Duane Percox wrote:
 
>As I understand it, the r&d efforts are funded through new system sales
>and not support revenue. This creates an interesting problem for a mature
>stable platform like the HP 3000. As new sales growth flattens out there
>is not enough $$ to keep r&d going at a rate to keep the platform current
>with technology. This then influences the growth rate negatively which
>lowers r&d $$. A downward spiral of epic proportion.
 
>IMNSHO this funding model guarantees a fast decline in any product that
>begins to see its sales flatten out...........
 
Duane is exactly right.   There is and has been a fundamental
flaw in HP's business model in this area.  As long as someone
is using a product, be it the OS or a subsystem, the need for
enhancements *never* goes away.  Some products may
"stabilize" more than others, but in the real world the idea that
you can freeze a software product and have it remain viable
for the long term doesn't fly.
 
>Why couldn't we (the users, Interex Advocacy, etc) propose to HP that they
>slightly adjust their funding model. ........
 
>Simple funding model
 --------------------
>Add $50 to every HP 3000 annual support contract and give this $$
>to csy for continued r&d on the platform.
>If there are 50,000 supported systems then this will generate a
>tidy sum of 2.5 million $$ per year.
 
Heck, we would pay $100 a month, if we could be *very sure* that
the additional funds would all go to CSY and the compiler labs.
SIDEBAR:
Note that SIGRAPID pressed HP to increase the support fee for
Transact from $50 to $100 a month a couple years ago;  HP
declined to do so.
 
I don't have any inside knowledge of just what the total number
of HP 3000's are still on support around the world, but I would
bet some money that it's less than 50,000...  However,
your simple funding model is still completely valid...  It wouldn't
take much from each site to generate the same $2.5M in any
case......  Also note that $2.5M is probably the cost per year
of 20-25 HP software engineers, when you include overhead.
 
Which is why I like my number:  An average of $100 per month
from each site (sliding scale depending on system size) =
$1200 per year.   Say there are 30,000 HP 3000's still on support.
That would yield $36,000,000.... Now we are talking like 360 more
HP software engineers;  that would be enough people to do
some serious work on MPE, Image, and the compilers.....  Well,
there would be some carrying and processing "fees" I suppose,
but since this would go on existing contracts the overhead to
collect the extra dollars should be minimal.  And given that HP
is supposed to be a computer company, I would hope they
could find some reasonably automated way to channel the $$
into Harry Sterling's cash box...
 
.....  Remember, the cost of an extra Response Center caller is
$100 per month.  Getting some more serious R&D on the 3000
is certainly worth as much to us as having an extra RC caller.
We have invested several million dollars in developing and
deploying our HP 3000 system... Another ~$100 a month to
help insure its continued viability for the long term is a real
easy call.......
 
Ken Sletten