Path: utzoo!utgpu!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!ames!pasteur!agate!ucbvax!vaxb.roe.ac.uk!MJP
From: MJP@vaxb.roe.ac.uk
Newsgroups: comp.sys.transputer
Subject: Long link cables
Message-ID: <20200BD2_00064104.0091BE7D3D8D3960$17_1@UK.AC.ROE.VAXB>
Date: 16 Nov 88 01:43:47 GMT
Sender: usenet@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 13
X-Unparsable-Date: 15-NOV-1988 16:34:37 GMT 00:00

Does anyone have figures on the performance of links implemented over long 
twisted pair link cables (>15m) or fibre-optic links?  There is presumably
a point at which the propagation delay down the cable delays the 
despatching of the acknowledge bits sufficiently that the data transfer 
rate starts to drop.  I've had a look in Rygol & Watson "Connecting Inmos 
Links", but while there are design notes and guidelines, there's nothing 
about hard, bitter experience!  

All and any info gratefully appreciated,

Cheers,
Magnus Paterson,
Royal Observatory, Edinburgh