Re: [FIXES] 2.1.125

Ely Wilson (plexus@ionet.net)
Tue, 13 Oct 1998 03:04:07 -0700 (MST)


On Mon, 12 Oct 1998, Mathew G Monroe wrote:

> Excerpts from internet.computing.linux-kernel: 10-Oct-98 Re: [FIXES]
> 2.1.125 by Oliver Xymoron@waste.org
> > > [btw: can someone explain me why in linux kernel I can't declare var
> > > inside of the proc, but instead it has to be at beginning of it. I think
> > > in 'normal' c programs you can do that]
> >
> > Because you've been writing your 'normal' C programs with a C++ compiler.
> > In ANSI C, variable declarations must preceed all other statements inside
> > a given scope. Further, variable initialization is limited to simple
> > assignments.
> >
> > This is one of the few features of C++ that would apply well to the
> > kernel, I think - there are numerous multiple-page functions that could be
> > clarified by local declarations. The kernel already heavily relies on
> > inline functions, which are standard in C++ but not in C.
>
> C9X drafts have included this feature for quite some time. So hopeful
> sometime next year when C9X becomes standard this can be used. Now that
> brings up a question, what do people thing about using it?
>

I think it is bad programming practice to declare variable in such a
manner. It makes variable declaration harder to track if anything. What
next? Dynamic deallocation in the middle of a proc using delete? Isn't
that C++ as well <cough>.

C was designed and is rigid for good reason, please don't tell me some
committee has decided that variable declarations can happen anywhere in a
C script. "ugh"

----------- Developer and Network/Systems Engineer
----- ely - Agape Specialty Products, Phoenix AZ
----------- <plexus@ionet.net>

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/