Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to sales@march-hare.com.
Oliver Koltermann wrote:
>> In a way it makes sense - it's returning the versions from the correct
>> local time.. it's just the timestamps that cvs then shows you are from
>> UTC, so you end up with confusion.
>
> This is the way it is documented in the section "Common command
> options" of the CVSNT manual btw.:
Ah, thanks! I missed that section.
I still would like to suggest two improvements:
- In the documentation (manual, help file) always mention whether it's UTC,
server local or client local whenever a time/date is mentioned. Even when
it is mentioned that it is local time, it doesn't say whether it's server
or client local time.
- Likewise, in the command help ("cvs -H command" etc) always indicate what
time it is. Or here maybe only when it's not UTC -- but in any case state
whether it's client or server local.
For example, it seems that -T (ls, log) is server local. And I'm not sure
whether -D (checkout, update, ...) is client or server local. (Server-local
wouldn't really make a lot of sense, IMO.)
> And the sticky option must be UTC to be comparable between working
> copies around the world. So after all it makes sense!
Well, halfway. every date/time the server returns (normally) is in UTC, to
me it would make more sense to use UTC for the -D option. The other thing
is that we need to know for every local time that it is local, and whether
it's client or server local. And it may not always be trivial to know the
server's timezone... But that's just me, of course :)
Gerhard