Problems with SimulTask 386 and MS-Kermit 3.1

Joe R. Doupnik jrd at cc.usu.edu
Wed Jun 26 08:59:26 AEST 1991


In article <13374 at uwm.edu>, dbm at csd4.csd.uwm.edu (Dennis B Meilicke) writes:
> In article <1991Jun23.153720.1674 at unhd.unh.edu> pss1 at kepler.unh.edu (Paul S Secinaro) writes:
>>I'm using AT&T SimulTask 386 version 2.0 (which is actually VP/ix 1.something)
>>on an AT&T SVR3.2.1 system.  I'm having a problem with MS-Kermit 3.1
>>
>>When I run Kermit, with Simultask set up to allow access to the COM2 port, I
>>get the error message:
>>
>>?Warning: no harware for this serial port
>>?This port will be operated through the Bios as BIOS2
> 
> Well, I'm not sure, but...
> 
> I have a similar problem running MS-Kermit 3.1 under MS-DOS.  My modem
> is on COM4.  When I tell Kermit to use COM4, it tells me there is no
> hardware for this port, and uses BIOS4.  When I connect, I get a
> screen full of characters.
> 
> I'm not sure, but I think Kermit's port definitions are screwed up.
> After a lot of trial and error, I found that I can get Kermit to work
> by telling it to use COM2.  *All* of my other programs use COM4.  Very
> strange.  Anyway, you might try using COM4.
> 
> Good luck...
> 
> Dennis
-------------------------------
	Aw, come on fellas. By now everyone knows that there is no definition
for COM3 and COM4. MS-DOS Kermit has the good sense to look before leaping,
more than can be said for some comms programs. If you want COM4 with Kermit
then you can do one of two things.
	With MSK v3.10 you can give Kermit commands
	SET COM4 \x0nnn irq		where nnn is the hex port, irq is irq
or	SET COM3  ditto
Or you can put the port address of COMx in the segment 40h work area where
everything rational looks for it. The four words (two bytes) starting at
40:0h are the port addresses for COM1..4. Kermit looks there, the Bios looks
there. The POST part of the Bios is responsible for filling in the numbers
but IBM (hence everyone else) looks for only the first two ports; you can
add to the system's knowledge.
	Joe D.



More information about the Comp.unix.sysv386 mailing list