VC>> Despite using the libs from 3.03b release as suggested by Bo I still
 VC>> get seg faults when entering menus options such as file areas or
 VC>> message areas.

 BJ> Have you compiled any MEX code?  If so, remove the 
 BJ> compiled mex files.
 BJ> That can cause seg faults all over the place because a lot of the
 BJ> default BBS displayed info is actually output from Mex code, and the
 BJ> mex virtual machine is broken, broken, broken, 
 BJ> broken....  But the Mex
 BJ> compiler does execute when run.....  So, take a look in things like
 BJ> etc\m and other places, and see if temporarily moving certain files
 BJ> (say to a temporary subdirectory) keeps you from seg faulting.....

 BJ> If you think you can work on the mex virutal machine, let me know.
 BJ> Wes was scratching his head on this one....  :(

 VC> Ok, mv  etc/m/*.mex etc/m/mex

I believe the *.mex files are the ASCII form of the files.  I forget what
the extension is for the compiled version of the mex files....

 VC> and rerun with the following:



 VC> + 18 Aug 16:14:39 MAX  Begin, v3.03 (task=14)
 VC> + 18 Aug 16:14:40 MAX  Vince Coen calling
 VC> # 18 Aug 16:14:43 MAX  Given 1441 min.
 VC> ! 18 Aug 16:14:49 MAX  MEX:  can't read file 'm/headfile'
 VC> : 18 Aug 16:15:15 MAX  File area SDS.MAXIMUS
 VC> : 18 Aug 16:15:59 MAX  File area SDS.FRONTDOOR
 VC> : 18 Aug 16:16:25 MAX  File area AALOCAL.GENERAL
 VC> : 18 Aug 16:20:48 MAX  Msg.area AALOCAL.LOCAL
 VC> # 18 Aug 16:22:40 MAX  External program: "/bin/sh"
 VC> : 18 Aug 16:22:45 MAX  Return from: "/bin/sh" (0)
 VC> + 18 Aug 16:22:55 MAX  Vince Coen off-line. Calls=8, Len=8, Today=8
 VC> : 18 Aug 16:22:55 MAX  End, v3.03 (5)

 VC> So I guess that bit works now to reload some .mex files.

 VC> Can I assume that if the mecca compiler does not report a fault when 
 VC> compiling that the compile is good or not?

I'm forgetting which program hanldes the mex files for compiling, but I
believe the program compiling the mex code on Linux / Unix based systems is
not generating error messages when run.  It is only when the mex virtual
machine (Mex VM) is started *and* loads a compiled mex program that the
Maximus BBS code is seg faulting.  And from memory, the seg fault is on the
start of execution of the mex code (i.e. on the 'main()' statement).  :(

There is probably a pointer error in loading and starting the mex code in the mex VM.

So, from your log file, I think you have things working like we have stuff
working.....  The mex VM machine needs someone to work on it.  Wes
(temporarily) switched off working on the mex VM machine to  work on serial
/ modem support to give him a break.  He was planning to go back to working
the Mex VM stuff once he got the serial / modem stuff working. 
Unfortunately, Wes is busy with family issues and with work issues right
now.  

Take care.....

Bob Jones, 1:343/41 (on 1:343/40)

--- Maximus/UNIX 3.03b