Path: utzoo!attcan!uunet!cs.utexas.edu!usc!samsung!umich!terminator!dabo.ifs.umich.edu!rees
From: rees@dabo.ifs.umich.edu (Jim Rees)
Newsgroups: comp.sys.apollo
Subject: Re: Socket programming under SysV ?
Message-ID: <1990Jul9.151235.1181@terminator.cc.umich.edu>
Date: 9 Jul 90 15:12:35 GMT
References: <1990Jun21.061232.2256@santra.uucp>
Sender: usenet@terminator.cc.umich.edu (usenet news)
Reply-To: rees@citi.umich.edu (Jim Rees)
Organization: University of Michigan IFS Project
Lines: 22

In article <1990Jun21.061232.2256@santra.uucp>, s31786w@kaira.hut.fi
(Matti Tapani Koivisto) writes:
    I'll have to do some "socket programming" or interprocess communication
    over a TCP/IP LAN. There is no problem as long as I can use BSD system
    calls, but the big boys have decided that we will write only SysV software
    (no BSD system calls allowed).

I've had the extreme displeasure of trying to use several "real" sysV
machines, and like Apollo, they all use sockets, not TLI, to access TCP/IP.

Apollo comes with streams, but no protocol stacks.  You are of course free
to write your own TCP/IP stack, or buy someone else's in source form and try
to port it.  The port would be tough because the commercial TCP/IP stacks
usually have kernel dependencies beyond streams.

You can tell the "big boys" that it will cost them US $400,000 to develop or
import a streams TCP/IP stack and figure out how to use TLI, or you can use
those nasty BSD socket calls for free (cost estimates are my own and do not
necessarily reflect reality).

By the way, the socket calls all work fine in the Apollo sysV universe.
You'll have to import the include files, of course.