MacPPP 2.1SD Read Me
What is it?
Is a 115K or 230K DTE rate really going to be faster than
57.6K?
Caveats
What do I need to use it?
Does it work with a GeoPort Telecom Adapter?
How do I use it?
Copyrights & Credits
Version History
Known Problems
Comments or Questions?
This a modified version of Merit's MacPPP Link Access Protocol module for
MacTCP that supports DTE rates of 115Kbps and 230Kbps on 660AVs, 840AV and
PowerMacs. There are no other changes from Merit's 2.0.1 release but I
thought the support for higher rates justified a bump of the minor
version number. I also wanted some way to differentiate this version from
an official Merit release (this most definitely isn't) and my initials are
SD so that's why they're there. I'm releasing this since there are lots
of Internet providers adding V.34 dialups now that the spec has been
finalized and I'm sure there are other folks like me that want to get the
maximum possible speed out of our nice expensive new modems.
It depends on the data being transferred. The V.42bis data compression
standard is theoretically capable of 4:1 compression. Therefore you'd
want a 115Kbps DTE rate to fill a 28.8Kbps V.34 link w/V.42bis
compression. In the real world most data doesn't usually compress better
than 2:1 so a DTE rate of 57.6K is usually sufficient. On the other hand
I've seen better than 8000 bytes/sec using Fetch to ftp a highly
compressible text file (400K of repeating 'abcdef') over a 24K V.34 link.
Using the Electr onic Frontier Foundation's "Big Dummy's Guide to the
Internet" as a test file I've gotten in the neighborhood of 6200
bytes/sec. BTW, does anyone regularly get 28.8K connects? My normal
connect rate with a USR Courier V.34 on each end of the line is 24K with
occasional 26.4K and very rarely 28.8K.
If you have a Hayes Optima 288, it supports a DTE rate of 230Kbps to feed
their modified version of the V.42bis compression routine that's
theoretically capable of 8:1 compression (as opposed to normal V.42bis'
4:1 max). Unfortunately I don't have one so I can't say what sort of
throughput my test files would have at the 230K DTE rate.
The changes to support the higher DTE rates make use of an undocumented
and unsupported serial driver call as well as some direct manipulation of
the Serial Controller Chip. This has been working for a few months on my
660AV and PM6100 but there are no g uarantees.
You need a Mac that supports serial DMA (i.e. a 660AV, 840AV or any
PowerMac) and a true V.FC or V.34 modem attached to the built-in Modem or
Printer port. The increased interrupt load for the higher rates is too
much for non-serial DMA Macs so I'd stron gly recommend against using it
on one although I don't currently check the Mac type. If you insist on
trying it on a IIfx or Quadra 900/950 you must make sure that you have set
the serial ports to 'compatible' mode using the Serial Switch Control
Panel.
If you don't, you'll crash your Mac when you try to connect. Since I
already said you shouldn't be using this software on one of these machines
I don't consider this a bug so consider yourself warned.
Of course if your dialup PPP provider isn't using V.FC or V.34 modems with
a 115K or 230K DTE rate on their end you won't see any difference. That
part I can't help you with. :-)
No. The GeoPort Telecom Adapter pod isn't really a modem. It's just the
hardware required to connect your Mac to the phone lines. The rest of the
modem functions are implemented inside your Mac. Apple may implement a
V.34 version of the Express Modem software but I wouldn't hold my breath
waiting for it. I wouldn't recommend it on the PowerMac either as the
current V.32bis implementation of the software seems to be good for
approximately a 30% slowdown on my 6100AV.
If you aren't sure how to install MacPPP, see the accompanying 'Installing
MacPPP' document. Once you have installed MacPPP 2.1SD, open the Config
PPP Control Panel and hit the Config... button. Select 115K or 230K from
the Port Speed popup menu. That' s it. You should now be talking to your
modem at a rate that supports the maximum possible throughput. If you
selected a port other than the Modem or Printer ports you'll get an error
Alert when trying to connect stating that the higher rates aren't sup
ported for other ports and the rate will be changed to 57.6K.
MacPPP 2.1SD is based on MacPPP 2.0.1 which is Copyright ©1992-1993 Merit
Network, Inc. and The Regents of the University of Michigan.
The changes from MacPPP 2.0.1 to 2.1SD are Copyright ©1994 by Steve
Dagley. As is the case with Merit's statement regarding MacPPP 2.0.1, I
make no guarantee about the performance or reliability of this software.
Any risk in it's use is your own. Not f or profit distribution is
allowed/encouraged. This Read Me document must accompany any
distribution. If you want to include this version in some sort of
software collection (i.e. CD) all I want in exchange is a copy of the
collection.
Thanks to Bill Coleman of Hayes for reminding me to think async instead of
sync when programming the SCC for the higher rates. I've been doing code
for 13-bit sync serial data for so long I forgot what async was. :-)
Thanks to Larry J. Blunk and the rest of the MacPPP programmers for making
the source code to 2.0.1 available.
2.1SD - 12/11/94
First public release.
As always, revised this Read Me.
2.1SDb3 - 12/10/94
Third semi-public beta.
Due to popular demand (from Lyman Green) I have now updated the Config
PPP Control Panel to show the 115K and 230K rates in the Port Speed popup.
Revised this Read Me once again.
2.1SDb2 - 12/08/94
Second semi-public beta.
Oops - left in a modified version of the modem init code that occasionally
caused timeouts when trying to establish communications with the modem.
Cleaned up and expanded this Read Me.
2.1SDb1 - 12/06/94
First semi-public beta
Cleaned up the code somewhat and actually check for a suitable serial
driver before making the calls.
2.1SDa3 - July 1994
Never publicly released.
Made sure selected port was modem or printer.
2.1SDa1 - July 1994
Never publicly released.
First build.
Apparently the software for the Axion serial port switching device makes
the test for the internal serial ports fail and revert back to a DTE rate
of 57.6K. I don't have one of these critters so it's kinda hard to figure
out a workaround other than turning off their software.
You can e-mail me at sdagley@zeno.fit.edu
Enjoy!
Steve Dagley
December 11, 1994