On 26/10/2007, Dustin J. Mitchell <[EMAIL PROTECTED]> wrote: > > On 10/25/07, nick aschberger <[EMAIL PROTECTED]> wrote: > > > ... and then finally, the last thing printed: > > > > <path > > > > copyfrom-path="/branches/moose-support-1/omap/src/chips/w3gdigrf" > > copyfrom-rev="202" > > action="A">svn: Write error: Resource temporarily unavailable > > Huh -- looks like an svn bug, rather than a svnmerge bug. Googling > doesn't turn up much about that error, but you might want to try to > run the same 'svn log' commands as svnmerge is running, and see if you > get the same error. I suppose that possible reasons could include a > corrupted repository (svn log looks at old revisions that an > update/commit wouldn't necessarily touch) or network problems. If you > can replicate it outside of svnmerge, you may have a clearer idea of > what triggers the problem. > > Dustin >
Thanks Dustin, I have been thinking this myself actually - the subversion storage server is remote (in the US, and I am in Australia) so it may be some kind of weird timeout. I do not think it is a corruption in the repository (how do you even diagnose that?!), as everything else works fine. It is possibly a dropped packet/connection. However, the "svn log" command seems to work fine outside of svnmerge.pyscripting! There is a huge amount of log output - maybe the input buffer inside python fills? I don't really know about how python manages that memory. Having a look at the script, I'm not really sure either what you are doing with all the xml output - but I take it that it's all required? cheers Nick
_______________________________________________ Svnmerge mailing list [email protected] http://www.orcaware.com/mailman/listinfo/svnmerge
