Re: [ICANN-EU] Re: Don't waste your endorsement

[lutz@iks-jena.de (Lutz Donnerhacke)]

* Marc Schneiders wrote:
>On 7 Sep 2000, Lutz Donnerhacke wrote:
>> So a de jure solution has to be found. No requirement for new TLDs.
>
> Good luck with the way you think this can be handled. How is ICANN going
> to pay all those lawyers they will then need?

As typed again and again the the last few days: They will need lawyers
anyway and will always lose with there contracts against (inter)national
law.

> Maybe by introducing new gTLDs with the customary tax of $0.33 for each
> domain? (Did you know about that by the way? Not many seem to do.)

I heard about such proposals but--as you can imagine--I was not able to
follow all news (and olds) while laying down at doctor's couch. So I'm
really proud of nearly dealing with all mail of the last weeks. If I missed
some questions, please mail me.

As you may noticed on the IPv6 debate, I'm far from perfect but able to learn.
A short search on icann publications did not come up with results regarding
this tax (mailing lists searched, too). So please drop me an URL.

>I did not and I do not intend to. Being on a Board makes a person less
>free to reiterate his own views, once it is clear they will not prevail, 
>ad nauseam. It is a fact.

You are surely right. But should this change my mind? Explaining my true
views to the topics allowes you (and others) to preestimate my opinion on
similar questions. If you can not imagine myself in this board, please
endorse a better suited person. I'll not change my views in order to become
elected.

>>> In a concensus model one has to be able to say: OK, we have it your
>>> way. And then try to participate in making the best of it. Being on
>>> the ICANN Board will mean having to accept the introduction of new
>>> TLDs and working towards a decent launching of them. (I may go on
>>> saying NO :-)
>> 
>> Definitly. But you miss the discussion and voting period in such a board.
>> There personal opinions are very interesting. I'm pretty sure you do not
>> expect somebody saying 'Do what you want, I'll support the result anyway.'
>> as a director.
>
>It won't come as a surprise that I am all with you here. My point is
>above.

And so my answer is: I can actively support compromises I feel bad with.
But I'll not completely switch to the other side after losing a battle.

>> Somebody not able to working with consensus results even if they are
>> contradictory to his own opinion will fail as a director. It might be a
>> reason to kick him off the board. This does not imply that every decision
>> made is indisputable.
>
>No, witness UDRP.

There are board members actively fighting against UDRP after passing it? URL?
I probably misunderstand you.

>> Yep. This is the easiest solution. But it is a bad one for the net. The
>> technical problems arising will not be solved by ICANN or the registries,
>> but by the admins ...
>
>What technical difficulties? Really, it will not mean that more DNS
>traffic, which will, I presume, be distributed among a larger number of
>servers (new ones for the .xyz and .123 gTLDs I mean).

'host.sld.tld' will be resolved as 'host.sld.tld.my.local.do.main' first.
In order to keep reachability to new TLDs you have to carefully remove
existing DNS entries or deal with customer queries. This problem becomes
more prominent if the TLD is a trademark. So lawyers may sue innocent
companies for there internal naming scheme.

If a TLD is used by itself as a hostname (ccTLD ai doing so, there I got the
idea for bofh from), those issues exponentiate.

Impersonating and privacy issued not mentioned.

>> In the short term (two years). Then we will have an expotentially increasing
>> number of TLDs including mercedes, coca-cola, cola, coca, germany, icann,
>> "brand", ... and a deep performance impact on the root name servers.
>
>Well, let's archive this and decide who is a true prophet in a couple of
>years time :-)

Yep. Hanno, please take over.

>> The reason for structuring the DNS is to keep a good performance on
>> queries and updates. Can you imagine what happens if the query
>> performance goes down? Can you imagine the legal consequences of
>> overloaded servers unable to process updates timely?
>
> Yes, but who should this happen? Remember a huge amount of the domains
> registered is not used and hence does not generate DNS traffic.

A deeper level domain is not queried from the root servers.

> Also many new companies etc. will register MyName.Web in stead of
> MyNameIsThisButItWasGoneSoITookNumber1234.com. Again no extra traffic.

How long will a new TLD be free to do this? Most dotcom companies will
register the same name as soon as possible under the new TLDs. So you will
increase the trial and error traffic.

> The extra traffic comes much more from the growth of the internet than
> from new names in the DNS. My guess.

You might be right. But if the domain names become shorter (and ultimately
TLDs) the whole DNS query traffic will be more centralized. DNS is
hierarchical in order to prevent this.

If more and more TLDs are announced, I'll change DNS to my hash idea
proposed as a kind of joke three weeks ago: Instead of querying for
'xxx.yyy.zzz' you hash 'yyy.zzz' with (say) MD5 and query for
'xxx.HA.SH.ED.SL.DN.AN.DT.LD...'

> And again: I assume there will be separate servers for the zones of those
> new gTLDs.

man Single Point of Failture.

>The TLD has an IP address assigned to it? Yes, and? Do it with . if you
>like: lutz@. 

The sentence "God is at <.@.>" is a old running gag in the Usenet.