g++/groff for i386? (emacs, and other things too)

[John Wilkes]

In article <1990Jul10.122545.6963 at athena.mit.edu>, dquah at athena.mit.edu
(Danny Quah) writes:
> >>
> >>Since V/386 is so prevalent on the net now, I'm wondering if
> >
> >Time to pound on the sources again...
> 
> Why keep regrinding the wheel (or "pound on the sources"?)
> This just doesn't seem to be enabling us as a community.

About a week ago, I posted to this forum asking about problems building
emacs ("ESIX neophyte has questions") and most of the response said
something like, "Use COFF and the standard unexec().  It should work." 
Well, it does not.  I have tried various different options in the s- and
m- files, taught sysdep.c that TIOCGWINSZ and TIOCGETC need to be
undef'ed for ESIX, and I am now fighting with process.c.

Folks, the standard emacs distribution WILL NOT BUILD ON ESIX without
some source hacking.  I cannot believe that others have not been here
before me.  Why is there no "s-esix.h" file, or a set of patches?  Am I
the only person in the universe to have these problems?

To anyone who has actually done it:  how do I build a dumpable emacs
under ESIX?  Answers along the lines of, "The standard distribution
should work," are not correct, so don't bother.

Thanks.

John Wilkes

wilkes at mips.com   -OR-   {ames, decwrl, pyramid}!mips!wilkes