[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Patch for timezone handling in cvs log

From: Bart Robinson
Subject: Re: Patch for timezone handling in cvs log
Date: Wed, 28 Apr 2004 11:57:14 -0700 (PDT)

Thanks for the feedback guys.  I can see how calculating the GMT
offset once, based on "now," is not useful.  I will investigate
the MT response stuff on my next wave of inspiration.

-- bart

On 2004-4-28 Derek Robert Price <derek@ximbiot.com> wrote:
 > Hash: SHA1
 > Larry Jones wrote:
 > >Bart Robinson writes:
 > >
 > >>These changes add to log/rlog a -z option, which specifies the
 > >>timezone to use for output.  This is very similar to the -z option to
 > >>RCS's rlog command.  It works for remote repositories as well.
 > >
 > >
 > >No, it doesn't; it's broken in exactly the same way as RCS's rlog
 > >command.  Timezone conversion *cannot* be done correctly on the server,
 > >it *must* be done on the client.  The fatal flaw is that the offset from
 > >GMT is not constant in any timezone that observes, or has at any time in
 > >the past observed, "daylight saving time" (a.k.a. "summer time").  And
 > >what's worse, the correct offset depends on the the particular time
 > >being converted, not the current time.  Because timezone rules are so
 > >baroque, it is not reasonable to expect the server to know about the
 > >rules for the client's locale, only the client can be expected to know
 > >them.
 > If you want to patch something, I think the correct way to do this is
 > using the client's MT request to send tagged dates and then let the
 > client recalculate the date before printing it if it likes.
 > Derek
 > - --
 >                 *8^)
 > Email: derek@ximbiot.com
 > Get CVS support at <http://ximbiot.com>!
 > Version: GnuPG v1.2.1 (GNU/Linux)
 > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
 > iD8DBQFAjzHyLD1OTBfyMaQRAjzcAKDDvOJjg9FQeQwdcgC5jAaxG/jgcgCgpEu3
 > Usc+RNtWmSDwRoDpJw9Eihs=
 > =PDix
 > -----END PGP SIGNATURE-----

reply via email to

[Prev in Thread] Current Thread [Next in Thread]