Re: Basic usage question

[Steve Greenland]

On Mon, Jan 21, 2002 at 05:02:27AM -0800, James A. N. Stauffer wrote:
> --- Kaz Kylheku <address@hidden> wrote:
> > If people cheat by committing only the files that they modified, they
> > can get around the up to date check. But that is a bad idea because
> > the changes you make in one set of files can semantically conflict with
> > changes in another set of files.
> 
> So are you saying that committing all changes at once is better than
> committing each change individually?  If yes, how can distinct messages be
> used for each modified file when they are all committed at once?

No, he's saying that you can get around the "up-to-date" check by
specifying specific files on the command line. If you do so, it will
check only the files specified. If you specify a directory (or nothing
at all), it requires all files be current before checking in any (or so
Kaz is saying; I didn't know that).

But typically the best way to operate (I think) is:

1. Edit files
2. Test build. If fail, go to 1.
3. Test function. If fail, go to 1.
4. Update project (or current module, or whatever level 
   you feel is appropriate.)
5. If 4 changed anything, got to 2.
6. Commit modified files as you desire.

Kaz's point was that you can, to some extent, avoid CVS complaining
about your sandbox being up-to-date, but usually you don't want to.

Steve