Tom Jackson said:
> If something like Ns_SetIGet exists, maybe use that,
Good idea, it does and I will.

> I know you were not trying to follow the cookie RFC's on this module, but
> maybe point that out so folks don't think that any cookie value will work
I'd say it is a matter of pointing it out. I'd hate to do my own escaping
as they may well end up setting cookies using other mechanisms and then
have them unreadable, or v.v.

> it possible to update a set-cookie header? I chose to delete all cookies
I don't see why not and I probably should. I'd expect the last one set
would be the one the client will use anyway, but better safe than sorry.

> What are the limitations of session data? For one they expire if the
> server restarts. I understand that this is by design in this case,
> but of course sessions which rely on persistent data will fail.
That is something I have though about as well. My problem with it is that
sessions should be fast and avoiding slow storage. On the other hand, I do
see the value in having sessions last beyond restarts, even if I'd never
write my app that way. :)

I'll look into it. I'll probably do something like mark the session dirty
if a change has been made and register a trace to write it to
'modules/nssession/$session_id'. If a request comes in for a key that I
don't have in memory, I'll look it up in the file system first. Enabling
this would be a config option obviously.

I'd say I am probably best off just opening and closing files, that way I
never run out of handles and I don't have to serialize this. (just create
a $conn_id.tmp file and rename it to $session_id when done)

Shouldn't take that long to implement.

Thanks for the feedback!

Cheers,
Bas.


--
AOLserver - http://www.aolserver.com/

To Remove yourself from this list, simply send an email to <[EMAIL PROTECTED]> 
with the
body of "SIGNOFF AOLSERVER" in the email message. You can leave the Subject: 
field of your email blank.

Reply via email to