Reverse or inverse ARP from windows/linux - no way (!?!?)

=46irst off the TCP/IP stack is 4 levels, and the ISO model is 7 layers. Thus there is intrinsic mismatch. Much of the TCP/IP model tools span=20 two or more layers in the ISO model. The real deal for TCP/IP protocol=20 definitions comes from the IETF RFC library.

Please note that this RFC information is freely available for any use.

Wikipedia tries to allocate levels but fails sometimes:

And another opinion:

Personally i place IP and UDP at the ISO link layer and ISO the transport= =20 layer. After all they are sending the message. (R)ARP is about = different=20 network functions and talks to the same layer (going towards Phy) as TCP=20 and IP; however it functions in link, transport, and network ISO layers. = =20 So where do you want to place it?

No. UDP does not use IP all. It operates in parallel with IP, and = provides=20 a different service. Al least that is what the RFCs say.

rovides

UDP (and TCP) both use IP as their lower layer. UDP operate in parallel with *TCP* and provides a different service. They *both* run over IP.

It is most emphatically *not* what the RFCs say. In fact, the opening paragraph of RFC786 (which defines UDP) is:

"This User Datagram Protocol (UDP) is defined to make available a datagram mode of packet-switched computer communication in the environment of an interconnected set of computer networks. This protocol assumes that the Internet Protocol (IP) [1] is used as the underlying protocol."

And the prior poster is correct - it's IP that basically defines the Internet - the whole Internet exists to move IP packets around, and protocols like ICMP, UDP and TCP are built on top of IP. A number of other less common protocols also are built directly on top of IP (rather than on top of UDP or TCP). For example, the communication between routers running OSPF happens over IP, but uses protocol #89, which is distinct from TCP (6), UDP (17) and ICMP (1).

As for DHCP, it does, in fact, use a UDP packet, although in many implementations that's more a formality and the IP header is largely faked during initial system startup. Having it an actual UDP packet makes it rather easier for the (DHCP) servers and the rest of the network, though.

You are mistaken on this point.

UDP is built on top of IP. TCP is built on top of IP. ICMP is built on top of IP.

These three protocols are peers, at the same level in the network stack.

rovides

You are aware that UDP is IP encapsulated, just like tcp? If tcp is one level above IP then UDP also is. But like I said, I am not particularly interested in that sort of "official" side of things. dhcp is indeed running over UDP, but during the gray period of network setup; then at least the request is broadcast etc., some level cheating is applied during the process.

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

------------------------------------------------------

In comp.protocols.tcp-ip Didi wrote: (snip)

There is internet, and there is the Internet. You can have a network, even an internet (network of networks) not connected to the Internet (the rest of the world).

Any of them may or may not have a DHCP server on them.

-- glen

note quite true anymore ... the copyright notice use to say that author gives unlimited rights including any derivative works as long as the IETF copyright notice was included. A couple years ago, that was changed ... and now it ways that RFC authors may retain rights (so RFCs published since the change are subject to the new copyright rules).

look at RFCs related to IETF Trust ... for instance 5377

current documents carry the following

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents

... i got hit with it and had to hire a copyright lawyer out of my own pocket (situation where an attempt was made to try and apply the new rules to RFCs published under the old rules) ... and suggested that they needed to make it much more clear to the authors about copyright changes.

The ISO and the IETF are both wrong. There are nine layers :)

rick jones

In comp.protocols.tcp-ip Rick Jones wrote: (snip)

The story I still remember is that between Europe and US one wanted six and one eight (I don't remember which) so they compromised on seven. Also, for ATM, one wanted 32 byte payload, one 64, so they compromised on 48. (I would have thought it would be geometric mean, though.)

-- glen

That's only true for a meaning of the term "internet" that IMHO rather strictly contradicts with your self-proclaimed expertise in internet technologies. What exactly did the 'I' in IP mean again?

One on which your personal opinion is not the normative reference for the meaning of a technical term like "internet".

... says the person who obviously doesn't understand USENET well enough to recognize that his cross-post is obviously off-topic in one of the three groups he wrote to and on the fence of topicality for a second, who didn't bother setting a Followup-To on top of that, and replies to technical suggestions with ad-hominem attacks undiluted by technical arguments. Cute.

Interesting how you believe you know what I looked at and what I didn't.

Who do you believe you're kidding?

Yeah, right. Writing the above really saved your time, as compared to, say, not bothering to write it at all. Sure.

But if you're so sure I don't understand the issue, how about you start arguing against my statements instead of against my person?

Many networks that have nothing whatsoever to do with the Internet have adopted its protocol. You can't have the Internet without IP, but you can certainly have IP with the Internet.

DS

For this thread to make any sense you should be able to define exactly what an 'Internet' is. Off you go...

Mark.

I skipped over that part. When he mentioned his 'customers', I was thinking of people outside of his LAN (on the Internet), which would assume the operation of the entire name service infrastructure.

On LANs, the easiest protocol to implement (for discovery of peers) would be NMBD. That's a part of the Samba s/w suite

=20

protocol=20

Thanks for the update, i hadn't realized that the publishers sickness had= =20 gotten to IETF. Damn publishers, insisting on hanging on to a=20 technologically obsolete business model from the dark ages.

Not at all, both of those are applied liberally to all levels of=20 any stack currently; they actually enclose the whole stack. =20 Worse they are trying to get in between the existing layers/levels.

DHCP

This erroneous statement alone shows that you are ill prepared to = participate=20 in this discussion. If you think UDP "sits" on top of IP you are very = mistaken. It is a parallel (at the same level) link/transport protocol.

Moreover, you desired email reply shows arrogant disregard for USENET = norms.

no

DHCP

for

and provides

Damn, it seems i got things upside down again.

Op Thu, 01 Apr 2010 06:09:46 +0200 schreef JosephKK :

If you think that the link layer, internet layer and transport layer are all the same layer, then yes. But alas for you, OSI and TCP/IP do not agree with you.

Well, at least he didn't top-post.

Em 1/4/2010 02:09, JosephKK escreveu:

mistaken.

If you think you have a lesson to teach please:

Can you show the erroneous part of the statement explaining us how an UDP datagram could be *sent* to a network without the IP layer?

Remember *you* wrote they're 'parallel' so they can be used in exchange of each other, right?

Your not very informed reply shows an arrogrant ignorance for the Internet RFC or equivalent ISO models.