Is Transcript's English orientation a plus or minus? (rather long)

jbv jbv.silences at Club-Internet.fr
Wed Feb 11 13:55:14 EST 2004


>
> >>  because we have a
> >> language that thinks like we do, not like the compiler does."
>

Well, even if I somehow agree with the above sentence (because I more
or less understand what it implies), I must confess I don't really like
it...

IMHO computers and languages DO NOT THINK. Only programers
(and human beings in general) do. And in the context of computer
programing, they even think better (faster & more efficiently) when the
underlying concepts of programing have been well understood.
I already made that point in a post from yesterday, but surprisingly
nobody commented (which means it was incredibly smart or totally
dumb). Please allow me some more comments on that topic.

In 1980-81 I've been involved (as a programer with some background
in psychology & ergonomics) in an experiment in which a Logo machine
(Seymour Pappert's language) had been introduced in a classroom for
kids of 9 & 10 years of age. It used the french version of Logo, but
nevertheless the syntax was so closed to everyday language that the kids

were able to concentrate only on the basic concepts of programing
(sequential
set of instructions, variables, loops, if-then-else, etc).
They used the machine only a few hours a week, but at the end of the
school
year, most of them completed amazing projects (mostly animated 2D
graphics).

To me, HC (and X-talk in general) is just another offspring of the Logo
concept : it allows beginners to approach (and understand faster) the
basic
concepts behind programing more than cryptic languages like C or Java.
But in all prog. languages, these basic concepts ARE THE SAME.
And the natural feeling of the syntax shouldn't hide the fact that basic

concepts should be well understood before trying to build any serious
project.

The drawback of the above is that the "natural" syntax of x-Talk can
lure
beginners by letting them think that programing has become "plug &
play",
but they might quickly face desillusion (being unable to debug their own

code) or produce code that works more or less, but is so awfull and
ineficient that it becomes useless.

The main point I want to make is that this discussion about introducing
C-like
syntax in x-Talk is totally pointless, especially if the only goal is to
make
Transcript look less amateurish...

Please consider the following : in theory Rev sets the limit of a script
size to
very high level (don't remember exactly, but several Mb anyway, may be
even
more). This raises 2 questions :
- do you really think that a script with several hundreds or thousands
of lines
doesn't need very accurate and careful conception & structure, and of
course
comments (just like any C program needs) ? And don't you think a serious

and solid background as a professional programmer could help ?
- do you really think that the current Rev / MC IDEs allow us to
maintain &
debug scripts with hundreds or thousands of lines ? Of course not... But

CodeWarrior (for instance) does...

So in conclusion, I'm tempted to say that the language itself doesn't
look
amateurish at all (and DOESN'T need to include any C-like syntax).
But OTOH the IDE (especially the scripting & debugging environments)
REALLY look amateurish... And what you gain in productivity because of
the syntax, gets lost at the same time by the poor scripting tools...

Thanks for reading.
JB



More information about the use-livecode mailing list