Arthur Barrett wrote, On 11/15/2007 05:21 PM:
Todd,

or will the CVS version of validate_repo work with CVSNT's format?

No idea - I suspect there would be a good chance it'd consider some
things as errors.


Understanding that fact is a good thing.


Seriously though - the problem with "validation" is that (some) people
come to rely on it as 'proof positive' that everything is all right -
which it potentially is not.
On the TortoiseCVS newsgroup recently someone was saying they had given
up on TCVS/CVSNT because they had their repo on a laptop and not backed
it up for years and the hard disk died.  Apparently this means that
CVSNT is a poor version control system...

were they paying enough that you did not call for your really big LART!? :)

I think it means that as the use of CVS/CVSNT becomes more widespread
with non-commercial-big-iron-admin types that someone needs to do more
to publicise the importance of backups and data integrity checks (eg:
checkout old things and try "build" etc).  I'm sceptical that some
"verify" script will help - in fact I'm reasonably convinced it'll
discourage users from performing more meaningful tests...


In the hands of NEWBIEs I fully understand your stance, with more advanced users (who most of the time would not be hitting these problems with out hardware errors, or are trying to cleanup after a NEWB) it is helpful to me when I can run a tool that checks for the obvious errors that have been seen in the past, so that I have and idea of whether it is worth trying to work with the broken bits or just restart from the ground up. Of course I am one of those who has reconstructed rcs ,v files from the remains of a raid0 because after the original project lead left no one knew that the project was not being placed on the periodic backup tapes, and I now make sure that in addition to the general system backup there are project backups on separate media. Pain is such a wonderful teacher.

Again, thanks for the update. It lets me know that pointing someone with a broken _CVSNT_ repo towards the validate_repo script is probably a very bad idea.
--
Todd Denniston
Crane Division, Naval Surface Warfare Center (NSWC Crane)
Harnessing the Power of Technology for the Warfighter


Reply via email to