Clemens Schwaighofer wrote:
> On 02/08/2005 02:38 AM, Monks, Peter wrote:
> > I could be wrong, but it looks to me like Subversion is using ISO8601
> > format [1] for the date/time, in which case the "Z" indicates that the
> > value is in UTC (which is almost, but not quite, identical to GMT).
> >
> > AFAIK this isn't configurable, and I'm not sure whether allowing it to
> > be configured would make sense (consider a chronologically distributed
> > team, where each team member could be working on any file in the
> > repository - which timezone should Subversion use for each file?).
>
> $Id: index.php 273 2005-02-07 04:15:30Z gullevek $
> $LastChangedDate: 2005-02-07 13:15:30 +0900 (Mon, 07 Feb 2005) $
>
> but as you can see the Last change date has the local time (in this case
> Japan) and in brackets the difference to the UTC (Z) time.
>
> I would _love_ (really love) to have exact the same in the $Id$ tag. it
> would be 100% consitent. and 5 digits more, I think thats worth :)
Here's one against:
When I export a version of my project, I want all files to carry an identifier
by which I can _exactly_ determine lateron to which revision this file
belongs. This ID should be the same whoever and wherever checks out that
file, so localtime should not be a part of it.
Yes, the information content is exactly the same if I use localtime plus the
offset to GMT/UTC, but I wouldn't want to complicate things by requiring
people/programs to first parse and convert this to a common format before
comparing - after all, it is an ID and an ID should be most of all unique.
So: I'm against making $Id$ configurable, but I'm not against providing
something equal that is not called $Id$ that provides the features you need.
just my 2cc
Uli
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Tue Feb 8 18:58:35 2005