On 6/10/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
I agree with Tijnema he's hit the nail on the head. And the inability to handle
daylight time really is a big potential snag - who in North America isn't on
daylight time at some point in the year? If the timezone data is unreliable
then no thinking developer will use it. The question then becomes: As a
developer why WOULD you use unreliable timezone data?
Sincerely,
Rob
Yes, I think that you can compare it with the Referrer header, it's
unreliable too, and which developer uses it these days?
Tijnema
>
> On 6/10/07, Stefanos Harhalakis <[EMAIL PROTECTED]> wrote:
> > On Sunday 10 June 2007, Tijnema wrote:
> > > To get back to the point, I think that the timezone should be defined
> > > on what time it actually is at his PC, and what time it is on
> > > time.nist.gov for example, and not lookng at some setting... Timezone
> > > setting is often wrong, people just update their time to match the
> > > time of their watch. This sets the UTC time wrong too, but still
> > > displays the right time to the user. So if you compare this time to
> > > time on web servers that are allways right, and you compare the
> > > difference between that, then you know the timezone that is probably
> > > right.
> >
> > Timezone is a property of the user and some times of the session. It is not
> > a property of a system. Different users of a system may use different
> > timezones (unless they use Windows where they can't).
> >
> > Anyway, I strongly believe that the issue of providing the correct timezone
> > should be a concern of the end user and the browser. Any error checking and
> > workarounds should be performed by those two and not by the server side
> > scripts.
> >
> > In any way, there is no guarantee that the timezone information provided by
> > clients will be correct. That's why it should only be used for informational
> > purposes and not for security etc.
>
> Sure, but if this setting will be incorrect for 90-95% of the time,
> then there won't be a lot people that are actually gonna use it,
> because why do we want to know information that is probably wrong?
>
> Tijnema
>
>
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php
