Re: ada and robots

[bobduff@world.std.com (Robert A Duff)]

In article <33A7427F.6646@epix.net>,
Matthew S. Whiting <whiting@epix.net> wrote:
>...  I'm not sure how to reconcile the disagreement
>between Jon and Robert with respect to whether 1.5 of the points made
>were legitimate or not, ...

Well, I guess I was being generous.  The ".5" was for "variable length
argument lists".  I was admitting that they can be a nice convenience.
However, I must say that I completely reject the idea that C in
particular is "better" in this regard, since it sacrifices type checking
for variable length argument lists.  And since you can conveniently
concatenate strings for output, the printf example is not compelling.
My position is: it would be nice (a little bit) to have var-length arg
lists, but only if the language did them in a type-safe way.

Anyway, I thought the issue was whether Ada can deal with low-level
programming such as in device drivers.  And the printf example, and
var-length arg lists, have nothing in particular to do with that.

The "1" was for "lack of bindings", and I still agree with your
colleague that this is a problem in choosing Ada.  On the other hand,
it's not *that* bad -- Ada 95 has quite nice facilities for interfacing
to C.

>... but this provides a good reason for me to study
>Ada a little more to make my own judgement.

Good idea.  I'd be interested in hearing what the outcome is --
e.g. whether or not your colleague ever changes his mind, or comes up
with a real example of device-driver-ish code that can't be written
easily in Ada.

- Bob