On 2007-04-02 18:58-0500 Maurice LeBrun wrote:
> Alan W. Irwin writes:
> > Therefore, I am hugely relieved that I _don't_ see this issue for the backup
> > that I did of the final frozen version of our CVS repository that was used
> > for the subversion conversion.
> > tar ztvf plplot-cvsroot.tar.gz_20070329 |grep Attic |wc -l
> > 1347
> > Maurice, it smells like you had a bad backup on March 4 (although I don't
> > see how rsync could be that unreliable). Do you get good results if you
> > take a cvs backup now? rsync of our frozen CVS repository should continue
> > to work, and your results on the number of Attic'd files should be identical
> > to mine.
> My last rsync of the locked repo looks ok. I don't know why, but there have
> been a lot of bogus ones lately. Here are all the Attic files in my recent
> backups ("last" being Thursday morning before the switchover).
> plplot-cvsroot.20070225.tar.bz2 1341
> plplot-cvsroot.20070304.tar.bz2 51
> plplot-cvsroot.20070311.tar.bz2 51
> plplot-cvsroot.20070318.tar.bz2 51
> plplot-cvsroot.20070325.tar.bz2 51
> plplot-cvsroot.last.tar.bz2 319
> plplot-cvsroot.locked.tar.bz2 1347
> So at least the final result looks good.. no idea what happened with those
Maybe we just got out in time? :-)
I am really glad you confirm the 1347 Attic'd files in the final result.
Also, if your late February and most of March results are an indication of
some lingering CVS problem at SF, it appears I picked a good time to do the
>From my CVS versus SVN consistency tests I am completely confident that we
have everything that could be reasonably extracted from the locked version
in our current SVN version. However, if there were something wrong with the
locked CVS version (such as missing Attic files), my tests would not have
Therefore, Maurice, thanks for making this important cross-check, and I
encourage you to look further for other anomalies (if any) in the "locked"
CVS version that I have depended on for the conversion. Also, I encourage
all developers here to read the trunk, tags, and branches of our current SVN
to make sure it delivers what you think it should for the current HEAD and
the historical branches/tags that interest you. Better to find out now if
there are any lurking problems, rather than two months down the road.
Alan W. Irwin
Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).
Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the Yorick front-end to PLplot (yplot.sf.net); the
Loads of Linux Links project (loll.sf.net); and the Linux Brochure Project