Re: [MiNT] bigger pppd problem

[Jim DeClercq <jimd@panix.com>]

I am replying to myself, so I already know what this says.

The solution is to change ISPs, because mindspring was a bit
mindless, and has been for a while. First, they had a mailserver
that did not get stuck on badly formed (meaning no line space
between header and body) messages, and replaced that with the
latest and greatest, which did not handle those correctly, and
choked up. They had a web site reachable by lynx so one could
delete the problem message, and replaced that with something
newer and better not reachable by lynx.

My response to this was to move my wanted mail elsewhere, and let
my mailspace there fill up with unwanted commercial messages.

Then they installed the latest and greatest version of ppp, which
did not work with any earlier version of ppp. The RFC for ppp is,
I think, 1661, and it calls for a start of header of binary "21",
for the LCP, and as a practical matter, that was followed with
the text "LCP"  for Link Control Protocol.

This latest follows the "21" with "IP". That may be about the
same thing for a human reader, but it makes a difference to
TCP/IP stacks that look at both the header top and the text that
follows. Every earlier ppp version looks at those two things,
concludes that procedure is unknown, and link negotiation fails.

I notified them by letter of this, they took no action, and now
my dialup is as jimd1@gbronline.com. Great Barrier Reef has no
connection with Australia. They are in Joplin, Missouri, USA.

Their phone people have been assured that they are using all the
latest programs, but they are not, and thus MiNT ppp works with
that ISP. They are cheaper, too.

-- 
 /"\      Jim DeClercq--jimd@panix.com--Sylvania, Ohio, USA
 \ /      ASCII ribbon campaign | I'm a .signature virus!       |
  X       against HTML mail     | Copy me into your ~/.signature|
 / \      and postings          | to help me spread!            |
.


On Thu, 19 Jun 2003, Jim DeClercq wrote:

> Hello, all
>
> With a great deal of help and support from Martin Tarenskeen, and
> a wild guess on my part that "passive" was a necessary thing, I
> got ppp to dial and connect, and to tell me that what works
> with STiNG for DNS numbers is not what the current DNS is, and
> there things get seriously stuck.
>
> All went well until
>
> Unsupported protocol 'IP' (0x21) received
> sent LCP [ProtRej
>
> and then my ISP shut down IPCP, and then LCP, with appropriate
> replies from EasyMiNT pppd, and quit.
>
> I tried this twice, with exactly the same results. ppp on the
> far end has a packet not recognized by SpareMiNT ppp.
>
> And that is what I expected to run into in the first place.
>
> OK, does anyone have an idea for a next step?
>
> My provider is not run by technically competent people, but I do
> not know of a good one in this area. Still, I will take
> suggestions.
>
> --
>  /"\      Jim DeClercq--jimd@panix.com--Sylvania, Ohio, USA
>  \ /      ASCII ribbon campaign | I'm a .signature virus!       |
>   X       against HTML mail     | Copy me into your ~/.signature|
>  / \      and postings          | to help me spread!            |
> .
>
>
>