Re: [Groff] geqn problem?

[Werner LEMBERG]

> > Didn't groff used to *always* install links to the groff verions
> > of the "standard" macros (ms, me, mm) with a "g" prefix?  (-mm was
> > available as -mgm) Didn't it recently change to *not* installing
> > these prefixed versions *unless* the system already included other
> > versions of those macros?
> 
> This might be true -- I looked through the documentation and did
> not find a description of the process.

Sigh.  I'm documenting and documenting.  This is a never-ending
story.  Here's my roadmap:

  a) Finish documentation of troff.
  b) Document output devices.
  c) Document preprocessors.
  d) Document installation.
  e) Document macro packages.

Volunteers are welcome to help with b)-e).

> I also did not see any easy way of telling groff during installation
> that the troff macros I use are not in /usr/share/lib/tmac or
> /usr/lib/tmac, but rather in /v/lib/tmac.

The current behaviour of looking up some `standard' places for UNIX
troff macro packages is ugly, very ugly.

To avoid these checks you should say

  ./configure
  tmac_wrap= make

or something similar.

Some solutions to this problem:

  a) Provide a configure option to activate/desactivate this test
     instead of always doing it.

  b) Modify the wrapper scripts (look at the `stamp-wrap' target in
     tmac/Makefile.sub) to call the UNIX troff macro packages only if
     in compatibility mode.

  c) Introduce an environment variable GROFF_SYSTEM_MACROS.  Only if
     set, the UNIX troff macro packages are called.


I prefer a) and/or b) but haven't had time yet to implement it
(i.e. modifying the `stamp-wrap' target).  Additionally, I don't have
access to UNIX troff so I can't validate any changes.  Any help highly
welcome.


    Werner