Singletary, Mark wrote:

> <>As to the topic of screen abandonment, I am afraid a must strongly 
> disagree.
> While it is theoretically true that in a stateless protocol like http we
> should be able to just abandon a screen, most modern web applications 
> do in fact
> keep a session state associated with the user(hence the log-in) and as 
> a result
> are quite statefull. In statefull systems a load is imposed on the server
> when users abandon a screen. Session maintenance and cleanup require
> significant machine cycles and memory, and UIs that encourage screen 
> abandonment can
> create significant issues for server sizing and concurrent user loadings.
>
Totally true but try telling that to users ;-)  We can encourage them to 
"exit" properly and "cancel" but many of them will just leave if they 
want to abort.  I know I just did it for an accommodation reservation I 
was making.  I proceeded to the point of finding how much it would cost 
then clicked on a discrete exit button to exit and cancel the room that 
was being held.  However, I know many users would have just left the 
website without properly cancelling their transaction.

Whilst we can encourage users to do the right thing, surely systems and 
sessions should be designed to accommodate typical user behaviour e.g. 
automatic time outs to reduce server load?

cheers
Tania Lang

    --------------------------------------------------------------
    Tip of the Day: Quote only what you need from earlier postings
     CHI-WEB: www.sigchi.org/web POSTINGS: mailto:[log in to unmask]
              MODERATORS: mailto:[log in to unmask]
       SUBSCRIPTION CHANGES & FAQ:  www.sigchi.org/web/faq.html
    --------------------------------------------------------------