[132347] in North American Network Operators' Group
RE: Introducing draft-denog-v6ops-addresspartnaming
daemon@ATHENA.MIT.EDU (George Bonser)
Sun Nov 21 18:14:19 2010
Date: Sun, 21 Nov 2010 15:14:14 -0800
In-Reply-To: <AANLkTinb4i2rkR9=sFoJWsnXE57SC3BjLOtOvKG+bN24@mail.gmail.com>
From: "George Bonser" <gbonser@seven.com>
To: "William Herrin" <bill@herrin.us>
Cc: nanog@nanog.org
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
>=20
> An option w/ movable separators:
>=20
> 260:abc:1234:9876:fe::1
>=20
> Actual IPv6 standard (and also allowed w/ movable separators):
>=20
> 260a:bc12:3498:76fe::1
>=20
The problem with movable separators is in handling zeros. If the
separators are a known distance apart, zeros can be deduced. The
example above has only one zero. Imagine it were a different address:
260a:0:8:65::1
Now with movable separators the the zeros would be explicit because you
don't know how far apart the colons are n the address. In your example,
it would become:
260:a00:0000:0800:65::1 because there is no way to tell from a movable
colon address how many places the colon was moved and how many zeros
there are between them. You can move colons as long as zeros are
explicit but you must have fixed colons if zeros are implicit.
Otherwise there is no way to deduce where the zeros go from simply
parsing the address.
