Re: Pre-Ada95 books still worth reading

[wrp <i3text@gmail.com>]

On Jul 14, 2:13 pm, wrp <i3t...@gmail.com> wrote:
> ...I have combed through the older
> titles, looking for books that might still be worthwhile reading for
> the Ada95/05/12 programmer.

It turns out that somebody around here did have some of these books,
so I've been able to review them.

Ford. 1986. Scientific Ada.
    Although scientific computing is not my area, I found this book
really interesting. It gives an in-depth discussion of issues in
creating numerical libraries and the use of Ada 83 for that purpose.
Several of the contributors worked on NAG libraries in Fortran or
Algol 68, so they had a lot of practical knowledge to draw on. They
were writing at a time when people thought that Fortran was going to
fade away and soon everyone would be using Ada, so they spent a lot of
time discussing aspects in which they thought Ada wasn't as good as
Fortran. Their biggest complaint was about access types, but I think
later Ada answered those complaints. They have several other concerns
that still seem relevant, though.
    For the general programmer, Ch. 5 "Ada and other Scientific
Languages: A Critique", will be the most interesting part. It briefly
compares Ada 83 and Fortran 77. In short, they say Ada is much nicer
for general programming but much worse for numerics. Of course, modern
Fortran is little like F77.

Gautier. 1990. Software Reuse with Ada.
    This is about 200 pages of small print, containing contributed
papers from a workshop on the subject of the title. The content is for
the most part general observations you could find in any text on
software engineering. About half the book describes usage of
particular Ada features. While the information may be some use, my
impression was that most of the advice was obvious to an experienced
programmer. There are also many criticisms of Ada, but they seem to be
things that were addressed in Ada 95.

Gehani. 1984. Ada: Concurrent Programming.
    Well, writing is choppy, coverage is sketchy, code is old-style,
and he uses some obsolete features, but Gehani occasionally presents
an idea more clearly than do Burns & Wellings in _Concurrent and Real-
Time Programming in Ada_. Some interesting references in the annotated
bibliography.

Habermann. 1983. Ada for Experienced Programmers.
    The goal of this book is to present major features of Ada to
experienced Pascal programmers. To that end, it has 16 chapters,
following the format 1) present a problem, 2) discuss how to solve it
in Pascal, 3) discuss how to solve it in Ada. The writing is not
polished and explanations are minimal. Even if I knew Pascal well, I
wouldn't choose this book to learn anything about Ada. Also, the
quality of the Ada code doesn't seem great, maybe because it's pre-Ada
83.

Keeffe. 1985. PULSE: An Ada-based Distributed Operating System.
    This is a sketchy description of a project to create a Unix-like
distributed operation system with Ada. They fudged quite a bit on the
Ada part, as they wrote the kernel in C for efficiency reasons. There
are only two subjects on which they get into interesting detail. First
is their distributed file system, which turned out to give very poor
performance relative to Unix. The other is their great dissatisfaction
with Ada's tasking model, which they discuss in a 35-page appendix. I
think, but am not sure, that Ada 2005 dealt with their issues.