Re: Who should we contact for a proposed new feature of gcc, g77 and g++?

[craig at jcb-sc dot com]

>Specifically, we would like to know if it's acceptable that the new feature
>we provide becomes a standard feature of gcc and g++. In the case of
>FORTRAN we modified f2c just because in that case a lexer file was publicly
>available: we would like to do the same thing with g77, but we need to
>contact an expert.

I can't look at your demo (I run Lynx, not Communicator, on my machines
here at home), but it shouldn't be too much trouble to modify g77 the
way you want.  Whether we'd want to make that a standard feature is
another question -- it really depends on the quality of the patch, the
general usefulness of it, and so on.  Particularly of concern is the
fact that the lexer is probably going to be rewritten as part of the
long-delayed 0.6 rewrite.  If your patch is clean and simple enough,
that shouldn't be much of a problem, but it's hard to say.

One thing that might make things easier with g77 is that all tokens
formed by the lexer are passed through a single point, a routine
named ffelex_send_token_.  That might change in the new rewrite, but
my thinking is currently that it'd probably be changed to send N tokens,
rather than just one at a time.  It's hard to say, since I haven't
yet done the research necessary to determine what the best design is
for reliability, performance, support of wanted features, long-term
maintenance, and so on.

Note that the level of cooperation you get from free-software enthusiasts
and volunteers (like myself) might be improved by supporting free-software
browsers (like Lynx) in your project.  In particular, I, and the Free
Software Foundation as well, support the "Best Viewed By Any Browser"
campaign (http://www.anybrowser.org/campaign/).

        tq vm, (burley)