🚀 go-pugleaf

Following is a letter I wrote to IBM and sent to their technical
support Internet address (cset2@vnet.ibm.com):

Dear Sir/Madam:

I would greatly appreciate if you could forward this message both to
the director of the C/Tools++ project, and to your president.

I am appalled at the lack of response I have experienced trying to
get technical support from you via this email address.  Approximately
two thirds of my messages go unanswered, even after several days.
What kind of way is this to help developers build applications that
will make OS/2 more appealing to the public?

OS/2 is a great operating system, and I'm delighted to be using it,
and developing software for it.  However, although I would like to
say only nice things about it, I have to be honest.  IBM has still
not totally changed.  When I call IBM's phone numbers for the various
departments relating to OS/2 (DAP, etc.), I never speak to anyone who
has any techincal understanding.  Is IBM a marketing company?  We
are technical people with technical needs.  We need you to give us
technical people to talk to.

There must be dozens of different phone numbers and departments
relating to OS/2 where there could be just one.  IBM is a huge
company, but OS/2 is a small part of it.  It would be nice if there
were a central point of information, or, at least, *fewer* points.

I have tentatively invested a large part of my career in OS/2.
Technically, it is an excellent operating system.  I'm a little
scared, though, that IBM will not be able to shed the rest of its
bureaucratic baggage, and that it may return to its ineffectual,
inefficient, and frustrating ways of the past.  Is the excellent
result of the OS/2 development team a result of lasting changes in
style at IBM, or is it just a fluke, and will the hardliners (or the
old habits) take over again?  I do not want to regret my decision to
move to OS/2 next year or two years from now.

Another possible source of regret is if OS/2 does not gain popular
support.  We developers are your allies here.  Please recognize that.
Anything you can do to help us is greatly appreciated, and is in your
interest as well as ours.

There are already many barriers to discourage developers from moving
to OS/2.  Here are some ways to lower them:

1) improved free technical support
2) subsidized developers training seminars and videotapes
3) online technical support software for searching for bugs,
   explanations, etc.

It is important that these not be prohibitively expensive.  If you
view developers as a source of income, and charge the kind of prices
IBM is accustomed to charging, many will not be able to pay the price
of entry.  You may wind up losing much more from the lack of
applications than you would have by funding developer support.  I
applaud the low cost of OS/2 and the special C/Set++ price.  However,
it is not enough.

1) You need to offer much better, and more, technical support.  I recall
attending a user group meeting that featured the Next machine and its
operating system.  The presenter said that WordPerfect was several
months *early* porting to Next than they had estimated.  (Wouldn't
*that* be nice!)  One way you can help us get those applications out
more quickly is to eliminate the tech support bottleneck.  If I have
a problem that is holding me up, your tech support might be my
critical path.  This business of not answering for days, or at all,
is infuriating, and only gives credibility to those who opposed
the use of OS/2 and are eager to say 'I told you so'.

2) Although I have been programming in C for 10 years, PM programming
was quite a shock.  There were so many new concepts to learn.
Although I purchased a couple of books to supplement the manuals, I
still got stuck often.  When the system would hang, or the window
never showed up, those books and manuals were not that helpful.
Although I have learned almost everything I know in other areas from
reading and doing, I would have preferred to learn PM from a teacher.

Employers and clients are not often willing to invest thousands of
dollars in training.  Costs that may not have been prohibitive a few
years ago are unthinkable to them now.  I have been to two training
seminars in the last five years.  The first time, my employer paid
the costs, but the time I was there was unpaid.  The second time, I
bore the complete cost; tuition, transportation, lodging, food, car,
and my time.  Unfortunately, I am not in a position to do that
anymore.

You could do a great service to developers, and to the OS/2 cause, to
offer training at greatly reduced rates.  If you look at the ultimate
effect of this, and not just the initial expenditure, you may find it
to be well worth it.

Remember, when we travel out of town to a seminar, we are already
paying more in out of pocket costs and time than the tuition itself.
Even if you were to offer free training, it would still cost a lot to
attend.  Your training seminars run about $1500 per 5-day course.
Even with a modest discount, this is impossible for most of us to
consider.

A second best alternative would be to put these seminars on videotape.
Although this would not be as good as face-to-face training, it would
cost you a lot less, and would not require out of town travel.

3) I realize the number of OS/2 tech support people that are highly
qualified and available is limited.  Therefore, it would be
helpful (again, to you *and* us) to have some kind of automated
help/advice/troubleshooting program available via telnet and
IBMLink.  (I haven't accessed IBMLink yet, so I don't yet know
what's on it.)

We developers do not need help in running some kind of search or
text retrieval software, which is all some tech support people
do anyway.  Let us do it ourselves, so you can save the human
resources for the truly difficult problems.

**********************************************************************

OS/2's main obstacle to success is switching costs.

By switching costs I mean the sum of all costs to a user or developer
relating to switching to OS/2 from another operating system.

One such cost is the cost of the software itself.  To your credit,
you have reduced the price of many software products.  Software cost
is therefore not that big a barrier.  (That is, unless you decide to
start charging high prices for upgrades, etc. later.)

There are other switching costs though, that are much higher than the
price of the software.  Developer learning time costs thousands of
dollars, and dwarfs the software cost.  I have described three
concrete ways that you could reduce this cost.  Any progress you can
make in this direction will result in more and better application
software for OS/2, and greater incentive for customers to buy and use
it.

Please let me know what your plans are regarding these issues.

Sincerely,
Keith Bennett
-- 
--------------------------------------------------------------------------
Keith Bennett                            Bennett Business Solutions, Inc.
C++/C Software Development               1605 Ingram Terrace
kbennett@access.digex.net                Silver Spring, MD USA 20906-5932