[697] in Coldmud discussion meeting

Re: assignment operator

daemon@ATHENA.MIT.EDU (Fri Mar 10 23:57:56 1995 )

Date: Fri, 10 Mar 1995 20:54:07 -0800
To: coldstuff@MIT.EDU
From: jeffpk@netcom.com (Jeff Kesselman)

WARNING: Alex and I have been having a violent discussion about the
assignment operator and he value of adherance to standards.  I didn't
realize it wasn't going to the list so you are comign in in the middle.
Alot of the less emotional and positive thing have already been said.
Sorry.

JK

>> > it?  (hint: it should NOT be '=')
>>
>> Why not?  Your talking about giving at property unique to C (assignment
>> returning a value) and THEN changing it so it doesnt LOOK like C?  It
>> doesn't make alot of sense to me.
>
>Sigh.. I was kinda hoping you'd actually bother to read the messages on the
>list leading up to all this, but I guess I'm gonna have to explain it all
>AGAIN..
>
>The decision to make an assignment-expression operator in C as '=' was one of
>the most innately stupid and boneheaded moves that Kernighan and Ritchie ever
>made.  It causes countless problems, not only comprehension problems but simple
>typos, which can do a remarkable amount of damage (and have been known, on
>occasion, to completely undermine security and even wipe out vital portions of
>the database in MOO, which uses this scheme also).

Guess what? I agree wuth you about K&R. It does NOT change the fatc that
K&R DID make this decision and that it is now a standard in the ONLY
standardized programming language in the indusrty, the one that 99% of new
computr scientists learn first, and the ONLY one to have the kind of
operation you are suggesting (an assignment expression rather then an
assignment opration.)

I kinda hoped YOU would have read that in my last pst.

>Are you actually suggesting that this is in some way a good thing?

Good is irrelevent.  Standradization and ease of learnign is the issue. See
above. In a language named <foo>C, being like C is a good thing, yes.


>
>> Its not a question of "because theya re there."  It a very practical
>> consideration. C, good or bad, has become a defacto standard.
>
>Umm, excuse me, but if you'd look at what you're saying you might notice that
>the term "de facto standard" means _PRECISELY_ "because it's there".  sheesh.

Not because they ae there. Because IT is there. Stop picking on words and
read the sentance ::sheesh::

>
>As I said, stupidity shouldn't be blindly copied in the name of so-called
>"standards",

I disagree.  Standards are NEVER optimal, becaus ethey are by definition
compromises. Standard are however a hgood thing for comprehension, or don't
you agree with that?

> C doesn't have
>dynamic typing, or an inherent OO database-model basis, several important
>constructs (such as 'for' and '?:') are different anyway (and should be), as
>well as many other important changes.

As I said in my original post (which you ignored) ColdC IS different from C
WHERE THOSE DIFFERENCES REFLECT INHEARENT DIFFERENCES IN THE ACTION BEING
PERFROMED.  The varient syntax of "for" makes sense because "for" is a set
operator, not an ordinal one liek in C.  Your sugegstion does NOT make
sense because the operations are identicle.

>Your whining about "but it's not like C"
>is really one of the most insignificant aspects of this whole issue I can think
>of.  It was NEVER exactly like C, and it's not supposed to be!  Changing this
>one operator won't suddenly make people go "oh, I could handle all the other
>changes, but I just can't learn this language if assignment isn't '='"..
>sheesh.

Youa re taking this MUCH too personally and obviously were scared by an
assignment operator at a young age.

Actually your mention of the fact that you never learned to avoid this C
pitfall in your earlier post explains a great deal.  MOST competant C
programmers make this mistake twice and never again.  I consider your
"whining about how it was a bad K&R decision" the msot irrelevent thing in
ths dicussion.  One mans logic is another mans whining irrelevancy.  You'ld
do well to grow up and learn that if you EVER plan to do ANY programmign in
a competant group environment.
(In case you're wondreing I've been programming professionally, both singly
and in groups, for about 10 years now.)

>
>> What you have is a standard.  The three pedals on the floor of my car are
>> probobly far from optimal, but you woudl have ALOT of trouble selling a
>> car that didn't conform to that, and a dispropotionately high number of
>> accidents involving people who didn't know you had switched the pedals
>> around.
>
>Yes, but the your three pedals on the floor aren't arranged in a way that
>causes hundreds of drivers to go careening off cliffs every year.
>
>-R
>
>PS:  Are you aware that all these messages aren't being sent to the list?  This
>appears to be the second time you've responded to something I've said on a list
>(the first time was on MOO-Cows a little while back) and haven't CC'd to the
>list..

Nope, I am not, thank you.   I will repost this, though peopel will be
confused by the sudden thread and joinging in the middle. I wish you had
said so soone. I hATE mailing lists.  Conferences make MUCH more sense...