RE: Website development

[Chris Chambers]

If you can't do different sandboxes for each developer, you won't be liking
concurrent source control systems. For shared resources, you need locking
version control (RCS or something).

Even with a shared "cvsdeveloper" account, you'd have issues with
overwrites.

I've been staring at this ugly fact for a while... if it's at all possible
to do the gruntwork required to create a sandbox for each developer, I'd
recommend going that route. Otherwise, I recommend a locking model.



> -----Original Message-----
> From: address@hidden [mailto:address@hidden Behalf Of
> Atkinson, Chip
> Sent: Thursday, February 15, 2001 12:08 PM
> To: 'address@hidden'
> Subject: Website development
>
>
> Greetings,
>
> I'm trying to set up CVS for a web development project and have come into
> what appears to be a minor problem and am looking for advice on
> the best way
> to fix the problem.
>
> We need to all work out of the same directory, which is where the
> web server
> works out of and we can't just set up multiple web servers
> because it's not
> just a regular Apache or IIS setup.
>
> When I check out the files for the web site they are "mine" even
> though they
> are in the common directory.  When a coworker makes changes and tries to
> commit the file, cvs needs my login and not my coworker's.  If I change
> CVS/Root to have my coworker's login, then he (or I as he) can commit the
> file in question.
>
> The solution to the problem that occurs to me is to give everyone the same
> login, but then that seems to get in the way of logging.  Every change is
> from somedevacct or whatever.
>
> Is there some other way that we can all use the same sandbox without
> resorting to a generic login?
>
> Here's the setup:
>
> CVS Server: Linux box
> CVS Clients: W-2K boxes using wincvs
>
> Thanks in advance,
>
> Chip
>
> _______________________________________________
> Info-cvs mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/info-cvs
>