Re: testing code inside CVS, but not increase rev #?

["Daniel E. Markle" <syntax@ashtech.net>]

On Thu, Mar 25, 1999 at 07:30:14PM -0800, Aaron D. Turner wrote:
> 1) I have to run two commands to get the file on the server.  That's one
> too many.

This is still a strange, intermittent problem we are having with the cvs
scripts.  The cvs config at linuxkb explicity says to run the scripts, yet
they still don't run for some.  Perhaps it is a cvs version conflict, can you
send me the versions at the locations you use?

> 2) Having so many rev numbers makes it virtually impossible to find the
> rev you want.

The only way this should happen is if people are not logging properly.  As
long as everything is well logged, it should be easy to look back through and
see what you are looking for.  I can certainly see how this is annoying,
though.  Could you do a checkout into your linuxkb.org/~aturner/webdev, for
example, directory, then work from there, committing when done?

> 2)  Learn a way to overwrite a rev.  Ie, commit a file(s) without
> incrementing the rev number.

As far as I can tell, this isn't possible, but I will continue to look into
it.  It looks very unlikely, however.

I think CVS is really starting to show it's code-specific nature, where
everyone has a local development environment and these issues are
non-existant.  No one would happen to know of a revision control system for
websites, would they?

-- 
-------------------
Daniel E. Markle
syntax@ashtech.net
-------------------