(ichimunki) Re: CVS and perl

I've found this to be a good resource: CVS Book.

I use CVS for all kinds of stuff, not just coding (system configurations, home directories, etc). I would set up one server to do CVS for all projects at all stages. Maybe once a project gets to a certain point you can strip the history by doing a release and then checking that in as a new project.

Comment on (ichimunki) Re: CVS and perl

Replies are listed 'Best First'.
Re: (ichimunki) Re: CVS and perl by edebill (Scribe) on Nov 25, 2001 at 08:24 UTC
Maybe once a project gets to a certain point you can strip the history by doing a release and then checking that in as a new project. That's basically what we do where I work. Although we typically don't `commit` the entire checkout to the new module - we build it up a few files at a time so we have an opportunity to clean things up as the new module grows. If nothing else, having a single CVS repository makes keeping track of your `CVSROOT` easier. This gets to be a big thing when you have 10-12 people who aren't all CVS or even UNIX literate to support. I wouldn't recommend shared CVS logins, though. Just define a cvsusers group and sgid the repository directories to it. That way you'll still be able to see what user just `commit`ed a particular bit of brain damage.	[reply]

Replies are listed 'Best First'.

Re: (ichimunki) Re: CVS and perl
by edebill (Scribe) on Nov 25, 2001 at 08:24 UTC

Maybe once a project gets to a certain point you can strip the history by doing a release and then checking that in as a new project.

That's basically what we do where I work. Although we typically don't commit the entire checkout to the new module - we build it up a few files at a time so we have an opportunity to clean things up as the new module grows.

If nothing else, having a single CVS repository makes keeping track of your CVSROOT easier. This gets to be a big thing when you have 10-12 people who aren't all CVS or even UNIX literate to support. I wouldn't recommend shared CVS logins, though. Just define a cvsusers group and sgid the repository directories to it. That way you'll still be able to see what user just commited a particular bit of brain damage.

[reply]