From: | Georg Bauhaus <rm.dash-bauhaus@futureapps.de> |
Newsgroups: | comp.compilers |
Date: | Fri, 08 Jun 2012 16:16:50 +0200 |
Organization: | Arcor |
References: | 12-03-012 12-03-014 12-06-008 12-06-010 12-06-013 |
Keywords: | i18n |
Posted-Date: | 08 Jun 2012 17:28:19 EDT |
On 07.06.12 14:00, Joshua Cranmer wrote:
> Sure, you can make the editor do things
> for you (this is what happens for APL IIRC), but to most programmers,
> it just seems like you're making their lives hard for no reason.
There is a huge blind spot near ASCII, indeed.
Programmers might find it hard to tackle characters, so they
continue writing in---if I may put it that way---monkey mode:
one key press, one character. (Remember no-dead-keys?).
You can watch them work like an experimenter of behavioral
psychology will observe his animals: stimulus, response.
Engineering does not have to be smarter than that, does it?
After the fact, though, ASCII makes programmers' lives really hard,
because of the effects that the 7-bit table assumption will cause.
This assumption is fixed in the minds of programmers,
and is reinvigorated by defining every new programming language
on top of C's char together with a de facto 7-bit assumption.
A recipe for disaster, making software integration fail,
making languages hard to develop later. Character
set support in Ruby was proof of that; Python means a continuing
effort at getting libraries to work in the presence of the new (!)
str/unicode decisions for version 3.0 (!) of the language.
A similar road block seems to have appeared on the way to ATS v2.
If programming languages trigger adopting characters as
objects of proper data types just like they do with all
other objects, the quality of software will necessarily improve,
because the programmers handling characters will start
to work as carefully as they do with objects of types other
than char.
Maybe there is a pride issue in the way of accepting proper character
sets. Programmers being asked to learn how to use a computer's keyboard.
Embarrassing. The humiliation! And all for turning '<' '-' into that
fancy 'b'? No, for making you aware of plain characters in B%, b,, O,
rC4le play, etc. A language that offers a persuasive way of
making engineers type universal characters will have delivered
a killer feature.
-- Georg
Return to the
comp.compilers page.
Search the
comp.compilers archives again.