Re: NetSol power grab begins in earnest

On Fri, 11 Jul 1997, Carl Oppedahl <carl[_at_]oppedahl.com> wrote:
>
> On 07/10/97, Carol Shepherd <arborlaw[_at_]aol.com> wrote:
> > How is NSI any different from Experian?
> > Experian didn't author my mother's maiden name and my social security
> > number, but it can certainly compile a protectable database of factual
> > information about me.
>
> Certainly? Under Feist? I am not so sure.

Umm.. Experian certainly can't claim copyright in my personal information. THey can claim copyright in a compilation using some arrangment which has some minimal creativity -- but I don't see how Experian could say they are arranging their data in ANY matter (its probably just a big database -- there really is not neccesarily any arrangment there at all). Similarly, I fail to see the arrangment, original or otherwise, in NSI's database. The thing works as a big UNARRANGED database. The whole point is that a DNS server performs an ad hoc query on the database (basically -- I'm simplifying the structure of DNS for the purposes of arugment), and (from what I hear), a bunch of Perl scripts run off and look up the information. Its a dumb search, not an original arrangment.

> Neither the root level server file nor the Whois file contains the IP
> number to which a domain name is translated. In each case, the file
> merely contains the IP number of a domain name server (operated by
> someone other than NSI) that will, upon request, map the domain name
> to an IP number. A good reference on this subject is "DNS and Bind"
> from O'Reilly (http://www.ora.com/).

And there is not arrangement at all. Its not like NSI has made a pretty collage out of IP numbers and domain names...

> > I'm talking about the compilation of the information, not the individual
> > fields and their values.

Right. But Feist explicitly (!!) said that mere compilation in a purely mechanical way is clearly not protecible under copyright. There is absolutely NO human involvement in storing and retrieving the domain name data. The only non-automatically generated information is stuff like address, domain name (which was created by the domain name holder!), etc. The handles for various persons in the whois database are (as said before) automatically generated.

> > If the DND only contains domain names and IP numbers, in two fields, and
> > it is "the set of all such things, in alpha"--then this is looking to me
> > a lot more like a Feist case. If it contains all the WHOIS-type
> > information fields and values, and if there are addresses out there that
> > could be included, that aren't recognized by some servers but are by
> > others (isn't this the .WEB controversy?), then the DND is a subset,
> > involves editorial choice by NSI, and that moves it down the continuum
> > away from being a pure Feist case.

There is no "subset" here. Thats akin to saying that Pacific Bell must have a copyright in the Bay Area phone book because they select a "subset" of phone numbers, specifically, phone numbers beginning with 415/408/510/707 that are in the bay area. Isn't this specifically what was rejected in Feist?

The reason that some TLD's (ie .web) aren't recognized by most DNS servers is that most DNS Servers aren't configured to see the alternate DNS hierarchies ("what?" you say..) Basically, DNS servers work by being responsible for a certain slice of the net. DNS servers answer the question "What is the IP number that corresponds to this domain name" Generally, if a request comes to a DNS server that it doesn't know about, it passes that query up to a DNS Server which knows about "more" of the net (generally, this is done by parts of adomain name). So, here at aimnet.com, there is a Domain Name server (say, ns.aimnet.com), that can answer authoritatively all queries about machines ending with .aimnet.com Queries about a.aimnet.com, b.aimnet.com (ie what are the IP numbers for these machines) are handled by ns.aimnet.com.... If, however, I ask about www.whitehouse.gov, then ns.aimnet.com has to look elsewhere. In particular, it goes to the ROOT DNS Server which tells ns.aimnet.com to go the .gov DNS server(s) which has responsibilty for the .gov domains. The ..gov DNS Server tells ns.aimnet.com "For the whitehouse.gov domain, go ask ns.whitehouse.gov and here's the IP number to talk to ns.whitehouse.gov".... Its quite a bit more complicated than that, but thats the basics.

Whew! So, you see a hierarchy here. The .gov DNS server is "higher up" (and handles a TOP level domain name) in the hierarchy. The ROOT DNS server is the highest of all. ns.aimnet.com in my example has to know where to go to find info on where to find the .gov DNS server so it looks to the ROOT DNS SERVER (which is "above" the top level domain servers)

If the ROOT DNS SERVER(S) doesn't know about .web (because the .web TLD is not approved by the establishment running the ROOT DNS SERVER), then the ..web domain is not visible to the part of the internet which uses the DNS hierarchy underneath the ROOT DNS SERVER in question.

In other words, the ROOT DNS SERVER simply claims ignorance. (and NSI doesn't register domain names for .web, for example). Thus, it is not "filtering" or making editorial choice, it is simply a matter of not configuring the DNS root servers to respond to .web queries and not interact with the root servers which do know about .web, and NSI only registering certain domain names.

> We're going to have to do better than a string of "if"s here. I am
> unaware of anything that would budge NSI's position away from the Feist
> end of the continuum. I am unaware of any creativity exercised by NSI
> in compiling the database. I am unaware of any discretion exercised by
> NSI in compiling the database.

I agree with Carl completely (!!)

> A company that is compiling, say, a list of phone numbers in the 212 area
> code isn't exercising discretion when it excludes telephone numbers that
> are in non-212 area codes. In the Feist case the phone company was
> compiling numbers in some town in Kansas. Did the Supreme Court given
> them "protectable expression" points for excluding Michigan phone
> numbers from that directory? I think not.

Agreed as I said above.

> > > In what way does NSI decide what can and can't be in the Whois
> > > database? Are you saying NSI could, at a moment's notice, delete
> > > the America Online entry from its database? Is that what makes
> > > NSI's case more like the DJ case?

NSI could delete the AOL entry, but I don't see how this has anything to do with creativity. The fact of the matter is that if NSI went around arbitrarily messing with the DNS database (which seems to me is the only way they would be able to claim any sort of originality), they'd lose control of the DNS registration within days. Its rather scary that they have that power..

> > NSI exercises the discretion to deny registration to some domain
> > names that are not already taken, from what I understand. Remember
> > "ADULTSRUS.COM"? This looks a lot like editorial control over
> > inclusion in a compilation to me.

I don't know exactly on what bases NSI retains the right to deny registrations, but I guarantee you they are functional and/or legal reasons, and Carl's comments in the next paragraph are right on target.

> Copyright never protects functional things. NSI's decision not to
> register, say, ADULTSRUS.COM is a decision about function, namely that
> such a domain name will not physically function on the Internet. Then,
> as an echo of that functional decision, it happens that ADULTSRUS.COM
> doesn't turn up in the computer file.

        -Gabe

Gabriel Wachob
<gwachob[_at_]aimnet.com> Received on Tue Jul 15 1997 - 02:43:03 GMT