RE: [LiveUser] LiveUser - session_set_cookie_params()
| From: | Koistya `Navin | Date: | Fri, 08 Jul 2005 02:24:49 +0000 |
| Subject: | RE: [LiveUser] LiveUser - session_set_cookie_params() | ||
| References: | 1 | Groups: | php.pear.dev |
| Request: | Send a blank email to pear-dev+get-38485@lists.php.net to get a copy of this message | ||
No-no guys, I was talking about another thing - time() which leads to
unexpected results at least with Internet Explorer 6.0.
Original not working:
======================================================================
(time() + (LIVEUSER_DAY_SECONDS *
$this->options['session_cookie_params']['lifetime'])),
======================================================================
And that code is working for me:
======================================================================
(60 * 60 * 24 * $this->_options['session_cookie_params']['lifetime']),
======================================================================
OR
(86400 * $this->_options['session_cookie_params']['lifetime']),
OR
(LIVEUSER_DAY_SECONDS *
$this->_options['session_cookie_params']['lifetime']),
The problem is that with time() cookies sets not correctly in IE.
Sincerely,
Koistya `Navin
http://navin.biz
-----Original Message-----
From: Lukas Smith [mailto:smith@backendmedia.com]
Arnaud Limbourg wrote:
> I changed it to hardcoded number of seconds, 86400 for a day as it will
> not likely change.
Well I dont understand how hardcoding the value should fix the issue?
> I'm cc'ing LU list as they may have a reason for the constant for which
> I am unaware. For the record, I don't see the added value in using a
> constant (as we know constants are slow) when the number of seconds in a
> day will not change in the near future. Doing the math 60*60*24 at every
> file load seems like a waste.
I put it into a constant because we had an issue where the number was
accidentaly screwed up and I wanted to prevent such mistakes in the
future. However I dont mind recomputing the value in the constant
definition. Actually I dont care eitherway.
I just dont understand why there would be a bug here that is fixed by
replacing the constant.
Eitherway the constant is still used in two other places.