[comp.lang.c] Evolution of the language

dsill@NSWC-OAS.arpa (Dave Sill) (08/31/87)

In article <1025@argus.UUCP>, argus!ken (Kenneth Ng) writes:
>In article <9042@brl-adm.ARPA>, dsill@NSWC-OAS.arpa (Dave Sill) writes:
>> Would you call the changes in C from K&R to dpANS evolutionary?
>> Biological evolution is on clock so slow that thousands of years are
>> required to implement noticeable change.
>
>Biological evolution takes thousands of years?  How about the resistance
>of insects to pesticides like DDT?  Granted some of the insects may have
>already had the resistance, but I'd say that as a whole the insect
>species has evolved to become resistant to it.

Okay, perhaps I should have said "thousands of *generations*".  My
point still stands.  The language is changing at a rate that can
hardly be considered evolutionary.  For truly evolutionary change to
occur, the lifetime of source code would have to be shorter than or
equal to the intervals at which the language is changed.  Until this
is the case, which it won't be any time soon (software never dies),
backward compatibility will be a problem.

K&R consider the obsoleting of the =op assignment operators to be an
example of an evolutionary change to C.  In fact, this may be as close
as software can come to evolution.  It's certainly more evolutionary
than dpANS's function prototypes, which I think represent a
*revolutionary* change.

-Dave

gwyn@brl-smoke.ARPA (Doug Gwyn ) (09/01/87)

In article <9073@brl-adm.ARPA> dsill@NSWC-OAS.arpa (Dave Sill) writes:
>It's certainly more evolutionary
>than dpANS's function prototypes, which I think represent a
>*revolutionary* change.

Without getting into the (probably unresolvable) issue of the exact
distinction between "evolutionary" and "revolutionary", I want to
point out that function prototypes in C are not an invention of
X3J11; they were in fact introduced in Stroustrup's "C with classes",
which evolved into C++.  There had been sufficient experience with
them to convince X3J11 that they were both practical and desirable.
The only difficulty was in figuring out how to "grandfather in" old
code that doesn't use function prototypes.  Now, if X3J11 had simply
outlawed such code in the standard, that WOULD have been revolutionary.

V4007%TEMPLEVM.BITNET@wiscvm.wisc.EDU (Mike Brower) (09/01/87)

Dear Dave:

      Account v4007 of TempleVm is no longer Mike Brower's account.  I
 am Franky Choi and if you are still interested in sending me mail please
 by all means do.  However, I am not an expert in C language.

V4007%TEMPLEVM.BITNET@wiscvm.wisc.EDU (09/01/87)

Dear Doug:
     Mike Brower is no longer with account V4007 TempleVm.  This account is
 now under a new user named Franky Choi.  Franky is will to accept interesting
 mail.

peter@sugar.UUCP (Peter da Silva) (09/03/87)

How many backslashes do you think this subject line will end up with?

And who's putting them in there?
-- 
-- Peter da Silva `-_-' ...!seismo!soma!uhnix1!sugar!peter
--                  U   <--- not a copyrighted cartoon :->