🚀 go-pugleaf

In article <323449C2.31DFF4F5@hobbes.ss.org>,
David Collier-Brown  <davecb@hobbes.ss.org> wrote:
>Richard Sexton wrote:
>[in a discussion with Brian Tao <taob@zap.io.org>]
>> The EDNS hack, searcing the ALternic root servers if the
>> current root serves don't resolve a name seems sound in theory
>> and has worked well in operastion. I don't see the harm in it.
>
>  Well, I like the experiment (although Karl seem convinced
>I don't), but do have a scalability concern.
>
>  If a second company starts up in competition to Alternic,
>and starts providing root service, either as a primary source of data 
>or as a fallback after the normal ones. then they can sell another
>copy of cybercafe.biz to someone else entirely...
>  and the end user customer will get 
>the one in Toronto if their ISP subscribes to Alternic, the one in
>Podunk if their ISP subscribes to  NotherNic, and one in Tnuctipun
>if their ISP subscribes to MarsNic.


Obvioulsy a real concern.  The Alternic have been very careful
not to tep on anybody's toes. Suggestions that they resolve
names that have been cut off for lack of payment were met
by "That would be politically unexpedient".

The InterNIC can screw up bigtime (and do) and no harm will
come to them. Not so the Alternic. As a purely commercial
concern they have to A) Try harder (being #2 and all)
and B) Tread lightly. 

Their reord so far indicated they are doing just that.

Alternic has not caused any namespace collisions, and I would
put forth the notion that the party that does, is liable,
whether it be ALternic, IANA or MarsNIC.

-- 
Richard Sexton
root@cabal.org