Farley, Peter wrote:

>This is a question that has been on my mind for some time now while
perusing responses like yours and others from the IBMer's who monitor this
list.  The consistent message seems to be that mainframe development
resources are very constrained, so only the most important/most useful/most
justified projects (FSVO most) can be undertaken.  (E.G., "functional
stability" for the TSO and ISPF and Rexx components, among many others.)

 

>If IBM will only stay in high-margin businesses (repeated many times by
company leaders over the decades), and the mainframe software/hardware
market is in fact such a business for IBM, then why are IBM's mainframe
development resources so terribly constrained that only "this must be done"
projects can be developed, but not "it would be really nice if ..."
projects?  Why are all possible/desired/required projects not staffed
properly to develop, test and support all of them (or at least most of them,
again FSVO "most")?

 

>It seems to me that IBM is cutting off its nose to spite its face by not
having a much larger complement of development resources to handle far more
of the mainframe customer's needs and wants.  If the mainframe margin is
high, it should be supportable to significantly increase development
resources from the shareholder point of view in order to increase the volume
and thus (in the long term) the profit.  It would appear that short-term
financial gain is still being used to restrict "expenses" like staff in
order to maintain/increase short-term profitability at the expense of
long-term market growth.

 

>Just a question.  I don't really expect an answer, since we'd have to
engage in an NDA-protected dialog with top management and the company board
to get anything approaching a real answer.

 

>My real fear is that the mainframe software/hardware market is not one that
IBM's management and board sees in its long-term future (FSVO "long"), and
that the short-staffing of development resources is a symptom of that
failure to see.

 

Obviously I have no crystal ball, nor connection to the IBM PTB, but I think
the answer here is pretty simple: those areas are considered "mature" and
with little specific growth potential (revenue or  platform), so IBM sees no
reason to invest in them. You and I can argue about the accuracy of the
latter half of this assertion, suggesting drag-along value and so forth, but
the first part is certainly correct. 

 

And this is one of the problems with commercial software (or at least
differences between it and open source): it's not a labor of love, so it's
almost always 75% "done" at best, because "good enough is good enough" and
it's time to move on to the next pasture where the grass is easier to mine
(how's that for mixing metaphors?).

 

It can also be the problem with traditional metrics of "quality": Rexx has
relatively few APARs against it because it's very solid. Some other products
have relatively few APARs against them because they're so horrible that
people can't be bothered. Other products are great and loved and widely used
- and have many, many APARs because folks use them in widely varied ways,
well beyond what their designers anticipated.

 

The PC software model has always struck me as (mostly) more broken than the
mainframe model, because you're driven to add features so you can justify a
new version and thus new revenue, rather than just keeping your customers
happy. Adobe is a perfect example: FrameMaker is at V12, and is full of ways
to crash it by doing tricky things like clicking on the wrong part of the
screen. By Version 12, one would hope that stability was baked in, but no.
And this has been true with every version I've used, from 6 or 7 on up (so
it isn't just one buggy version). Since Adobe is managing to keep selling
upgrades, why fix the bugs? (I do have to wonder how many of those upgrades
are from folks who are stuck with Frame and are hoping for better
stability-but we still buy it!)

 

So as an upper-level manager, how do you decide which products to invest in?
Many APARs or few? Popular or no? Last year you sold $xM of product Y. This
year your Board would like you to sell more. Are you going to do that fixing
bugs and adding incremental features to a product that's mature, or by
putting those same resources on new&shiny? Not even worth asking the
question, really.

 

.phsiii


----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to