RE: cvs log and UTC

[Jim.Hyslop]

Derek Robert Price wrote:
> Mark D. Baushke wrote:
> > a bigger question is if that change should also go into 
> > things like $Id$
> > and other RCS keyword expansion strings...
> 
> 
> Got me here.  I tend to avoid RCS keywords as much as 
> possible.  I know
> that they are useful via the `ident' command, but that is 
> supposed to be
> human readable, I think.  Even if it is intended to be 
> machine readable,
> I would say standards would be better.  What do other people think?
I agree - consistent date format is good, and using an ISO standard is even
better.

I think the RCS keywords should also be expanded. Now, having said that, I
also have to point out that I'm not aware of all the other utilities that
may depend on the keyword expansion - I'd never have considered the impact
on `ident', for example.

Seems to me the best way to implement this would be to have a global option
which controls the date format, something like:

-l  Display local time for timestamps
-u  Display UTC for timestamps

Default for the next stable release would be -u, then eventually it could be
changed to -l.

And, of course, all timestamp displays should go through a single function.

> Regardless, when we're just talking about rewriting scripts using CVS
> output to parse _more standard_ dates, I personally don't consider the
> change drastic or the effects catastrophic.  In the long run it would
> certainly be a good thing.
Agreed here, as well.


-- 
Jim Hyslop
Senior Software Designer
Leitch Technology International Inc. (http://www.leitch.com)
Columnist, C/C++ Users Journal (http://www.cuj.com/experts)