I've observed with interest the recent discussions
about adding an http protocol library to the freeware
collection of resource we have for the Ada community.
It would clearly take a lot of work to do this, but,
having looked closely at the competition now for the
last several months, I think that the result by itself
would not be nearly enough to convince anyone to move
over to Ada. I believe it would be a waste of effort
and resources to undertake a project like this.
As you may know, I have a relationship with Microsoft,
and, as a result, I see a lot of what they are doing.
I also teach a course in Visual Basic at a local
university, and I end up seeing their work even closer.
Today, universities using Microsoft products (there are
over 350 in the Microsoft Academic Cooperative last
I looked) can put resources into the hands of students
that make what we are talking about re the http issue
pale in comparison. Last Sunday, I prepared a demo
program for my class in which I wrote a custom web
browser using Visual Basic. It took under 1 hour, and
I wrote about 20 lines of code. The http, ftp, and
gopher protocols were built into the reusable component
I used at no extra effort on my part, and the basic
functionality of the browser came online less than 15
minutes after I started. My students will probably
take longer to do the same thing, but not that much
longer, and they are college Freshmen for the most part.
Likewise, I also wrote programs that accessed database
files and spreadsheets with equal ease. I wrote programs
that produced full-color graphics also with the same
ease. Here are the writeups, including the source code:
With our current infrastructure, we cannot compete
with something like this. It shows where some of the
$2.5B/year Microsoft is putting into R&D is going.
It also shows where the $55M+ investment in its
Academic Cooperative is going. I have the Microsoft
Developer's Network Library (MSDN) online on my PC,
and it contains over 1.3G of material (1/2 the size
of the ASE) with its host of hypertext manuals and
40,000+ sample programs. It updates every three
months and comes out on DVD and CDROM. My students
paid $50 for their text book, which was bundled with
a learning edition of the compiler and the MSDN library
section on Visual Basic.
I'm an Ada fan as well, but I think Ada has its target
in arenas other than those in the more popular venue.
We need safety critical software for the subways,
airlines, medical systems, and other systems where
lives are at stake. At the part of Lockheed where I
work, I have not heard a peep about moving in any
direction other than Ada for our aircraft, and that's
just because it makes sense.
So, I think that we should focus our energies in the
direction of the safety critical systems. We don't
have enough resources to divide them otherwise. Ada
is already well-enough designed to let it work with
these other technologies (I've already put up a pair
of demos on my university website showing a web-based
interface to Ada and a Visual Basic front-end to an
Ada engine). I think we should let the others evolve
as they are going and we should focus on moving Ada
along its line of strength ... a line that distinguishes
it from the others, and a line where it is needed,
particularly since the others clearly don't fit this
need. If you want to see the interfaces I've mentioned,
here is my university website for them:
Just my opinion,
Richard Conn, ASE and PAL Manager