lkml.org 
[lkml]   [2000]   [Jan]   [29]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: all zeroes/all ones used in host IP's...
Date
From
From: "H. Peter Anvin" <hpa@transmeta.com>
Subject: Re: all zeroes/all ones used in host IP's...
Date: Fri, 28 Jan 2000 16:04:41 -0800

> Magnus Danielson wrote:
> >
> > From: hpa@transmeta.com (H. Peter Anvin)
> > Subject: Re: all zeroes/all ones used in host IP's...
> > Date: 28 Jan 2000 11:55:20 -0800
> >
> > > You are, indeed, completely wrong. There rules are that neither the
> > > *entire* network portion or the *entire* host portion of the address
> > > can be all 0 or all 1.
> >
> > No, he is correct if you go by RFC 1122. However, if you are running CIDR you
> > are right. It seems most posts on this topic actually misses the pre/post CIDR
> > conditions and you really have to read the full section of RFC 1122 and think
> > of what class the address at hand has in order to fully interprent the
> > paragraph that he quoted correctly.
> >
>
> No, he's not; he claimed there was something magic about each *octet*.
> That has never been true, although people have, in the past, been keen
> on putting the barriers at octet boundaries.

This is true, he is off on the octet issue. But since his subnet was on octet
boundary so it got things confused. The rules he refered to is valid however
for the examples he gave _GIVEN_ that you do not run CIDR in your network
(that is, on your routers).

> However, RFC 1122 is obsolete. Everyone uses CIDR these days, and once
> you have that, you have the rules I gave. CIDR is Best Current Practice
> per RFC 2050.

RFC 1122 may feel obsolete, but it is still labeled standard.
RFC 1518 is labeled proposed standard
RFC 2050 is BCP, but covers INTERNET REGISTRY IP ALLOCATION GUIDELINES and only
suggest CIDR, not specify it as standard. It does not replace
RFC 1518.

BCP was invented to supply a way of endorsed practice while keeping it out of
the standard process. The CIDR spec is in for the standard track and has yeat
to become standard, until that is the case RFC 1122 does apply and actually is
standard. So, RFC 2050 provides the BCP for IP Allocation guidelines and is
using the framework set up by the CIDR RFC's but the CIDR RFC's does not yeat
overshadows the statements in RFC 1122 if you read things strictly.

You are however correct in that everyone is using CIDR and everyone should be
using CIDR. So, for the applicability to the Linux kernel you are infact
correct.

Cheers,
Magnus

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:56    [W:0.108 / U:0.096 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site