php3-list | 199903
Date: 03/18/99
- Next message: Jordi Robert-Ribes: "[PHP3] Knowing if page moved permanently?"
- Previous message: Michael A. Smith: "Re: [PHP3] last update"
- In reply to: Kip Keil: "Re: [PHP3] Cookies lifetime"
- Next in thread: James: "RE: [PHP3] Cookies lifetime"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Correct, Kip.
from the Netscape Cookie spec:
<quote>
The date string is formatted as:
Wdy, DD-Mon-YYYY HH:MM:SS GMT
This is based on RFC 822, RFC 850, RFC 1036, and RFC 1123, with the
variations that the only legal time zone is GMT and the separators
between the elements of the date must be dashes.
</quote>
Note the bit about GMT. Dennis I suggest you check the configuration of
your client/server and determine if they know what timezone/planet
they're in/on :).
Kip Keil wrote:
>
> Are you certain? I mean, looking at this purely from a logical perspective, this
> doesn't make any sense.
>
> My server is in MST (GMT-5) and a client logs in from EST (GMT-3) with a cookie set
> for 1hr using time()+3600. According to your theory, the cookie will have expired
> an hours before it was set. Nevertheless, my client gets in and works for an hour
> before the cookie expires.
>
> Just imagine some in Hawaii (GMT-8) trying to log in and use a service on a server
> in Japan (GMT+9). Unless the cookie lifetime is set for 18hr, the Hawaii user will
> never get on. Yet I've never heard of this type of thing happening.
>
> Then, there're the people who buy a computer, take it home, set it up, and never
> bother with the clock settings the first time. Or who pack up and move across
> country....in which case, they don't get nagged to check the time setting.
>
> And what about the change from xST to xDT and vice versa? While note used in a lot
> of places (even some areas of the US have wised up and dropped this now-archaic
> practice), but this would certainly complicate matters.
>
> No, I think the basic practice is that HTTP servers and clients _both_ look to
> GMT. As you've explained it, the HTTP server uses GMT, but the HTTP client uses
> local time. In fact, neither one actually uses GMT (meaning, by direct reference
> to GMT via NIST, et al), but each uses local time to derive GMT. If the local
> clock is off, neither the server nor the browser knows about it.
>
> So, if my server is running at GMT-4:50 (as opposed to GMT-5), and your computer is
> running at GMT+2:10 (as opposed to GMT+2), then your cookie is already 20min old
> when it gets set. And nobody knows. All my server knows is to pass a variable
> that represents an arbitrary mark in time. All your client knows is it received a
> variable that represents an arbitrary mark in time. Neither your client, nor my
> server, knows, nor cares, that the relative period until reaching that arbitrary
> mark in time is different from what the programmer intended.
>
> Now I could be wrong, cuz I haven't gone digging into the specs on this stuff. I'm
> just spewing forth hyperbole based on observation. As I recall, we do have few
> people on this list who have taken the time to get far more intimate with specs
> than have I. Maybe one of these esteemed individuals can shed more light on this.
>
> Kip
>
> Denis Thuillier wrote:
>
> > I finally find the solution.
> >
> > The problem comes from time zone, IE4 and Netscape3 calculate cookie
> > lifetime with GMT time. The server and the remote machine have to be in the
> > same time zone.
> >
> > Example:
> > If the server is set to be in GMT (Greenwich) and the remote machine is set
> > to be GMT (Paris, Madrid) (+1h)
> > the cookie lifetime will be 1h(cookie time) - 1h(Greenwitch-Paris) = 0.
> >
> > (In fact I have set my cookie time()+3700 then 3700-3600=100s=1mn20)
> >
> > -----Message d'origine-----
> > De : "Denis Thuillier" <dthuillier <email protected>>
> > Groupes de discussion : php3.general
> > Date : vendredi 12 mars 1999 15:28
> > Objet : [PHP3] Cookies lifetime
> >
> > Hello to everyone.
> >
> > We are using php3 on a NT server with IIS3.
> > We have noticed that cookies don’t run the same time depending on the
> > navigator and on the computers.
> > The navigators are IE4 and Netscape 3. On some computers, a cookie which is
> > supposed to run for one hour runs in fact for about 1 min 20.
> > The problem doesn’t seem to come from the navigator configuration as the
> > same configuration on two different computers gives different cookies
> > lifetime.
> > If Netscape 4 is used, there is no problem, whatever the machines.
> > If anyone has any suggestions, thanks to let me know.
> > Thank you.
> >
> > Denis Thuillier
> >
> > --
> > PHP 3 Mailing List http://www.php.net/
> > To unsubscribe send an empty message to php3-unsubscribe <email protected>
> > To subscribe to the digest list: php3-digest-subscribe <email protected>
> > For help: php3-help <email protected> Archive:
> > http://www.php.net/mailsearch.php3
> > List administrator: zeev-list-admin <email protected>
> >
> > ---
> >
> > --
> > PHP 3 Mailing List http://www.php.net/
> > To unsubscribe send an empty message to php3-unsubscribe <email protected>
> > To subscribe to the digest list: php3-digest-subscribe <email protected>
> > For help: php3-help <email protected> Archive: http://www.php.net/mailsearch.php3
> > List administrator: zeev-list-admin <email protected>
>
> --
> Kip Keil, Sr. Programmer, V i s i o n N e t
> http://www.vsnet.com | http://kip.vsnet.com
> MC, Ad Hoc P-3055; MC, Advancement T-1022; MC, Ad Hoc T-175
> AA, Ceremonies El-Ku-Ta 520, Great Salt Lake Council, BSA
> --We all learn from history . . .
> ...either by study, or by repetition.
> -- Kip Keil, 1998
Gerry.
-- "Have you passed through the portal of interest yet?"-- PHP 3 Mailing List http://www.php.net/ To unsubscribe send an empty message to php3-unsubscribe <email protected> To subscribe to the digest list: php3-digest-subscribe <email protected> For help: php3-help <email protected> Archive: http://www.php.net/mailsearch.php3 List administrator: zeev-list-admin <email protected>
- Next message: Jordi Robert-Ribes: "[PHP3] Knowing if page moved permanently?"
- Previous message: Michael A. Smith: "Re: [PHP3] last update"
- In reply to: Kip Keil: "Re: [PHP3] Cookies lifetime"
- Next in thread: James: "RE: [PHP3] Cookies lifetime"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

