Thats what's confusing to me is that I just realized (after coding the servlet)
that the getLastAccessedTime() would be reflective if a *single* user is
accessing this servlet, but it would have to be reflective of each user.
What I need to do is access a servlet MyServlet?Command=New, which when
accessed for the first time, would save the time, then when its accessed a
second time go MyServlet?Command=Old and the output should be the time that has
passed in seconds.
- Jon
Hans Bergsten wrote:
> Jason Hunter wrote:
> >
> > Hans Bergsten wrote:
> >
> > > "Returns the last time the client sent a request carrying the
> > > identifier assigned to the session, or -1 if the session is new."
> >
> > It's kind of like when it *is* Friday and you say "last Friday" which
> > one are you referring to? :-)
> >
> > > To me this means it has to be reset every time the engine receives a
> > > request that includes the session ID; the current request is the "last
> > > time" a request for the session was received.
> > >
> > > Maybe I'm missing something here, so please correct me if I'm wrong.
> > > To return the time for the previous request (n-1) the engine must have
> > > "look back" feature, i.e. remember when the previous request was
> > > received and set the last accessed time to this value when it
> > > receives the current request. That sounds a bit convoluted to me.
> >
> > The implementation has to a bit convoluted (involving two vars), but
> > n-1 is the only thing that's useful, and it's the way JSDK 2.1 is
> > implemented. When the spec's unclear we should go with the JSDK
> > reference impl.
> >
> > BTW, if the server returned the time at which you *just* called
> > getSession() then the -1 wouldn't make much sense.
>
> Well, if the session is new no request with this session ID has been
> received so -1 would be appropriate. Note that it's not calling
> getSession that sets the time stamp, it's when the request with a
> valid session ID is received.
>
> Anyway, if this is the way it's implemented in the JSDK I'll buy your
> argument, but it needs to be clarified in the spec.
> James, are you listening?
>
> --
> Hans Bergsten [EMAIL PROTECTED]
> Gefion Software http://www.gefionsoftware.com
>
> ___________________________________________________________________________
> To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
> of the message "signoff SERVLET-INTEREST".
>
> Archives: http://archives.java.sun.com/archives/servlet-interest.html
> Resources: http://java.sun.com/products/servlet/external-resources.html
> LISTSERV Help: http://www.lsoft.com/manuals/user/user.html
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Jon Baer / MTV Networks
1515 Broadway - 29th Floor
New York, NY 10036
T. 212-846-5984
F. 212-846-1801
E. [EMAIL PROTECTED]
=-=-=-=-=-=-=-=-=-=-=-=-=-=-
___________________________________________________________________________
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff SERVLET-INTEREST".
Archives: http://archives.java.sun.com/archives/servlet-interest.html
Resources: http://java.sun.com/products/servlet/external-resources.html
LISTSERV Help: http://www.lsoft.com/manuals/user/user.html
