user based pricing was 270 quote

["Steve Richter" <srichter@xxxxxxxxxxxxx>]

-----Original Message-----
From: Jon.Paris@e400.com <Jon.Paris@e400.com>
To: midrange-l@midrange.com <midrange-l@midrange.com>
Date: Wednesday, September 05, 2001 8:24 AM
Subject: Re: MIDRANGE-L digest, Vol 1 #178 - 11 msgs


> >> Some sort of user based pricing would work better for us.
>
>Please don't suggest that to them - it would be a disaster.  I was part of
>a couple of teams that wrestled with that in my IBM days.  Just what _is_ a
>user when it comes to a compiler?
>


Jon,

I am not enthusiastic about it, but base the P5, P10, P20, ... grouping on
the nbr of concurrent users. A user being a point of entry ( ip addr of the
connected to socket, twinax device addr ) or a point of output ( printing to
a printer ). Concurrent is defined as activity within the last 24 hours.

Then price the software and hardware, with a giveaway to me, depending on
the P group.

When the user count is exceeded, roll out cfint to degrade system
performance.

-Steve Richter