Path: utzoo!mnetor!uunet!swlabs!jack From: jack@swlabs.UUCP (Jack Bonn) Newsgroups: comp.dcom.modems Subject: Re: Quote from Feb 88 Computer Shopper Message-ID: <2093@swlabs.UUCP> Date: 5 Feb 88 22:49:22 GMT References: <2237@charon.unm.edu> Organization: Software Labs, Ltd. Easton CT USA Lines: 35 Summary: sarcasm was lost on me In article <2237@charon.unm.edu>, cs2531bn@charon.unm.edu (Ernie Longmire) writes: > I would bet he researches better than you do . . . if you'd read the article > more carefully, you would have realized that the author was making *fun* of > the so-called "experts" who used to insist that dial-up modems could never, > ever be designed to communicate faster than at 300 baud and that you'd need > leased lines to "maybe -- just maybe" run at 1200-2400 baud...and who now > insist that 9600 is the upper limit. Upon further examination, I too believe the "expert"s line was intended to be taken in a sarcastic tone. He just must have forgotten his ":-)". :-) On the other hand, I stand by my statement regarding the limited research shown by the article. It would have been appropriate to include a description of the Trailblazer in the article, if for no other reason than to disprove the "experts". Does anyone outside of the UNIX community even know about the Trailblazer? As an aside, what is the "real" theoretical maximum data rate that can be pushed through a data line? I guess I can set up a straw man. If one could control all the bits on a companded digitally sampled line (a big if), then 8 bit samples * 8K samples per second gives us a 64 kbps channel. They steal 8 kbps for signaling here in the US, so that leaves us 56 kbps. Although 14.4 kbps is using only less than 26% of this available bandwidth, how much more can we extract? At what point do you need to compensate for A-law/u-law companding? What if each end has different companding (for example on a transcontinental call)? What about phase jitter? At what point do these modems become "ISDN-killers"? (Sorry Rick.) -- Jack Bonn, <> Software Labs, Ltd, Box 451, Easton CT 06612 uunet!swlabs!jack