Re: Counting users

Leif,

>===> I'll adopt a simple operational definition:
>If any of my application's programs are on your invocation
>stack, you are using my application. what can be simpler?

This would be fairly easy to implement too, just by having the programs place a
non-exclusive lock on some object dedicated to the purpose.  Comparing the lock
count to the number of authorized users is then simple.

But Ron has already stated he is willing to be more lienient than this, and
allow multiple sessions on the same physical device to have the application be
in the call stack and still allow it to count as one user.  He is trying to find
a reliable way to give that same flexibility to multi-session dumb terminals.

The only two ways I can see to approximate this are:

 1) Assume jobs with the same user ID on the same WS ctl/port as one device

or
 
 2) Same as above, but further check the device serial numbers for a match

However, many older and/or non-IBM displays will not report their serial number,
so the system reports it as 00-00000.  Should these still count as a single
device, or do you say it only counts as one user if you have a terminal capable
of reporting a serial number?

I suppose the argument can be made that Ron is already being lienient in trying
to detect multiple sessions, and that it is not his fault if your terminals are
not capable of reporting a serial number.

Doug

+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---