Re: a language is a language

Arthur J. O'Dwyer wrote:
On Wed, 13 Dec 2006, goose wrote:
Arthur J. O'Dwyer wrote:

Forget resource limits; even God's C compiler has to pick a value
for sizeof(void*). :)

Any particular reason that the value has to be less than infinity?

printf("%z", sizeof(void*)); must produce a result.

I agree, but there are no bounds on the result that is printed.
Anything from 1 to an infinite sequence of non-repeating digits
is allowed by the C standard - your implementation may, of course,
differ but the DS9000 can, for the above code segment, print
digits repeatedly and unendingly until the computer is switched
off.

Any limit is defined by the implementation.


See the archives (May 2003) for this dead horse;
I'm going to try not to beat it any more.

I checked, I did not find anything relevant; could you possibly post
a link as I don't remember this coming up in the last 3-6 years.

http://groups.google.com/group/comp.programming/browse_thread/thread/6f8da49abad957a2

ARGH! That thread made me want to dig my eyes out! I don't want
to go so much into detail; this subthread arose because I disagree
with the assertion that someone (sjdevnull, perhaps?) made along
the lines of "The C standard imposes limits, therefore the C language
is limited *even* *if* *the* *implementation* *isn't* *considered*".

I cannot think of any bit of the C language that implies that
an implementation has to have a language-defined limit. Things
like INT_MAX and the rest of its ilk need not be bounded by anything
other than hardware.



Can't we pick a different language this time around? Let's discuss
the Turing-completeness of SML. That way I'm not tempted to jump in.

Actually, I have an even better idea. Let's just decide whether
Hofstadter's FLooP is a programming language. It's obviously not T/C.


Not familiar with that. Nevertheless, I think with this point, I
leave this thread: The C standard does not require a conforming
implementation to have any limits; the limits of any conforming
implementation must at least meet the minimum requirements but
may exceed the minimum requirements towards infinity.


goose,

.

Relevant Pages

  • Re: a language is a language
    ... resource limits. ... C standard places an *upper* limit? ... is considered a programming language without caveats like "if you ...
    (comp.programming)
  • Re: a language is a language
    ... C standard places an *upper* limit? ... The C language does not guarantee infinite limits. ... but most definitions of programming language ...
    (comp.programming)
  • Re: malloc and maximum size
    ... clauses in the standard. ... translate a program which exercises each of the limits /simultaneously/. ... [snip some elaboration of that] ... The purpose of the Standard is to define the language (albeit ...
    (comp.lang.c)
  • Re: truth/falsity of sentences in first-order logic
    ... language to objects and relations of the model) *do* provide a completely formal way to seperate language and meaning. ... Formal logic doesn't concern itself with human limits. ... what "infinity" is. ...
    (sci.logic)
  • Pernicious problem with vfork / qmail / qmail-scanner
    ... My connection limits on tcpserver are set to 60 concurrent connections. ... Resource temporarily unavailable". ... coredumpsize infinity kb ...
    (freebsd-questions)