[451] in Coldmud discussion meeting

Re: assignment, while (etc)

daemon@ATHENA.MIT.EDU (Sat Sep 24 15:40:57 1994 )

From: stewarta@netcom.com (Alex Stewart)
To: BRANDON@cc.usu.edu
Date: Sat, 24 Sep 1994 00:13:32 -0700 (PDT)
Cc: coldstuff@MIT.EDU
In-Reply-To: <01HHGH2HYPXUDE2OU9@cc.usu.edu> from "BRANDON@cc.usu.edu" at Sep 23, 94 04:06:45 pm

> ColdC (aka C-- (Greg's) is not C, nor is it C++).

Nor should it be.

Greg actually raised a few good points about things..  as did Erik.

However, the argument that you should use a while (1) ... if (...) break; ...
is in my opinion one of the stupidest things anybody could deliberately design
into a language.  If we're going to do hacks like this as standard practice,
why not just get rid of while entirely (since you're not using it's abilities
anyway) and just replace it with a "loop" operation that loops endlessly and do
everything this way?  more consistent and apparently the preferred way to
design programs..  well it's just plain a stupid way to do things in my opinion
if you have any choice, and I'd like to have some choice to do things properly.
(I refuse to be pressured into writing bad code by a bad language design)

Erik is probably right, however, that changing the while statement's syntax
could be more confusing than it's worth, which leaves us with the only other
viable option (as far as I can see):  assignment-expressions.  I actually agree
to some extent with Greg's position against indirect effects of evaluating
expressions, but I don't see any other real alternatives.  sigh.

I do not think that an assignment-expression should be a single equals sign,
however.  That's fine for assignment-statements because there's no room for
confusion, and that's fine for pathetic backwards languages like C++ because
they don't know any better, but this is a real language and it should be
designed properly.

I don't really care what the assignment operator looks like.  I didn't suggest
:= because it's like Pascal, I simply suggested it because it wasn't taken and
its appearance rather reflects its function (as colon is often used to
establish a correlation in human language as well).

I'd actually vote against += and -=, and definitely against ++ and --.  They're
unnecessary complexities to the language and, particularly if they're allowed
to be expressions (which I think they'd have to be if simple-assignment was),
can result in all kinds of really confusing expressions (at least with only
assignment, it requires people to spell things out more so people can tell
what's going on).

And I'd just like to say to all those who can't do anything except compare
ColdC to C, C++, and Pascal: Those languages are all flawed and mangled anyway.
If you're going to compare this to something, at least pick something decent
(and no, I don't have any suggestions.. I've got quite a bit of contempt for
almost all the languages I've come across so far.  That's why I'm making
suggestions based on design principles instead of conformity with some lesser
language).

-R
-------------------------------------------------------------------------------
   Alex Stewart - stewarta@netcom.com - Richelieu @ Diversity University MOO
             ftp://ftp.netcom.com/pub/stewarta/html/stewarta.html