Re: [IDNO-DISCUSS] Working Group C

Mark C. Langston (skritch@home.com)
Mon, 12 Jul 1999 09:49:09 -0700


On 11 July 1999, Karl Auerbach <karl@cavebear.com> wrote:

>
>
>> Karl, you make a good argument here, but you point up my one major conern
>> at the end of your description. Or at least I think you do. Comments
>> below:
>
>I'll see if I can add some more light to your concerns.

[...snip]

>
>It is cheap (relatively) and easy to start a system of root servers. (The
>hard part is convincing people to point to your root server group.)
>

Agreed. And it's the art of "convincing people" that plays to my
stated fear. While the people who actually maintain the DNS tables
for a company or organization may have one set of criteria (most
likely based on technical merit) for choosing one root over another,
the management may have another. And my fear is that the "convincing"
will be directed at the management. This has the potential of leaving
us with a net that in large part is running on roots that may have
technical issues.

>So, yes, there is a risk of "the big guys" establishing their own root
>systems -- in fact, I have heard that they have done so already in order
>to firewall themselves away from NSI-induced errors but are merely
>slavishly mirroring the "legacy root" system.
>
>So, I'm trying to recognize a fact-of-life that we can't prevent anyway.
>By having such a recognition, we make it easier for the smaller guys to go
>into the business and offer an alternative without having to endure the
>calumny of being told that they are risking and damaging internet
>connectivity.
>

But, if I understand what you proposed (and there's no guarantee that
I do, so please feel free to correct me), the startup root may or
may not carry one or more TLDs. Based on the decisions made as I
described above, this could effectively eliminate certain TLDs simply
because large companies buy into one vs. another root.

>It's important to recognize that I'm only talking about multiple root
>systems, not split TLDs. I think that it would be bad if
>"freds-barber-shop.tld-x" gives inconsistent results depending on which
>root system you enter through.

Ok, this is what leads me to believe I may have misunderstood you.
Could you clarify whether or not these independent roots would or would
not be required to carry all TLDs?

>
>(It's the same thing -- if I look up "freds barber shop" in my telephone
>directory, it had better give me a useable number to Fred's no matter
>whether I use the AT&t phone book, my local telco's, a CD-ROM, or
>switchboard.com's directory service. Whoever gives me "wrong" answers
>will be crossed my list of acceptable directory service offerings.)
>

But if your choice is a phone book that gives you more correct answers
than incorrect answers, but doesn't include Fred's barber Shop, or even
a category for barber Shops, you shut out and shut down Fred by merit of
the phone book you choose. Get enough "big players" to buy into
one phone book over another, and Fred's out of business, and so are all
the other barber shops.

>Yes, I see you fear that big guys will buy lucrative DNS root services
>("lucrative" still feels strange to me when linked to "root services". ;-)
>

Well, there's also a flipside to this, which I just realized: Let's
say (just as an example) a huge multinational company, Irradiated
Banana Mash, decides it wants to do all its business under a new TLD.
Just for argument's sake, let's say that TLD is .ibm. Now, this
company will only use those roots which carry this TLD. Likewise, those
who wish to do business with this company are pretty much forced to
use these same roots. And the first root to decides to carry .ibm
stands to gain. So we're now in a position where:

Companies can create TLDs almost at whim, and
There's a potential trademark/famous mark issue with TLDs.

I'm not at all sure how I feel about that, but I'm not leaning towards
being happy about it right now.

>But the alternative is to establish a strong Internet Governance body. I
>actually could be persuaded that that is "a good thing". However, given
>the reluctance of so many people to even admit that ICANN is governance at
>all, I doubt that we would see ICANN ever become both strong enough to be
>a strong governor. And we certainly see signs that it isn't willing to be
>anything less than a despotic one (albeit one that sees itself as a
>beneficient democracy in its dreams.)
>

Heh. While I wouldn't argue for an "internet governance" body, I would
be in favor of setting things up in such a way that these things can't
happen, or such that it's difficult for them to occur. This doesn't have
to be legislated, or imposed by a governing body, per se; it could be
implicit in the structure of the new root system, I think.

-- 
Mark C. Langston	     			Let your voice be heard:
mark@bitshift.org				     http://www.idno.org
Systems Admin					    http://www.icann.org
San Jose, CA					     http://www.dnso.org
-
This message was sent via the IDNO-DISCUSS mailing list. To unsubscribe,
send a message containing the line "unsubscribe idno-discuss" to
majordomo@idno.org. For more information, see http://www.idno.org/