Re: Compiling lposix.c (patch)

[David Jones <drj@...>]

On 5 Oct 2006, at 18:51, Sam Roberts wrote:

> On Thu, Oct 05, 2006 at 11:46:56AM +0100, David Jones wrote:
> > Indeed.  It would seem, from a bit of googling, that CLK_TCK went
> > (from the standards) a long time ago.  sysconf is a crucial
>
> "the standards"? X/Open defines CLK_TCK, warning that it might not  
> be a
> compile time constant (which it isn't on linux). The Single Unix
> Specification, v2 also continues to define it, as far as I can tell.

Wasn't SUSV2 replaced like _ages_ ago?  Almost as long as c99  
"replaced" c90? :)

Care to quote chapter and verse on that?

I'm slowly losing my grip (happily) between the exact relationship  
between SUS, POSIX, and Open Group.  I note that:

http://www.opengroup.org/onlinepubs/007904875/

"The Open Group Base Specification Issue 6" does note have CLK_TCK  
and notes (Change Issue 6 on the history section for <time.h>) that:  
"The symbolic name CLK_TCK is removed".  Apparently in order to align  
the spec with ISO POSIX-1:1996 .

> And more importantly, nobody has posted here that *it doesn't work*.

It doesn't work!  On OS X 10.4.

And I tried the _POSIX_C_SOURCE=1 voodoo.

> > cornerstone of POSIX, not an arcane interface.  Well, I suppose it
> > might be both. :)
>
> Guys, just because a new standard comes along, it doesn't mean that  
> you
> should change your code so it only works on new systems, or that all
> the old code is suddenly "wrong"!
>
> Notice how the lua source doesn't use "//"-style comments? Those are
> ANSI C now, but when you build lua for some grotty cell phone with an
> ancient C compiler, you will be thankful lua doesn't use //, even  
> if it
> is ANSI C.

And on a similar note, can we stop calling ISO C "ANSI C".  It wasn't  
long after the publication of ISO C (ISO 9899) that ANSI replaced  
their standard with the one from ISO.

Personally the more I look at more standards the more I see them  
becoming less and less relevant.  As you point out, you would be,  
let's say adventurous, to start making a widely deployed C program  
that relied on ISO 9899:1999 features (or even 9899:1995 features  
come to that).

> p.s.
>
>   Though some hasty zealots cry "not so; the Millenium is come, and  
> this
>   saying is obsolete and no longer need be supported", verily there be
>   many, many ancient systems in the world, and it is the decree of the
>   dreaded god Murphy that thy next employment just might be on one.
>   While thou sleepest, he plotteth against thee. Awake and take care.
>
> 	- The annotation to the 9th commandment for C programmers
>
> Maybe your external identifiers can be longer than 6 characters,  
> though :-)

I once had the pleasure of porting code to a system that was nearly  
that bad.  7 characters were significant in external identifiers.   
Not case sensitive mind.

drj