On Mon, 24 Jul 2000, Ury Segal wrote:
> > Umm..."b" is always supported, and the default mode is *always* text. It
> > just so happens that on UNIX, things happen to work even if you forget the
> > "b".
>
> So what you are saying that there is no standard meaning for
> "b" and "t". Exactly my point: No standartization in C for the
> most abvious and used functions.
No, I meant there is a standard meaning for "b" and "t": one gurantees
binary files, one gurantees a list of lines. Use the wrong one at your own
risk: just like *((int *)5000) = 1
> Who cares what is the minimum length?
Anyone who says "hmmm....this will need to hold at least 1 billion...I can
use a long and be safe"
> WHAT IS
> THE STANDARD ?
I've seen few standards which don't mandate implementation dependant
behvaviour. Just like undocumented APIs, if you rely on it, you deserve
getting snookered.
> >
> > > *) Is the "long long" datatype signed or unsigned by default ?
> > > Can you read my the "standard" on this ?
> >
> > "long long" isn't part of the C standard. It is part of the upcoming
> > C9x standard, however.
>
> What do we care? Most compilers have it, since almost
> no compiler followes the standard.
No, most compilers have *extensions* to the standard. Again, use
extensions at your own risk. You can program beautiful C without any
extensions.
> > But "far" isn't part of the standard.
>
> So why is it widely supported? Again, C compilers
> DO NOT follows the standard, exaclty like C++.
Extensions to the standards are different -- let's define "conform to the
standard" as "compiles any standard following program to something to give
the semantics mandated by the standard". C++ compilers do *not* conform to
this definitions, C compilers do (modulu a few minor bugs which quite a
few years of C coding left me completely unaware of).
> I know C, thank you.
Then why did you ask?
> So you agree with me, every C compiler have another
> version of C. `What a wonderfull, standartized language.
No, every C compilers handles your mistakes differently. That's a quality
of implementation issue. No 2 ADA compilers optimize the same either, no
give error messages which are completely the same. Live with it.
> Who care about the solution? I am showing you example of
> miss-standartization, nd you give me obvious, useless
> "solutions". Those #pragma-s are needed, you cannot
> just "not use" it.
I haven't needed it in well over 5 years of C coding. Apparently, it's not
needed that much.
> Please, stop with the hot air. Give EXAMPLES.
OK, it took me 2 months to port template code from MSVC to SunWS because
SunWS didn't support some weird corner case in templates.
> >. C, OTOH, is
> > pretty well implemented on all of them.
>
> Pretty well, but not from a standartization POV.
Will you stop the BS? I know you are intellegient enough to know the
difference between extension to the standard and problem in following the
standard, and I hate to see you confusing everyone here.
--
Moshe Zadka <[EMAIL PROTECTED]>
There is no IGLU cabal.
http://advogato.org/person/moshez
=================================================================
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word "unsubscribe" in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]
