Re: Printing Enum Variable Re: Linux World

["Michael Young" <nobody@all.org>]

Larry Kilgallen <kilgallen@eisner.decus.org> wrote in message
news:1999Mar7.074102.1@eisner...
> While those participating in this forum may be atypical, there is
> a belief that most of those with a C++ compiler are programming in
> C.  In that sense, this backward compatibility may be the greatest
> technical weakness of C++, as well as its greatest marketing strength
> (selling lots of licenses to people who don't really use the
features).

I agree wholeheartedly that C encourages certain poor programming
practices. However, Ada is not the only solution to these problem. I
hear in your postings: "Ada is better than C; C++ supports features and
practices we recognize as evil in C; therefore, C++ is bad. OTOH, Ada
rejects all of C; therefore Ada is good." That's pretty convoluted,
don't you think?

Ada can stand on its own merits, as any successful language must. To
hear your strong denials says otherwise, so I must be wasting my time
and effort here. After all, C++ already does everything I need and want
from a language (well, almost everything), in a way that I am more than
happy and comfortable with. What do I need with another "better C than
C"?

> > While I'm not especially fond of the old SDK, its structure looks
> > surprisingly like object use in Ada today. Instead of the more
succinct
> > C++/Java style aWindow.GetDC(), we see GetDC(aWindow).
>
> I am not sure how one views the first as more succinct, but my own
> feelings are more swayed by type-safety than syntax.

I was pointing out that the SDK exposed an object structured interface,
implemented in C, well before object oriented languages were seen
outside academia. I only meant to say that, because of the similarity in
syntax, this should have been more evident to Ada programmers than to
C++ or Java programmers. With that in mind, I found it especially odd to
see it bashed (along with everything not implemented in Ada).

> > Anyway, I'm not sure which you were bashing: C in its infancy, or
the
> > old SDK. Both were laudable in their days. I expect I might say the
same
> > of Ada95 in some not so distant future.
>
> If you think the days of that SDK are over, I think you are quite
> mistaken.  The "latest released" is not at all equivalent to "what
> people use".

There's a certain amount of baggage associated with backward
compatibility. Perhaps some day, Ada will reach a point of maturity
where this becomes a concern.

Also, I didn't mean to convey that the SDK has changed dramatically. It
hasn't. Most interaction in new code, though, is made through the MFC
(Microsoft's "thick binding" for the SDK).

In very large systems, interface stability is a primary requirement, not
merely an inconvenient liability. Many here are quick to point out that
"large systems" are the rightful domain of Ada. I therefore expected to
find in this crowd some understanding of the legacy issues involved. As
before with the SDK "object" syntax, I found the bashing unwarranted and
misguided.

Anyway, I came here to learn more about Ada, not to preach about large
system development. I'll refrain from further postings of this nature,
and hope you'll likewise use restraint in speculating about the
weaknesses of C++; I presume I already know more about those than you.

Michael.