Re: An Ada Advice Inquiry

[Maciej Sobczak <see.my.homepage@gmail.com>]

On 4 Maj, 05:01, <adawo...@sbcglobal.net> wrote:
> I recently received an inquiry from a colleague I have
> not seen in many years.   He is currently in a position
> where he is asked to advise a client about whether to
> choose Ada or C++ for a project.

Good question.

> He is not a language
> junkie

And it would be very bad if his decisions were made on such
foundations only (I assume the project is not a home-made-just-for-fun
project).
His concerns should be (arbitrary order):

1. Objective of the project.
2. Availability of tools (+price).
3. Availability of workforce (+their price).
4. Cost of training, if necessary.
5. Community support (yes, it's much more important than the support
from vendors).
6. History of previuos projects.
7. Opportunity for reuse of stuff that already exists.
...

And so on.

> Is anyone choosing Ada for new projects?

Not where I work (CERN - at least I can be public with it, there are
no stupid secrets here), which is telling something. If you cannot get
Ada rolling in the nuclear research facility, then in my opinion it's
quite strong argument that the subject is "not easy".

I think that there are only two types of teams where Ada can be
successful:

1. Small team of dedicated junkies (no training cost, etc.) that have
enough energy to reinvent the necessary wheels. They will succeed,
because with the level of competence that such junkies already have
they can succeed with anything, anyway.

2. Larger and already established teams with proven history of
successful Ada projects. Such a team has enough inertia to plow over
any potential problem. And of course, they have already reinvented all
the wheels while working on previous projects, which is a big bonus.

In other words: if you don't have only dedicated Ada junkies on board
and this is a new project, Ada is not the best choice.

--
Maciej Sobczak
http://www.msobczak.com/