Re: philosophical question about frequent saving of files

On Aug 29, 2005, at 1:40 PM, Ivan Soleimanipour wrote:

> I quite often diff, for the same purposes you want, against the
> previous delta. Shouldn't that be enough for accidents?

Not for new work. You add some new classes, or refactor, and there's  
no file history until you've checked in. For larger changes that  
accumulate several days worth of work, that's no good.


> If I need to make an experiment in a file, I need to check it (out of
> sccs) anyway. There's my backup.
>
> Aren't sccs deltas "personal"?

No, because the next putback will commit all the deltas for the world  
to see. I want to have clean file histories, where each checkin has a  
useful comment corresponding to a new feature or bug fix, and  
somebody can clearly diff say revision N and N+1 to see how I fixed  
bug M - not have to figure out that revisions N to N+25 are all  
related to work on a particular bug fix.

-- Tor