Re: Patch for timezone handling in cvs log

[Derek Robert Price]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Bart Robinson wrote:

>On 2004-4-29 Derek Robert Price <derek@ximbiot.com> wrote:
> > scjones@jones.homeip.net wrote:
> >
> > >Derek Robert Price writes [quoting me]:
> > >
> > >>>Yes, that's the general concensus.  The only tricky part is
that when
> > >>>you're *not* in client/server mode, there are no requests or
responses
> > >>>-- the server is the client -- so you have to do it in the
"server" in
> > >>>local mode but in the client in client/server mode.
> > >>
> > >>That is already handled on the local/server end by routing output
> > >>through cvs_output() & cvs_output_tagged().  Those functions just
> > >>print if the server is not active.
> > >
> > >
> > >That's what makes it tricky -- when you're in local mode, you don't
> > *want*
> > >to just print the UTC time, you want to convert it just like the
client
> > >would have done had you been in client/server mode.
> >
> >
> > Ah, good point - I missed that.  Anyhow, that should be easily handled
> > with a simple switch on server_active in cvs_output_tagged().
> >
> > Derek
>
>Yes, if I'm talking about the same thing, I have to convert in
>output_tagged as well as in handle_mt.


'Well, yes, but hopefully you could get most of the conversion
functionality into a single function and just call that function from
each location.

>I send the GMT date in the current output form, YYYY/MM/DD HH.MM.SS
>with a MT "date" tag.  If the client is old, it will just print it
>like usual.  


Yep.

>If the client supports it (and -z was given),


Dang.  I really wish we could make this the default operation.  It
seems so clunky to require a command line option to do things the way
they should have been done in the first place.

Hrm...  Then again, as long as we are printing the timezone in output,
UTC normally and only leaving it out when it matches the local
timezone, can't we trust script parsers and leave this out?  This
should be true with the exception of the log command, which uses UTC
without a timezone specifier...  What do others think?  Perhaps we
could change the default on feature for log and require an option on
stable?  I guess that might be a second issue - do others agree this
should go on stable?

>it
>will scan that and do the conversion in handle_mt.  


Yep.

>If we are in
>local mode, the conversion will be done in cvs_output_tagged


Yep.

>(and subsequently routed through cvs_output for the case of the
>client not supporting the MT command at all.)


Not subsequent to a local mode conversion, I would hope, as there
would then be no client, but I think I know what you are getting at.

Mostly sounds good.  Have you written a patch?

Derek

- --
                *8^)

Email: derek@ximbiot.com

Get CVS support at <http://ximbiot.com>!
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFAkWPNLD1OTBfyMaQRAonpAJ4ojDQ4hlD9SHSMCLMLMaztn83xQwCgwE2v
N7vdMOeooYHejPa7wastuYQ=
=UUAu
-----END PGP SIGNATURE-----