KDateTime: next iteration

[David Jarvie]

On Tuesday 29 November 2005 12:49, Nicolas Goutte wrote:
>On Tuesday 29 November 2005 11:11, Thiago Macieira wrote:
>> Nicolas Goutte wrote:
>> >On Monday 28 November 2005 21:50, Thiago Macieira wrote:
>> >> David Jarvie wrote:
>(...)
>> >It could become useful for example if Konqueror uses KDateTime instead
>> > of seconds since the epoch. In that case, you can use it in the FISH
>> > KIO slave to tell that you have no idea about the timezone, as it is an
>> > information thatthe FISH protocol does not transmit.
>>
>> FISH should be modified to transmit it. All it has to do is set TZ=UTC
>> before sending anything.
>
>Perhaps for FISH a solution could be found when trying harder.
>
>But there are other KIO slaves that have no idea about the timezone or the 
>time offset of the dates in the files, for example zip and tar, but also 
>floppy or mac. Probably the only thing that could be told is that the time 
>was local on the system having written those files.
>
>Similar to other office documents. The dates/times are supposed to be local 
>but the document does not give any clue in which timezone the dates/times 
>were written, or at least with which time offsets.
>
>>
>> >> Or unless we're talking about an "unknown" timezone (which
>> >> means this should fall into the "Local time zone", which is in turn a
>> >> special case of "time zone").
>> >
>> >I do not think so. It is better to be allow to tell the user that the
>> > timezone is unknown than to force it to UTC or to local time (in
>> > QDateTime's sence).
>>
>> I don't agree. We should keep in mind that this time will get transformed
>> into the standard UTC-seconds-since-epoch of Unix time, so this
>> information will be irreparably lost.
>
>May be we misunderstood us here. I was talking about a possible change of 
>Konqueror, assuming that KIO slaves would be changed too, to allow KDateTime 
>to be passed. (I do not think that it is planned now.) So there would no need 
>over converting to seconds-to-epoch.

The need for a local clock time is clearly described in http://www.w3.org/TR/timezone/. There is also a 
provision for such times in the iCalendar specification (RFC 2445), ISO 8601, and I'm sure in other places. 
So whether a particular ioslave needs it doesn't really matter - the need is quite clearly spelt out 
elsewhere.

--
David Jarvie.
KAlarm author & maintainer.
http://www.astrojar.org.uk/linux/kalarm.html