[427] in SAPr3-news
Re: CO: Iterative calculation of activity price etc.
daemon@ATHENA.MIT.EDU (TWallner)
Thu Jul 27 02:37:29 1995
To: sapr3-news@MIT.EDU
Date: 26 Jul 1995 13:48:01 -0400
From: twallner@aol.com (TWallner)
Reply-To: twallner@aol.com (TWallner)
Hi Corwin,
I have had some experience with the iterative activity price calculation.
On the old R2 system it was a CPU time hog (runtime on a more or less
dedicated 3090 was about one hour for ca. 4000 sender receiver
combinations.
So fare I had few problems on R3. But there I never had the same volume.
In any case, the run time should grow quadratically with the number of
sender receiver combinations.
I remember that with one old 2.1x release (last year) I found something
what I believed was a bug in the combination. That problem seems to have
disappeared with 2.2 (sorry, I forgot the details).
You also might want to check out the 3.0a release notes about this
subject. There are considerably more options in the calculation of
activity prices in release 3.0.
And to your question if the clients where able to understand the
calculation: Do you understand it? If the concepts are explained in an
easy to understand way there should be no problem. If you don't understand
the subject yourself you might have trouble explaining it to others.
Tom Wallner, RSA Companies
