Re: [RFC] HTTP timezone

On Sunday 10 June 2007, Tijnema wrote: > On 6/9/07, Richard Lynch <ceo@l-i-e.com> wrote: > > On Sat, June 9, 2007 8:06 am, Stefanos Harhalakis wrote: > > > Timezone: +0200 > > > > > > that will specify their timezone offset. This way scripts will be able > > > to > > > provide appropriate date/time strings/representations and/or content. > > > > It's pretty useless and unreliable since user's clocks/timezone > > settings are incorrect far too often... > > > > YMMV > > I agree with you, clock settings are incorrect way too often, I just > checked my own, and I see that the time is correct, but the Timezone > is at GMT, while i live in GMT+1, but in summer, it's GMT+2. > That brings me to the next point, what about DST? > I've read your complete draft, and it doesn't say anything about DST, > What should browsers send for my country? +0100 or +0200? In DST the offset from UTC is adjusted and thus it will become +2 in your case. Perhaps a note about DST would be usefull. The original idea was to use the POSIX 1003.1 timezone string that optionally includes the DST information but this introduces a lot of complexity and was changed to include just the current offset. > Ps. what's the next thing we send to the browser? We already sent a > lot of info through the user-agent header... Next year we send our > computers specs to the server so that we get a site that is made for > the performance of our computer? Perhaps you didn't had a need about this in the past. Have you ever tried looking at graphs that include time information in them? Just have a look at mrtg graphs or stock graphs and try to figure what time each point is supposed to represent.