Re: Trial Import of RCS to SVN - success, and statistics, and questions

In two separate emails, Mark Benedetto King asked:
> Would it be possible to make your RCS repository available for some
analysis?

Yes. The gzipped tar file is just under a half megabyte. Send me
directions on how you want it sent to you - email will be OK if you want it
that way, or FTP. And I wouldn't send it to the users list as a whole.
The file has a top-level directory cvs, with sub-directories inc, lib
(empty except for two sub-directories), lib/JL and lib/ESQL. If you
decide you want to see the SVN archive itself, a bzip2'd tar file is ready
and waiting - it is just over 2 MB. I also have a trace of the import
process - it is just under 0.5 MB uncompressed, but bzip2 reduces it to
less than 24 KB.

> What percentage of that is occupied by the log files?

You saw from a later posting that I've got just one log file left - and
even discounting that, the SVN archive is about 5 times the size of the RCS
(masquerading as CVS) archive.

> Also, what version of svn and BDB are you using?

I'm running on Sun Sparc (Ultra 10), Solaris 8. I've got SVN 0.37.0. I'm
using BDB 4.2.52, compiled separately from Apache -- I'm not playing with
the WebDAV stuff yet. I used GCC 3.3.2. The Python I used is 2.2.1
(back-version, but it seemed to work OK). The APR and NEON libraries are
those distributed with SVN 0.37.0. Tell me if you need any other version
information.

--
Jonathan Leffler (jleffler@us.ibm.com)
STSM, Informix Database Engineering, IBM Data Management
4100 Bohannon Drive, Menlo Park, CA 94025
Tel: +1 650-926-6921   Tie-Line: 630-6921
      "I don't suffer from insanity; I enjoy every minute of it!"