RE: Bad coding standards

[Robert Dewar <robert_dewar@my-deja.com>]

In article
<B6A1A9B09E52D31183ED00A0C9E0888C469959@nctswashxchg.
nctswash.navy.mil>,
> Why not?   Seeing how the Ada Reference Manual is
> the "bible" of the Ada language, written by the
> founders (or "pillars") of the language.

No, that's a peculiar description. It is a formal
standard.

>  Wouldn't it be reasonable to assume that the
> examples given in the reference manual would be
> written in a "good" style?

No it would be unreasonable. The examples are
supposed to illustrate the semantics, and the
purpose of illustrations is to show the semantics,
not to suggest how Ada should be written. The
semaphore done as a protected type is a good
example, everyone knows the semantics of semaphores,
so this example nicely illustrates how protected
types work, but it is not intended to suggest that
you should use semaphores for synchronization. There
are other such examples in the RM.

> You know, a reasonable guide.  What's that
> phrase? Oh yeah, a "style guide".

No it is an international standard for the language.
Not a style guide at all.

> Because I think it's ludicrous to assume those
> using the language manual, especially new users,
> wouldn't assume that the "examples" weren't a
> good style to follow.

Well new users coming the RM often form peculiar
conclusions because they do not understand the
nature of a language standard. Indeed, precisely
because the RM is (by comparison with some other
standards) quite readable, they may indeed make
the huge mistake of thinking it is something like
an Ada text book :-)

> I know later they could learn about style guides,
> but if it were another part of the standard, maybe
> the "Style Guide Annex"

Not completely out of the question if there was a
reasonable consensus on style, but there is not.
For example, a fair number of style guides contain
statements that "thou shalt not use feature XXXX
ever". Clearly no "standard" style guide could have
silly rules like that.

> then we would have universal approach.  Why not?
> Many in the  beginning didn't think we could get
> this many people, from different countries, to
> agree, within reason, on a programming language.

A very odd claim. Where does it come from? Many
programming languages have been standardized. The
fact that a language is standardized does not mean
that people agree it should be used for all purposes,
or even for any purposes. I was there throughout
the Ada standardization process, and I never heard
anyone suggest that they did not think it would be
possible to agree on a standard (after all it had
been done many times before for programming
languages).

> I'm not sure about the "Since Ada attempts to
> discourage the "copy principle,""
> part.  I'm not sure where that comes from.

From a fundamental design point of view that dislikes
redundancy in formal definitions.



Sent via Deja.com
http://www.deja.com/