Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!husc6!hscfvax!pavlov
From: pavlov@hscfvax.UUCP
Newsgroups: comp.dcom.modems
Subject: Re: How to pick a medium speed WAN connection?
Message-ID: <361@hscfvax.UUCP>
Date: Fri, 27-Mar-87 17:37:53 EST
Article-I.D.: hscfvax.361
Posted: Fri Mar 27 17:37:53 1987
Date-Received: Sat, 28-Mar-87 18:17:30 EST
References: <2619@phri.UUCP>
Organization: Health Sciences Computing Facility, Harvard University
Lines: 37

In article <2619@phri.UUCP>, roy@phri.UUCP (Roy Smith) writes:
> 
> 	As a relative neophyte a datacomm, I recently had tossed in my lap
> the assignment of "getting our computers talking with their computers".  I
> think we've figured out that running IP over a point-to-point link is what
> we have to do.  What I don't know anything about is how to set up the
> physical link.  My "normal" alternatives so far seem to be:
> 	1) Leasing a D1 conditioned analog line ....
> 	2) Leasing a 9.6 kbps DDS line for $600 up front and $360/month.
> 	3) Leasing a 56 kbps DDS line for $600 up front and $700/month.
> 	4) Leasing a T1 "superpath" line...
> ...  Assuming we can get by with 9.6 kbps, what is
> there to choose between the first two alternatives?  Are we going to see
> any better performance with a DDS line instead of the D1 analog line...
>  What if we wanted to really cheap out and
> get a POTS line and something like a Microcom AX/9624c which claims to be
> able to do 19.2 kbps over an unconditioned dial-up line?  Is this a viable
> alternative?  Considering how much cheaper it is, there must be *some*
> reason why it won't work, but I don't see it.
> -- 
  We have been leasing a 9.6k DDS line for 3 years (formerly 2.4, then 4.8k).
  Distance: apx. 450 miles direct, tho actual path is apx. 650.  It has been
  extremely reliable.  Apx. 8 hours' down time per year, apx. 15 hours' "flake
  time" per year.

  I do not believe that the Microcom modem operates in true duplex mode; it, and
  other modems like it, either "simulate" duplex or allocate a very small band-
  with to one direction (dynamically, in response to traffic patterns).  If 
  this is suitable to your application, I would suggest looking at the Telebit
  modem.  From my point of view it utilizes a more robust transmission scheme
  and I believe it is likely to yield better throughput under various adverse
  conditions (since baud rate can be varied in much smaller increments/decrem-
  ents).  But I am not well-versed in this and I have no experience with any
  of these modems.  But I would certainly like to hear from those who have,
  since I am considering purchasing several myself.

      greg pavlov, fstrf, amherst, ny