t think that is something we tested. (I would
> > have assumed it would have just been a regular revision, but maybe not).
>
> I would expect the same. You can see the log that leads to 7 diffs at
>
http://radu.ucd.ie/temp/svn.log> The first commit after svnsync is 520.
>
> I tried now to checkout the whole repo (as opposed to only the trunk).
> Now it scheduled 384 diffs, which is much better but, I believe, still
> not complete. Could StatSVN have problems when certain version numbers
> are skipped in the log?
------------------------------------------------------------------------------
_______________________________________________
Statsvn-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/statsvn-users