Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!uwm.edu!zaphod.mps.ohio-state.edu!usc!elroy.jpl.nasa.gov!turnkey!jackv
From: jackv@turnkey.gryphon.COM (Jack F. Vogel)
Newsgroups: comp.unix.i386
Subject: Re: 386/ix uucp throughput still stinks
Message-ID: <6717@turnkey.gryphon.COM>
Date: 16 Feb 90 14:47:55 GMT
References: <1990Feb15.070638.1086@esegue.segue.boston.ma.us>
Reply-To: jackv@turnkey.gryphon.COM
Distribution: na
Organization: Turnkey Computer Consultants, Westchester, CA
Lines: 32

In article <1990Feb15.070638.1086@esegue.segue.boston.ma.us> johnl@esegue.segue.boston.ma.us (John R. Levine) writes:

>.... I got a 16550A, unsoldered the 8250 from my internal Telebit
>modem, soldered in a socket, and stuck in the 16550A.  I can send stuff to
>other sites at 1200-1400 cps, which is what I hoped for.  Inbound news is
>still a lousy 500 cps.  What gives?  The modem is set up as ttyd1, with tty00
>being a mouse that I don't use much.  That wouldn't be a problem, would it?

John,	I would check with the site that is sending you the news and see what
kind of figures they show. I suspect what you will find is that their figures
show the 1200-1400 cps even though yours look slower. I found this when I
run the TB+ on the standard COM port, it must be something to do with the
buffering between the modem, the port and memory. However, with the modem
on my Bell ICC (which has 256K onboard) the figures are nearly identical
in and out (i.e. 1200-1600). The only reason I don't always run on the ICC
is that the crazy thing has a tendency for its onboard software to crash
or something, so when I leave town for a week I move it.

This brings up a question, are others out there using the ICC seeing this
sort of behavior? Basically the symptom is that DTR will go down, you could
still open the port, but things are just dead. You can simply use the supplied
idebug program and reset the card and then things are fine, but needless to
say this doesn't make for very good unattended operation :-}. This happens
intermittently, but quite frequently. I have never been able to connect a
crash with any particular event.

Disclaimer: IMHO only.

-- 
Jack F. Vogel			jackv@seas.ucla.edu
AIX Technical Support	              - or -
Locus Computing Corp.		jackv@ifs.umich.edu