RE: Bye-bye Ada ?

["Robert C. Leif" <rleif@rleif.com>]

Richard et al.
Firstly, Newport Instruments is a user of Ada technology, rather than an Ada
company. However as both a small user and formally the corporate fellow of a
large company, which created a medical device in Ada, I have my own
perspective on the Ada market. 
My approach to medical devices is best served by using XForms as the basis
for the GUI and data presentation. An Ada interface to XForms would allow
for a very simple means to exchange data. I also reported at SIGAda 02 on
the creation of two Ada packages to match XML data-types. The first is a
generic bounded string with character set and the second is ECMA numeric
names in both XML and Ada. This is all in preparation for my next medical
device. 
Unfortunately, the Ada tool vendors have been slow to see where the market
is going. I believe that the use of Ada would be significantly facilitated
by either the addition of Ada functionality to tools, such as Microsoft
Visual Studio, XMLSpy, or Tidy; or the converse, extending present Ada tools
to XML. For instance, the error messages for XMLSpy are pathetic compared
with GNAT and other Ada compilers. Ada tools could be extended to the
development and validation of XML schema. Ada GUI tools should use XML
languages and constructs, as opposed to Microsoft or Linux specific
(resource) files or constructs for screen and output creation. Since these
tools would be dual use including a very hot market, they should be
profitable.
Bob

-----Original Message-----
From: Richard Riehle [mailto:richard@adaworks.com] 
Sent: Monday, February 03, 2003 9:27 AM
To: comp.lang.ada@ada.eu.org
Subject: Re: Bye-bye Ada ?

Marin David Condic wrote:

> Besides, CASE tools and other programming aids are a tough market. There's
a
> highly elastic demand and a small number of people (with respect to more
> general software) who can actually make use of the end product. It would
be
> far better to pick some more general industry (like auto repair shops or
> beauty parlors or something "professional" like doctors or lawyers) and
> research what they'd potentially be able to use to automate their business

This is a sensible approach.   There is very little money in developing yet
more tools for creating software.  And that money is being drained away
by the open-source movement.

> Develop something for a narrower industry that has a large software
> component and some other sort of value-added product (like an integrated
> cash register or some on-line service - that's where you make the *real*
> money) and you've got something where the big players aren't going to
drive
> you out of the market.

This is called  a "vertical marketing" approach.  Develop expertise in a
particular
industry, learn its demands, its needs, and its way of doing business.
Then,
build
a product that is easier to use, more reliable, and more closely mapped to
the
way that industry does business.   Keep the product current as the
technology
advances.   Keep the code portable for variations in operating system
availability.
Robert Lief was on the right track with his medical equipment in Ada
approach.
I'm not sure whether he ever got very far with it.

I have some friends in Silicon Valley who specialize in developing software
for small medical devices on the I-8051.  They are experts at this after
many
years of doing it.  They quietly make money, don't have any intention of
going
public, love what they do, and have a clientele of satisfied and repeating
business.

Ada, although not hosted on the I-8051, could be used to serve a similar
market.  A lot of lab equipment needs larger processors now.    There are
robotic assemby lines that need to run on larger processors.  When one looks
around carefully, it seems there is end to the opportunity for creating new
software to upgrade what is already in place, and to sieze the market with
reliable software in embedded industrial applications.

The problem is that too many people with Ada expertise are not of an
entrepreneurial bent.   Long years in doing software by contract has
stunted their ability to do anything except on-spec.   The large users
of Ada such as Lockheed, CSC, Raytheon, etc., simply don't have
the kind of management that understands free-market risk-taking.  I
recall a conversation with one high-ranking official of one of the
large DoD software developers.  He asked what kind of commercial
opportunities I saw for Ada.   I listed several.  His reply astonished
me.  "But who will fund it?"    My answer, stunned him.  "No one.
You need to fund it yourself once you determine there is a market."

Richard Riehle