From kde-core-devel  Thu Nov 24 12:09:29 2005
From: Shaheed <srhaque () iee ! org>
Date: Thu, 24 Nov 2005 12:09:29 +0000
To: kde-core-devel
Subject: Re: KDateTime: revised version
Message-Id: <200511241209.29530.srhaque () iee ! org>
X-MARC-Message: https://marc.info/?l=kde-core-devel&m=113283446420231

On Thursday 24 November 2005 10:49, David Jarvie wrote:
> >(I am jut thinking: has Windows any named timezone? How could this class
> > be used in Windows?)
>
> Windows definitely knows about time zones. I don't know how it represents
> them, but the KSystemTimezone class will handle that. (It probably doesn't
> yet work with Windows - that's just one of the jobs for the Windows port.)
> KDateTime uses the KTimezone class to know about time zones, so it will
> automatically work on Windows once KSystemTimezone works there.

A bit off topic, but while I am still able to contribute...

Windows does indeed have the notion of timezone in the sense that David 
describes. From the comment in the 3.5 ktimezones.cpp:

    // On Windows, HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows 
NT\CurrentVersion\Time Zones
    // is the place to look. The TZI binary value is the TIME_ZONE_INFORMATION 
structure.

However, the API to the database is not exactly clear to me, nor exactly what 
is stored there. FWIW, at the time I wrote the code, I was not sure how the 
Windows port would fit in to the overall scheme...the good news is that 
David's work on the 4.0 version of the KTimezoneSource et. al. seems like an 
excellent direction to take.

Thanks, Shaheed