Re: Mistaken change in GCC (fwd)

[Phil Edwards <pedwards at disaster dot jaj dot com>]

On Thu, Nov 23, 2000 at 11:07:23AM -0800, Linus Torvalds wrote:
> 
> Now, this is probably going to be considered too revolutionary, but why
> not
> 
>  (a) admit that it's GNU emacs that is ugly, and that some of the onus
>      of maintaining that uglyness should fall on GNU emacs instead of
>      the compiler. 

Hear, hear.

I find it {odd,disturbing,amusing,discouraging} that the GNU Autoconf
manual's rationale section includes the reasoning that, "The fundamental
problem [with cpp-based approaches] is that the C preprocessor was
designed to preprocess C programs, not `Makefile's."  Further reasoning
is available; I picked a quote at random.

So I read that, and then I discover that GNU Emacs uses an extremely
wackyhack to do its work; using GNU Autoconf for a first pass (on
Makefile.in) and then the C preprocessor on a second pass (on the generated
Makefile.c).  And I thought using m4 to create my physics lab reports was
an abuse...

Why is this a bad practice when applied to Imake, but a clever reuse of
tools when applied to GNU Emacs?


Personally, even though I've never helped with cpplib nor cccccccp.c,
I'd like to think that I'm working on a portable, standards-compliant C
compiler, not "the Emacs compiler".  (Didn't we just have this flamewar?
Except then it was the Linux kernel doing things not supported by the new
compiler, and the argument was that GCC was not "the kernel compiler".)



Phil

-- 
pedwards at disaster dot jaj dot com  |  pme at sources dot redhat dot com
devphil at several other less interesting addresses in various dot domains
The gods do not protect fools.  Fools are protected by more capable fools.