In <176uZD2KcidF-pn2-Z6BmFfMjJ6VU@rikki.tavi.co.uk> "Bob Eager" <rde42@spamcop.net> writes:

RFC 1597 says:

The Internet Assigned Numbers Authority (IANA) has reserved the
following three blocks of the IP address space for private networks:

10.0.0.0 - 10.255.255.255
172.16.0.0 - 172.31.255.255
192.168.0.0 - 192.168.255.255

We will refer to the first block as "24-bit block", the second as
"20-bit block, and to the third as "16-bit" block. Note that the
first block is nothing but a single class A network number, while the
second block is a set of 16 contiguous class B network numbers, and
third block is a set of 255 contiguous class C network numbers.

So the range 172.16.0.0 - 172.31.255.255 is bigger than a class B (in CIDR
terms, it is a /12).

But I still do not see why 192.168.0.0 - 192.168.255.255 is to be regarded
as 255 contiguous class C network numbers rather than a single class B
(and the difference is material when it come to the meaning of
192.168.0.255 and of 192.168.1.0, as someone else has pointed out).

--
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl
Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5

In <1211617812.6424.0@proxy00.news.clara.net> The Natural Philosopher <a@b.c> writes:

It maybe that the netgear consists internally of a Switch connecting its
four local outputs together, plus a router connecting that whole bundle to
its ADSL line, but I think that is unlikely.

Because, with the Netgear set up with its default and recommended Netmask
155.255.255.0, traceroute showed that it thought 192.168.1.77 should be
routed up the ADSL line (though my ADSL provider was unable to send it any
further, of course).

--
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl
Email: chl@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5

On Mon, 26 May 2008 11:39:24 UTC, "Charles Lindsey"
<chl@clerew.man.ac.uk> wrote:

Take a look at RFC 1466.

--
Bob Eager
Use the BIG mirror service in the UK:
http://www.mirrorservice.org

Charles Lindsey wrote:
[snip]
FWIW, in terms of behaviour, all routers I know of appear to be
implemented exactly as you describe. In reality, there are varying
degrees of integration (with, as everywhere, a trend towards greater
integration in more recent products), but it's the behaviour which is
important.

Alex

On Thu, 22 May 2008 10:35:20 GMT, "Charles Lindsey"
<chl@clerew.man.ac.uk> wrote:

the original internet standards defined the "class" of an address by
the range in the 1st octet.

All of that supposedly has been dead for a long time (despite being
wirtten into lots of recent textbooks that should know better).

allowing "supernets" (old terminaology from when this started) is part
of meeting some later general RFCs and modern devices "should" do
that. However, old code never dies, it just gets patched........

I suspect the Netgear "as a router" supports 192.168./16, but the
supporting functions are struggling.

FWIW 1 obvious workaround if you dont want to readdress the
192.168.1.77 address is to move the router & DHCP ranges, together
with any fixed address devices.
i think your assumption here is that there is 1 cohearent consistent
debugged set of code in the router.

i suspect serious wishful thinking
The golden of bug avoidance is only change the defaults if you have
to.

The more you go away from the settings which get all the attention
during testing, the more likely you will stumble over stuff that never
got tested, never mind fixed.
--
Regards

stephen_hope@xyzworld.com - replace xyz with ntl

[snip]
It's a general discussion, not related to the needs of "IP address space for
private networks".

It implies that contiguous sets of Class C networks are now to be preferred
over single Class B networks, but only because of the lack of availability
of "real" class B networks. Ii also suggests that if there are few hosts
per subnet, the general wisdom is that small subnets should be used.

Thus the subnet size should be a matter of user choice, rather than forced
by convention. This is covered by "Classless Inter-Domain Routing"

The overriding conclusion is that the netmask - whatever its length - takes
precedence; so if your netmask means that A.B.0.255 and A.B.1.0 are within
the same subnet then they can be allocated to hosts.

Sadly, I suspect there are a lot of systems out there which use "class"
based defaults ...

--
Graham J

On Mon, 26 May 2008 21:01:14 UTC, "Graham J" <graham@nospam.zen.co.uk>
wrote:

I'm not saying it's good! Just saying where the 'to be regarded' came
from. Subnet classes are dead, anyway, and have been for a long while.

--
Bob Eager
Use the BIG mirror service in the UK:
http://www.mirrorservice.org