Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!tut.cis.ohio-state.edu!snorkelwacker!bloom-beacon!eru!luth!sunic!nuug!sigyn.idt.unit.no!lear!hta
From: harald.alvestrand@elab-runit.sintef.no
Newsgroups: comp.protocols.iso.dev-environ
Subject: Re:  ISODE and Apollo X.25
Message-ID: <1990Jul5.194228.24300@idt.unit.no>
Date: 5 Jul 90 19:42:28 GMT
References: <9006270711.AA26586@nuada.sics.se>
Sender: news@idt.unit.no (Usenet news admin)
Reply-To: harald.alvestrand@elab-runit.sintef.no
Distribution: inet
Organization: ELAB-RUNIT, SINTEF, Norway
Lines: 17

Well....I did make a gateway-based approach....
the 3 x.25 layers present in ISODE 5.0 were all using some kind of sockets,
all with their own bugs and weird ioctls.
the X.25 I worked with was Symicron's, which shows strong traces of having
come from the PC world, where one connection is a lot....
Anyway, what I did was to make a new x.25 layer based on an existing one
(as recommended), and to get one single application, the tp0bridge, working
with this x.25 layer.
The hacks involved were nauseating (how about one data-shuffling UNIX process
for EACH DIRECTION ON EACH CONNECTION!), but it worked, and with acceptable
performance for my application.
So, sometimes a bridge works....as a side effect, I got the possibility of
using the X.25 card in a single PC for transport connections from many.
Nice try!

                    - Harald A
                    patcher of what I have to