At 01:41 PM 2/19/99 -0800, Greg Skinner wrote:
>Dr Eberhard W Lisse <[EMAIL PROTECTED]> wrote:
>
>>> I happen to agree with the opinions expressed by Keith Moore, as they
>>> match with my experience.
>
>> So I happen to disagree, from my experience.
>> And?
>
>In the absence of hard data, people will argue their opinion based on
>their experience. No one knows for sure what will happen when the
>safe TLD limit is exceeded. Anyway, even if TLDs are added a thousand
>at a time, I've not seen any evidence that they will suffice to
>qualify all the trademarks or service marks that are necessary to
>avoid DN lawsuits.
You are arguing apples and kumquats. Mike Krieger posted some test results,
that were run at MCI, on the InterNIC DOMAIN-POLICY list (damn, I should
have saved it!). My own investigation into BIND AND the FACT that there are
over 1.5M SLDs in COM kind of prove that the sky *is NOT* falling. el is
right, the only limit *is* horse-power. Bandwidth isn't even much of a
factor. In addition, if every one used local root.zones the load is reduced
almost one order of magnitude, as well as taking the load off the
root-servers.net machines. In fact, if everyone did that, we wouldn't need
the root-servers.net.
>>> But even you would have to admit that by adding lots of TLDs, extra
>>> orders of magnitude are induced in the DNS process, because queries
>>> are going further up the tree than they would otherwise be.
>
>> So What? More horsepower. BIND can take it. And if it's really a
>> problem add on a SQL engine. Which one should perhaps do anyway.
There is where el may be wrong, the DNS system is *bunches* more efficient
and faster than an RDBMS. Otherwise, gregbo's argument is not correct and
purely supposition.
>> More Horsepwower...
>
>Once upon a time, there were some people who thought that if you added
>more bandwidth to the Arpanet, the congestion problems that were
>occuring at the time would go away. However, it took some studies by
>a control theorist to show that changes needed to be made to the TCP
>protocol to relieve congestion.
Nice fairy tale, however, it does not apply here. There has yet to be shown
a problem with even 10,000 TLDs. I personally loaded up to 1960 TLDs, on my
test rig, while investigating BIND, before I got bored with not finding any
errors. I even tested this across a 33.6K modem link, instead of my
100baseTX backbone. It was slow, but no problems. No I didn't take notes
and I'm not doing a rigorous report. There are limits to what I'll do with
out pay. You'll just have to take my word that I did the investigation. My
staff did part of the work anyway.
>You can't just throw more hardware or software at a problem and expect
>it to go away (unless you *know* that this will solve your problem).
>System-wide analysis is required.
Sure, but first you'll have to prove that there is a problem, Chicken
Little. Show me a failure mode that I can repeat. Point to code that shows
the architectural flaw. Yes, there is one small section, in the caching
code, that is slightly non-deterministic in certain conditions. However, my
personal examination did not yield any failure modes in the code. testing
specifically, and generally, also did not reveal any flaws.
>Also, it seems to me there's been a fair amount of objection to stiff
>requirements for operating a TLD registry. Adding more TLDs would
>certainly raise the bar, in terms of processing and bandwidth
>requirements.
Gregbo, I like you, but I have to say that is pure hokum. FYI MHSC is
registering our new tld, with USPTO, next month (decision just made by the
BoD today), for an infomediary we are planning to deploy. We don't intend
to ask ICANN's permission. If anyone else tries to use it, they'll be on
the wrong end of the trademark suit from hell. We have other moves in the
works as well.
>In such an environment, the well-heeled companies will
>be much more able to operate TLD registries than the struggling
>entrepreneurs. We might very well wind up with only a few large
>companies as registries, because the others just won't be able to
>survive financially.
I have never seen a market for a pure TLD registry, as a stand-alone business.
___________________________________________________
Roeland M.J. Meyer -
e-mail: mailto:[EMAIL PROTECTED]
Internet phone: hawk.lvrmr.mhsc.com
Personal web pages: http://staff.mhsc.com/~rmeyer
Company web-site: http://www.mhsc.com
___________________________________________________
KISS ... gotta love it!
