Re: Looking for good Ada95 book

[dewar@merv.cs.nyu.edu (Robert Dewar)]

Mike said

'Your characterization of the books is correct, Robert, though I think
it's going too far to describe the lexical style as "horrible." It
does what it is intended to do.
"

No, I stick by my characterization of horrible. It means that you are
teaching students to use a style which is not the standard style that
we want to encourage for Ada use. People tend to stick with the style
they first learn, so this kind of nonstandard choice causes problems
later on.

The result in my class is that some people use the Feldman style and
some use the more standard style that I (and most other Ada programmers)
use.

I understand, but find unconvincing, Mike's arguments for this nonstandard
style (in which keywords are capitalized).

Obviously tastes vary, but for me this disadvantage is sufficient to 
look at competitive text books that follow a more standard style. This
is obviously a question on which opinions will differ, but for me variation
in lexical style is very annoying, and it is a great advantage if a language
has a pretty standard style, either enforced by the language (COBOL), or
by generally accepted convention (C).

In the case of Ada, the original RM encouraged an ALL_CAPS style for
identifiers that many adopted, but many found intolerable because they
FELT IT WAS LIKE SHOUTING! Consequently we had a big mixture of clashing
styles, even sometimes I saw employees plain refuse to follow company
standards.

With Ada 95, we have something approaching a real consensus on style (use
lower case for keywords and Capitalized_Identifiers_With_Underscores). I
think this consensus is valuable for the community, and I think it is
damaging for a text book in effect to wage a rear guard action against
this consensus.

Mike feels that UPPER case keywords are superior from a pedagogical point
of view. Even if this were true (I don't accept this), it is not enough
to make it acceptable to undermine consensus style issues. Certainly
I don't see people seriously arguing this as an issue in teaching C
(though of cours there are many other legitimate issues when it comes
to teaching C as a first language).

SO, horrible may be too far for Mike, but I stand by it!