Xref: utzoo comp.mail.mh:1840 comp.sys.sgi:10676
Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!eagle!data.nas.nasa.gov!wilbur!dwallach
From: dwallach@nas.nasa.gov (Dan Wallach)
Newsgroups: comp.mail.mh,comp.sys.sgi
Subject: mh on SGI's: problems getting it to work just right
Message-ID: <1991Jun14.022021.14745@nas.nasa.gov>
Date: 14 Jun 91 02:20:21 GMT
Sender: news@nas.nasa.gov
Organization: NAS Program, NASA Ames Research Center, Moffett Field, CA
Lines: 58

Well, I finally got MH to compile, after hacking a couple files, and
I ended up doing the make by hand in parts, but I have it.  (I'll send
the config's and diff's to anybody who wants them, or post if I get
enough demand).

Problems I'm encountering:

When I try to show an article, I get nothing.  This appears to be mhl's
fault.  My showproc is mhl and my moreproc is less.  Relevant lines from
my .mh_profile are:

showproc:               mhl
moreproc:               less
mhl:        -nobell -noclear
show:       -noheader -form mhl.format -showproc mhl
raw:        -noshowproc -noheader

The following don't work:

% show
% raw | mhl
% raw | mhl -moreproc cat

The following work:

% show | less
% raw
% raw | mhl | less

It appears that mhl notices whether or not stdout happens to be the console
and then proceeds to break.  This only happens when mhl has a moreproc.

Problem #2:

(this is likely an SGI only problem...)
inc doesn't work on /usr/mail/dwallach, but it works just fine on any other
file out there.  inc will hang for a minute and then print:

inc: unable to lock and fopen /usr/mail/dwallach

Relevant file info:
drwxrwxr-x   3 root     mail       20480 Jun 13 19:08 /usr/mail/
-rw-rw----   1 dwallach mail       14016 Jun 13 18:52 /usr/mail/dwallach

FYI: I'm building MH 6.7.1 on a SGI 4D/380 running 3.3.2.  I had to mess
with the makefiles not doing quite the right thing in some places, and
I had to change h/local.h and h/mh.h  (local.h to include the correct
<sys/dir.h> and mh.h to know that SGI's have vfork() )

I needed to change sbr/makedir.c to again include some of the right files.

My MH configuration is essentially SYS5 (i.e.: no BSD42 or BSD43 in there).


Thanks for what help you can give,

Dan Wallach
dwallach@nas.nasa.gov