Re: GTK 2.10 compilation help



> From: Valdis Kletnieks vt edu
...
> > > > 2) Sun's compiler is actually standards-busticated, and defining that 
just
> > > > *seems* to work, but papers over any deficiencies. Also bad news.
> > 
> > Aside:
> 
> > Sun's compilation system is actually the "Gold Standard" of standards
> > conformance.  Since I work for Sun, I don't expect you to take my word for
> > it. If you can find contacts in the standards groups, ask them.
> > 
> > I hope these aren't fighting words.  If you disagree, please keep it to
> > yourself and just dismiss the above as one lunatic's delusion.
> 
> Listed for completeness only.  I was enumerating all the possibilities, and
> later on, said that quick investigation indicated this wasn't the case.

Just for the record, I took no offense.  Let's just say that if you file
a bug on standards conformance with Sun, it will be addressed at a high
priority.  Since I still see standards bugs happening, I know we (Sun)
aren't perfect.  I'm commenting here, because I had another comment below
that's a bit more important, so why not?

> -X.  *that* was it.  Why was I totally unable to find that when I was looking
> at the manpages and *looking* for the compiler switch that did that? :)

I assume that was -Xc, because these days, -Xa is the default.

Perhaps you missed it because:

31: man cc | wc -l
Reformatting page.  Please Wait... done
    6732

That's a short novel.  8^)

You were speaking of older versions.  You should remember a couple of things.

   1)	The compiler version is separate from the OS version.  Each delivers
	some portion of the compilation environment - Almost all the headers
	come with the OS.  Just remember, when analysing a problem, there are
	two versions to worry about.

   2)	Originally, the default was -Xt.  This was the "transitional mode"
   	where (in case of conflict) K&R semantics took presidence over ANSI
   	semantics.  At some point (I forget which *compiler* release), the
   	default became -Xa.  This is ANSI mode where obviously ANSI takes
   	precidence.  Fortunately, these differences are very subtle, unlike
   	the significant differences in -Xc and -Xs.

- Joseph Kowalski




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]