Re: Software fault forces Ford recall

[A E Lawrence <adrian.lawrence@xxxxxxxxxxxxx>]

Mark Booth wrote:
>
> Almost everyone I know that dislikes Occam says that syntax is
their
> biggest problem with it. All of that LOUD text and the fact that
> indentation is a big part of the syntax. The other thing most
people
> hated at University was being forced to use a folding editor.
> 
> Now while I have always liked these aspects of Occam, I have
felt for a
> long time that what Occam really needed was a C like syntax.
With a few
> changes to the syntax I feel that Occam 3 rebranded as something
like
> Par-C could actually make some headway out of academia.

This old thorny chestnut does seem to divide the world into two camps.
Fanatics, like me (us?), who *love* these aspects, and the majority who
hate them, On enquiry, members of the latter have seldom
taken more than a superficial look before rejecting the ideas.

But how do we account for this? Maybe

1) Upper case
   ~~~~~~~~~~

An excellent way to pick out the key words of the language with almost
no chance of the user wanting to use use them for other purposes. Clear,
simple good software engineering.

Why do people object? Speculation:

a) Association with childhood, and using capitals before learning
"joined up" writing?
b) Association with BASIC, regarded as primitive.

Both of which seem pretty childish in their own way. 

2) Indentation
   ~~~~~~~~~~~

By far the most economical and clean way to include necessary
delimitors.

Used in other languages like Python, of course.

Here the objections are more obvious:-

a) Without folding or a suitable pretty printer (with explicit
annotation for fold boundaries) simply impractical for large sections of
code. Mind you, the same is true for indented curly brackets in C-like
languages which is why editors usually have special facilities to handle
them.

b) Standard tools, particularly unix tools, don't understand the
semantics of the different sorts of white space. And most importantly,
things like emacs cannot fold properly. [Aside: Jamie Packer wrote
folding macros for emacs decades ago, but it was unable to cope with
indented *folds*. Does anyone know if those limitations of emacs have
been lifted?]

3) Folding
   ~~~~~~~

I fail to understand how anyone who has used folding would ever wish to
do anything else. I use it for almost all languages.

Used in things like Mathematica without drawing flack. Why not? Maybe
because objection (a) below does not apply in the Mathematica largely
closed environment.

Objections:-

a) Can't use things like emacs. A really major objection and why I still
write unfolded text for some purposes. This is a killer.....

b) The folding editors that are available are far inferior to emacs in
capability. No I haven't tried all of them, so perhaps there are
exceptions.

c) Standard tools mess up: see (b) under indentation above.

d) I have actually heard people object to folding because they "can't
see all the code at once", but rather have to [ENTER]and [EXIT]! Given
[OPEN] and [CLOSE], this is simply wrong, but even so they must have
extraordinarily long screens. 


Last objection:
~~~~~~~~~~~~~~

Applies to all of the above: it's different! 

Is this diagnosis correct? Have I missed anything significant? Should we
not make these objections explicit when introducing people to decent
syntax :-) Seriously, it is worth trying to uncover the real reasons for
prejudice and exposing the superficiality and shallowness if only to get
an objective view.

Of course a pre-processor is the simple answer to translate between a
vile C-like syntax and sanity, so people can pick and choose. They can
have all their silly clutter of spaghetti code full of curly brackets on
a 20m listing if they really like it. Not to mention 350 #includes and
trying to keep track of all the interactions between them, obscuring the
structure which is so transparent and encouraged in a folded
environment.


Adrian
-- 
Dr A E Lawrence