Re: Sessions proposal

I assume you mean where the user is geographically, where timezones are a factor. But still, what difference does it make where the user is? If you track the login time using the server's localtime, it doesn't matter what timezone they're in. Not the most elegant solution, but it's better than nothing. J André NæSs wrote: > > Maybe I was unclear. The problem is that to allow for 2 hour timeouts I > need to have gc_maxlifetime set to 2 hours, but for 99,9% of all session I > don't need than 15 minutes gc_maxlifetime since they have a cookie > lifetime of 15 minutes. > > Also, I can't log the user out after 15 minutes (it's better to use > cookie_lifetime for that), I haven't got any idea of where he is, so > storing the login time as a session variable won't let me change the > gc_maxlifetime. > > André Næss > > >>