Richmond,

I hope I won't trigger any flame war, but I'm afraid I totally disagree with 
your opinions, which I find rather
counter-productive and even ill documented...

>
>
> All computer languages are what they are; computer languages, and as such 
> cannot do anything but resemble human languages in a superficial fashion (for 
> starters, most human languages were not designed). To manipulated computers 
> via a computer language requires an odd sort of non-human logic which must be 
> learnt by any would-be programmer; and, while some languages attempt to 
> obscure that, without that nothing really effective gets done.

Actually, no human language exists per se in nature. no human language 
popped-up out of nowhere.
Almost all human languages have been gradually designed (by social, cultural & 
environmental interactions),
except for things like Esperanto or Volapuk (which aren't used by any human 
group, except for very very
small amounts of afficionados, simply because they're totally artificial & 
abstract constructs that don't fit any
social, cultural nor environmental needs).

Basically a language can be considered as a tool to act and interact socially & 
culturally...
A programing language is a tool to interact with computers, which are parts of 
our environment. Therefore,
there's no reason why a programing language should have a different structure 
than "natural" languages...

The only reasons (IMHO) why a lot of programing languages aren't like natural 
languages are :
- historically computers have been created by mathematicians to process and 
solve math-like problems (the
Manhattan project...). Therefore, as technical resources were limited (and 
political pressure high), they
had to focus on the essential...
- using "cryptic" techniques (hard to understand and intimidating to average 
humans) helped the first computer
companies preserve & extend their monopole on that market, by introducing 
themselves as the "know how's"...
In some ways, MS is still maintaining that old-fashion attitude with their 
undocumented features...
And BTW it's also a well-known feature of human languages (slang) to preserve a 
group from being penetrated
by outsiders...

>
>
> I don't think xTalk is like English; what I do know is that, in some ways it 
> is easier to achieve fairly spectacular results more rapidly than with a 
> language like PASCAL. But, under the hood, its the same thing.
>

when driving your car, you don't need to know what's under the hood (unless it 
breaks down, of course).
Therefore, why should a computer language remain closed to what's under the 
hood ?
Of course, when fine tuning some scripts for better performance, it's usefull 
to keep in mind all internal
registers and other technical details... But that's basically true only for 
assembler, and may be sometimes
with some C code...
These days, when programing an app, you don't move bytes from accumulator A to 
accumulator B...
Instead you "act" at a much higher level (processing sentences, data, figures, 
diagrams...) which is getting
farther & farther from what's under the hood...
Therefore, since the nature of today's programing tasks is getting closer & 
closer to our daily social tasks &
interactions, why should the language we use to perform these programing tasks 
be fundamentally different
from the language we use in our social environment ?

It's the fact that we're still living in a transitional era where bandwidth, 
processing power, storage
capacities still need some improvement to fit all our needs (according to 
Moore's law, in a not-so-far-away
future we'll have enough bw for whatever we want to do); and computer science 
is so young that we still keep
in mind the constraints of the beginning... and Transcript isn't 100% 
english-like...

But again, there's no valuable reason to stick to that situation where 
programing languages have to be computer
languages...

Best,
JB

_______________________________________________
use-revolution mailing list
[email protected]
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-revolution

Reply via email to