Re: [Texmacs-dev] Boehm-gc-texmacs gc behavior

[David Allouche]

On Tue, May 25, 2004 at 10:04:11AM +0200, Joris van der Hoeven wrote:
> 
> On Mon, 24 May 2004, David MENTRE wrote:
> > Joris van der Hoeven <address@hidden> writes:
> > > One strategy would therefore be to write marking routines for
> > > all TeXmacs data types. That it cumbersome, but would allow for
> > > essentially better conservative garbage collection.
> >
> > <deliberately provocative>
> > Another strategy would be to rewrite TeXmacs from scratch in a correct
> > language (Scheme (bigloo), OCaml or common lisp). Those languages are
> > much more suitable to write applications than low level languages
> > without gc like C and C++.
> > </deliberately provocative>
> 
> <deliberately provocative>
> You are right: I probably should write a Mathemagix compiler first.
> Scheme, Ocaml and common lisp are very low level too: they do not
> even support real overloading.
> </deliberately provocative>

<deliberately quarrelsome>
The suggestion was not an encouragement to some more NIH, which you will
probably get wrong again, but reusing the good, tested language
implementations of people who understand how modern computers work and
know a thing about system programming. But I have gathered that reusing
third party components was something you were reluctant to do because it
exposed you to the difference between software engineering, which is
real-lifely messy, and mathematics, where there are no system
integration issues.
</deliberately quarrelsome>

PS: eventually I might find and analyst and try to figure why I enjoy
flaming you sooo much.

-- 
                                                            -- ddaa