Newsgroups: comp.sys.transputer
From: s9010721@minyos.xx.rmit.EDU.AU (Aaron Brian McDonough)
Subject: Connecting 2 Transputer Boards (in diff. hosts)
Organization: Royal Melbourne Institute of Technology, Melbourne, Australia.
Date: 28 Dec 1994 09:33:30 GMT
Message-ID: <3drbda$q7n@aggedor.rmit.EDU.AU>

Hi all,
	I have two transputer boards in separate host PC's. One 'puter is a 
	Sprint board with one processor and 4 meg of RAM. The other board 
	is a Quintek Fast-9 ( 9 processors each with 4 meg ), set up as 
	pipeline . The compiler I am using is 3L's C compiler (v2.2.2, 
	I think). 

	The problem: My goal is to use one host to monitor suspect tasks 
	that are running on slave processors (by printing variables etc..to 
	the screen.)  To do this I am inserting the Sprint 'puter into the 
	chain of Fast9 slave 'puters. One link (at the end of the pipeline) 
	is taken from the back edge of the Fast9, over to the Sprint board 
	(in the other PC), and then back again to the Fast9 board (completing
 	the pipeline). 

	The Fast9 board is limited to one 'puter (Master) being connected to 
	host. By inserting the Sprint board in the pipeline, I am hoping that 
	the Sprint board will also be able to talk to it's own host at the 
	same time the Fast9 Master is talking to it's host. 

	Upon making connections from the back of each board, I ran the 
	Netmap program supplied with the Fast9 board. All of the 
	connections between the board's were recognised execpt for the 
	PC-to-Master processor on the *remote* PC. I ran the Netmap 	
	program on both PC's getting the same result, all connections are 
	recognised execpt for the remote PC-to-Master connection. I tried 
	to use the configurer supplied with the compiler, but the same 
	error occured. It could not recognise the other PC-to-Master 
	connection either.
	
	Are there any other software/hardware considerations I have missed?

	Note: I also have up-s on one board to the down-s on the other board.

	I know that this setup is possible (mentioned in the 3L manual), 
	so any help solving my problem is *greatly* appreciated.


-------------------------------------------------------------------------
Aaron McDonough		E-mail: s9010721@minyos.xx.rmit.edu.au
Applied Physics			aaron@arcadia.cs.rmit.edu.au
RMIT				aaron@bunyip.ph.rmit.edu.au
-------------------------------------------------------------------------



