Strange indeed, Howard:

> Strangely, when I used LSPPP v .764 as the packet driver, consistently
> I got slower download speeds from my zoom 56K modem, compared to epppd,
> although I do not see how a packet driver can directly effect modem
> speeds.

I got the opposite impression with a 56K modem (not a Zoom one but it
seems Hayes compatible), and I think - didn't measure - consistently.
One explanation could be modem/hardwired compression - is it set the
same in both cases ? (Which is to be set with the modem initation, and
thus with the dialler - i do _not_ use LSppp's built-in dialler but an
external one [chat]; the LSppp dialler had some problems, earlier, so
I never vothered with this.)

Another thing, linked to that, is the port speed setting:
> /B   sets the baud rate.
> I hope/assume this is the speed of the modem, not the speed the com
> port talks to the modem - which is faster usually?

This /B switch is definitely for for the _port_ setting - the "real",
inter-modem speed is negotiated between the modems, and by them.  But if
your port setting is too slow, you get lots of errors and retry/resend
requests.  That could be it.

> /A   sets PAP timeout and retries
> Must you use this switch to enable PAP authentication?

I think this is only needed if you want to force PAP. As far as I
understood from the doc (which indeed is not too verbose, <g>), LSppp
does do some sort of "fallback". I never had to bother about that with
different access providers in different telco empires around here.

Thus the command line gets rather short, at my home place:
lsppp764 /b:03e8 /i:5 /B:115200 /U:xxx /P:yyy

The major difference I experienced was with the ability to negotiate
DNSs; I think this is part of "IP negotiation" - according to LSppp's
author, it does DHCP; but AFAIK the implementation of that one could
indeed vary considerably from server to server.  And as LSppp does _not_
read a WATTCP.CFG, so it may be better to put in DNSs at the command line;
it could shorten IP negotiation too, perhaps.

Thus away from "base" I susually have it like this (the /z is a debug
switch and creates a log file - not very useful if you don't know
those codes of just one or two bytes which are exchanged as key words
between hosts and clients):
lsppp764 /z /b:03e8 /i:5 /B:115200 /N:213.151.60.2,212.35.2.2 /U:xxx /P:yyy

I never bothered about most of the other switches; some of them seem
relevant only for specific machine conditions (/V, for instance:
needed only if the machine is networked and that card then has its
driver at that address commonly used precisely for packet drivers.)
I wonder if the /F switch does anything at all - maybe even _slows_
the machine if changed from the default (of "1" - a commport which
does have a FIFO buffer would get polled slower; a port setting of
115200 is way fast enough, in comparison with any real line speed of
incoming bytes, even on stone-old '286s. "No buffer" means the same as
1 byte in a FIFO buffer and correspondingly, an irq request to get it;
I doubt if even with 56K modems there would be overruns.)

But the doc should be a little bit more verbose or rather, precise
indeed (there are trivial syntax errors in the decription of the
switches).

// Heimo Claasen // <hammer at revobild dot net> // Brussels 2002-04-20
The WebPlace of ReRead - and much to read  ==>  http://www.revobild.net

To unsubscribe from SURVPC send a message to [EMAIL PROTECTED] with 
unsubscribe SURVPC in the body of the message.
Also, trim this footer from any quoted replies.
More info can be found at;
http://www.softcon.com/archives/SURVPC.html

Reply via email to