Re: how to manage the case where only one file differs per branch

[Jean-Claude Gervais]

On Sun, 2007-03-04 at 06:39 -0800, address@hidden wrote:
> hello, i've decided to finally start using cvs and it happend with a
> particular project in mind. i'm working on a web-related project,
> where i have to test changes locally on my personal computer and then
> publish them on a different server for the web. now my problem is,
> that i have to manage two sets of certain variables, such as, for
> instance, at least local and remote usernames and passwords.
> 
> what is the right approach? should the two scenarios be considered as
> two branches? but if only one file (which contains the sensitive data)
> is tagged, i'd have to use the -f switch of update, if i'm not
> mistaken, but exactly how?

Put everything common on the head, leave the different file on a branch.

To do day-to-day operations, you need to automate your sessions so that
changing servers will be easy. Begin by writing the following
instructions in a script file.

Get the head.
Then fetch the one changed file directly.

Call the script SetA.

Repeat the entire sequence to create as many sets as you need.

An example could be

cvs co .
cvs co -r BRANCH_CUSTOMER1_VARIABLES dirname/variablefilename
 
J