Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!ucbvax!POSTGRES.BERKELEY.EDU!dillon
From: dillon@POSTGRES.BERKELEY.EDU (Matt Dillon)
Newsgroups: comp.sys.amiga.tech
Subject: Re: Parallel-Port network preliminary specs & demonstration code
Message-ID: <8909112204.AA05485@postgres.Berkeley.EDU>
Date: 11 Sep 89 22:04:48 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Lines: 31

:When we used the parallel port as an IO device for the Lorraine (the Amiga
:prototype), we connected it's strobe & acknowledge to acknowledge & strobe
:of our Sage code development machine.  Of course, we had fewer machines
:than Matt in those days, so we were not trying to network it :-).
:
:Did you consider tying the flag input to D5?  If the network is only
:sporatically busy, this would allow an interrupt wake up mechanism
:for prospective clients instead of having them poll it.  That would
:contribute to the loading problem, though, for which I've no spec to
:refer to.
:
:- Kodiak

	Ah!  That's a great idea!  Why the !@#$ didn't I think of it! Thanks!

	My first posting will just be the ND server (two machine network).
	The server will poll continuously at priority -1.  I'll post it to
	comp.sources.amiga.

	My second posting will be a new server and nddisk.device that use
	a new device 'parnet.device' which I am working on right now.  This
	device will properly implement the network (DGRAM level).

	I think I'll use your idea 'bout the flag input ... That's a cool
	idea!  D5 is used only for line aquisition and would provide a 
	great wakeup mechanism!  And it's negative edge, perfect!

	This would mean that the network would load only the machine a
	particular packet is destined for.

						-Matt