home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Columbia Kermit
/
kermit.zip
/
newsgroups
/
misc.19970929-19971216
/
000119_news@newsmaster….columbia.edu _Thu Oct 16 14:52:47 1997.msg
< prev
next >
Wrap
Internet Message Format
|
2020-01-01
|
4KB
Return-Path: <news@newsmaster.cc.columbia.edu>
Received: from newsmaster.cc.columbia.edu (newsmaster.cc.columbia.edu [128.59.35.30])
by watsun.cc.columbia.edu (8.8.5/8.8.5) with ESMTP id OAA28101
for <kermit.misc@watsun.cc.columbia.edu>; Thu, 16 Oct 1997 14:52:46 -0400 (EDT)
Received: (from news@localhost)
by newsmaster.cc.columbia.edu (8.8.5/8.8.5) id OAA00365
for kermit.misc@watsun; Thu, 16 Oct 1997 14:52:45 -0400 (EDT)
Path: news.columbia.edu!panix!news.eecs.umich.edu!nntprelay.mathworks.com!newsgate.duke.edu!news.eng.convex.com!cs.utexas.edu!news.cs.utah.edu!cc.usu.edu!jrd
From: jrd@cc.usu.edu (Joe Doupnik)
Newsgroups: comp.protocols.kermit.misc
Subject: Re: Kermit via PPP under DOS?
Message-ID: <$QmoOmiES1Xq@cc.usu.edu>
Date: 16 Oct 97 09:59:26 MDT
References: <k1c7kBQEU5Wv@cc.usu.edu> <omn2kcqegc.fsf@tees.cs.ualberta.ca>
Organization: Utah State University
Lines: 58
Xref: news.columbia.edu comp.protocols.kermit.misc:7898
In article <omn2kcqegc.fsf@tees.cs.ualberta.ca>, Vladimir Alexiev <vladimir@cs.ualberta.ca> writes:
> In article <5jzp7SiZjuUm@cc.usu.edu> jrd@cc.usu.edu (Joe Doupnik) writes:
>
>> >> >> > Is BOOTP impossible with a SLIP interface? How about DHCP?
>> BOOTP and DHCP require a local MAC address to work with, SLIP has no MAC
>> address
> Ok, now: Is BOOTP availability over PPP links sufficient incentive to warrant
> pursuing this issue?
>
>> continue the conversation with the authors of those drivers and see if they
>> will reveal the extent to which they perform the required emulation.
> Ok, I'll try that.
>
>> Proxy ARP et al have not the slightest to do with Kermit (or other
>> client) internals. That is outside of and unknowable to these clients.
> I know that. I was trying to figure out how does class 1 emulation work.
>
> For a PPP link, Kermit`s job is pretty simple: it has to send all it wants to
> send down the link, without caring mcuh about MAC addresses. That's why I
> think it would be an easy fix: Kermit could simply disregard some of its
> "doubts" about the faithfulness of the class 1 emulation and just go ahead.
You are still grasping the problem by the wrong end. "Just disregard"
may seem to be convenient for your situation but is plain wrong in general.
The problem is very likely improper emulation of Ethernet by the PPP drivers,
as I have speculated each time this thread has a new message.
>> MSK will report
>> 1. if it is unable to register with the Packet Driver for IP and ARP packets
>> 2. if another station responds to an ARP for MSK's own IP address,
>> 3. and if it is unable to receive a response to an ARP.
> This is the kind of info I need to figure it out (if at all possible from
> external observation only). Does the message "Unable to ARP resolve" mean item
> 3 above? Will these conditions appear in the order listed, so can I assume
> that 1 and 2 do not happen?
Unable to ARP resolve means just what it says: ARP request sent, no
matching ARP reply received. Malformed ARP replies are the same as no reply.
>> If those conditions are met and the driver proclaims to be Ethernet then the
>> driver had better behave like Ethernet... Kermit does tell you if
>> interfacing conditions fail
> What other errors can happen from that point on? (Ok, perhaps this is a stupid
> question :-) Can you think of other interfacing conditions that can fail,
> apart from the listed 3?
I say again, the emulation is likely broken. Understanding the overall
situation does require some knowledge of TCP/IP on Ethernet, and this is not
the place to provide a course on that. The authors of those PPP drivers are
the ones to take on the task, as has been nicely demonstrated today.
Joe D.
>> it cannot tell anyone about Ethernet simulation failures in an external
>> driver.
> I was hoping that we can figure out what more does Kermit expect from a class
> 1 emulator compared to other TCP apps (and you're starting this, with the 3
> conditions above). Now I'll try from the other end, ask emulator authors what
> less do these emulators provide compared to true ethernet drivers.