john mccabe wrote:

> There are a number of things we could write about, most of which have already
> been done at some point or other e.g.:
>
> 1) Dispelling the myths

No, do not write about negative things; no one would be interested.
Besides, none of those myths exist anymore.


> 2) Provide multi-language examples e.g. "I can do this
> in C/C++/Java/Assembler,
> how would I do it in Ada?".

Yes, write these articles. This is the real problem, the true myth
to overcome: we can't do things in Ada that we can do in
Python/C++/Java.

The things that fail to sell Ada in the cases where the programmer
prefers Ada are the things that that programmer does not know how to
do in Ada. Examples are:

  - doing distributed programming as easy as Sun Microsystem's JINI

  - doing web pages as easy as XSQL and Java

  - doing nested lists as easy as Python

  - grabbing keyboard interrupts as easy as C++

  - writing a networked, multi-player game with interactive graphics

  - creating an electronic organ with multiple simultaneous
keystrokes (chords)

  - doing text_io.stream_io in DOS.

Ada has come a long way and is at the one yard line in many areas.
But it needs some things to show people how to use it to make the
touchdown over that last yard.

In all of the above cases except the last, an article with example
code is probably all all that is needed. In the last case, a bug fix
is probably all that is needed, and someone who knows how to do it
could make the bug fix as part of writing the article. However, the
article is needed anyway because many of us have never seen
text_io.stream_io code to read and write binary characters, in
particular, across distributed systems, and would like to see how to
do it.

Not making these touchdowns has forced programming teams, composed
entirely of those who preferred Ada and were permitted to use it, to
use other languages.