[ipv6-wg at ripe.net] Re: [lir-wg] IXP networks routing

In message <20030304152015.B15927 at Space.Net>, Gert Doering writes:

>In the long run, one of the multihoming solutions might "really" solve 
>the problem, or we might end up needing "real IPv6 PI space" for the
>exchange points.  (And big enterprises.  And universities.  And so on.)
>
>But this "future plans" shouldn't stop people from deploying IPv6 for
>infrastructure things *today*.

These two sentences remind me of a sketch I saw many years ago on
british television:

The inventor for the rocket-car was interviewed.  He argued fewerishly
that all cars should be exchanged for rocket-cars right now because
it was much faster, cheaper, safer etc etc.

Only when questioned directly did he reluctantly admit that some
minor issues were not entirely solved yet, but it would of course
only be a matter of a short time and a little emperical data before
that would be sorted out.

Further questioning reveals that the "minor issues" are "how to
avoid collisions in intersections", "how to follow curves in the
road" and "how to stop if there is no brick wall".

It wasn't the pythons, but the style was similar.

IPv6 needs to fix the protocol errors in the political layer, soon.

Poul-Henning

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.