> Hi,
>
> we encountered (again) the error already discussed in the
> earlier thread
> "Corrupted FSFS commit" (starting February 25).
Update:
Server version: 1.6.5
Client version: 1.6.7
Martin
Kutter, Martin wrote on Fri, 26 Feb 2010 at 10:29 +0100:
> The reported numbers are unstable, and flip between positive and
> negative values.
They don't have any reason to differ across runs, do they? So, what's
next? Uninitialized memory? Valgrind? (and how to make it play nicely
with poo
> Daniel Shahaf wrote:
> Kutter, Martin wrote on Thu, 25 Feb 2010 at 13:29 +0100:
> > I got a strange error in one of our subversion
> > repositories: On checking
> > out a file from revision 3865 on, svn reports "Svndiff
> > contains a too-large window".
>
> > This is the error message added
Daniel Shahaf wrote:
> Kutter, Martin wrote on Thu, 25 Feb 2010 at 13:29 +0100:
> > I got a strange error in one of our subversion
> > repositories: On checking
> > out a file from revision 3865 on, svn reports "Svndiff
> > contains a too-large window".
>
> This is the error message added in 1
