To: process.mail@wipo.int
Subject: WIPO2 RFC-3 Comments
From: "Daniel R. Tobias"
Date: Sun, 15 Apr 2001 15:20:10 -0400
WIPO2 RFC-3 Comments by Daniel R. Tobias --
dan@dantobias.com
I'm not the sort of entity from which comments to WIPO or ICANN usually emanate, or about whom WIPO or ICANN give any sign that they care when drafting policies regarding Internet domain names. I'm not a government, corporation, or organization. I'm neither a trademark owner nor a domain speculator. I'm not trying to get rich off of either the current domain system or a proposed future system, nor am I trying to protect my current economic status against threats from either the current or a proposed future system. I'm merely an individual who has been familiar with the Internet since it was still the ARPAnet, and has been involved with it for years as a user, a hobbyist, and a professional developer. While I've made my living from the Internet for years now, I've never attempted to get rich from it (and, hence, haven't lost my shirt at it either, as have some "dot-commers" these days). I have nothing in particular to gain or lose economically through the evolution of domain name system policy, unlike most others who write comments to these RFCs. I'm writing merely from my own conceptions of why the domain name system was created in the first place and how it was intended to be used, and the ways in which it has been abused in recent years (leading to much conflict), and the ways which have been proposed to resolve these conflicts (in some cases just making the situation worse).
Unlike most of the governmental, corporate, and organizational respondents, I'm submitting my response in plain ASCII text, rather than as an MS Word, PDF, or other specially-formatted document. This guarantees that, when it's put up on the WIPO website, everybody will be able to read it no matter what browser they use and what auxiliary viewer programs they have. This is in keeping with my "Keep It Simple, Stupid" philosophy -- getting the information across is more important than being fancy and flashy. If people had been using the domain name system with this philosophy, too, things would be so much better.
I can remember when I first heard of the domain name system, back in the mid-'80s when it was first implemented. My feeling was that there was a big gap in the naming system -- no top- level domain existed for individual computer hobbyists, just for various categories of organizations such as government, military, educational, commercial, etc. I wished they had created a TLD for hobbyists, perhaps ".hob". While I understood that the ARPAnet of the time didn't permit anyone not affiliated with an organization from gaining a direct connection, I expected that this would change over time, and the naming system ought to accommodate it. As it turns out, I was thinking too narrowly myself. Current personal use of the Internet has expanded vastly beyond the computer hobbyist community. Finally, the proposed ".name" TLD provides a proper namespace for such use, though I would have preferred the earlier-proposed ".per" for "personal" (.name sounds silly to me -- aren't all domain names "names"?).
Though I found gaps in the namespace, I still understood its purpose and proper use. It was to replace the earlier chaotic naming of Internet hosts in a flat namespace, where every machine in the world that was on the net had to have a unique name -- if somebody at MIT named their net-connected computer "Foobar", then nobody else could. This was solved by creating a structured namespace where each entity with a net presence could have its own domain it could use and subdivide as it wished. There could be separate machines at foobar.mit.edu and foobar.cmu.edu without conflict. There also could be separate entities at foobar.edu and foobar.com -- one of them a university and the other a commercial company. With several different top level domains, and the unlimited opportunity to create subdomains and hostnames within any domain to the desired level of hierarchical nesting, there would be plenty of opportunity for anybody on the net to obtain stable and meaningful names. A nonprofit group called "FooBar" could obtain foobar.org, then delegate subdomains like miami.foobar.org and boston.foobar.org to its chapters -- they'd all have logical names, and so could the completely independent commercial outfit that also happened to be named "FooBar" and which could have its own site at foobar.com.
This system started breaking down when the Internet became commercialized in the mid '90s. A large influx of newbies arrived who were unfamiliar with the proper structure of the domain name system, and unfortunately, the commercial entities who were driving the expansion of the net found it more profitable (at least in the short term) to pander to their ignorance than to try to educate them out of it. Because a large number of commercial sites came onto the net at addresses of the form www.SomeName.com, the general public became convinced that all web addresses were of this form, and so all the "marketing types" then felt the need to obtain separate domain names for every single site, subsite, product line, or marketing gimmick, rather than to use subdomains like the logical structure of the DNS intended. Eventually, even nonprofits and governmental entities started getting .com addresses, though this was really stupid given that they weren't commercial, because "that's where the public expects to find the site." You've now got idiocies like "navy.com" for the U.S. Navy recruitment site -- even with the U.S. military's monopoly over the .mil TLD, they still feel the need to clutter up the namespace of .com, though they're not in any way commercial.
Naturally, with everybody scrambling to grab names in what they perceived incorrectly to be a single flat namespace, much conflict ensued. Once governments and nonprofits started thinking that their Internet sites ought to be in .com rather than the properly structured namespaces created for their sorts of entity, they got peeved if some commercial site managed to grab "their" name first. Maybe barcelona.com and southafrica.com are being used legitimately as commercial sites about their respective cities or countries, and the governments of those respective places ought to have their sites in their appropriate country code domain, but because the drooling imbeciles on the net these days expect everything to be in .com, commercial or not, then the government had better sue to get the name "back". This has produced lots of bad cases, both in the courts and in the ICANN arbitration process, where big governments, corporations, and others with money and power have tried (sometimes successfully) to bully a legitimate user of a domain into giving it up because it happened to resemble their name, even if it wasn't even in the correct TLD for the complainant organization.
Of course, domain name holders aren't always the "good guys" either. Many are speculators trying to get rich off of domains named after corporations and trademarks, either by selling the domains to the corporation or by "typosquatting" to draw traffic to some sleazy pseudo-portal that no Internet user would intentionally go to. For many of the domain disputes, I say "A pox on both your houses," having little sympathy for either side. If domain names were used as originally intended, as a manner of giving logical and stable addresses to things on the Internet, then in most of these cases neither the complainant nor the respondent would have any legitimate need for the domain they're fighting over. FooBar, Inc., which already owned foobar.com, could logically name the site of its Memphis branch office memphis.foobar.com, and wouldn't have to worry about whether some cybersquatter grabbed foobar-memphis.com already. The cybersquatter, on the other hand, has no legitimate need for this name either.
So what to do about the whole mess now? How about admitting that putting everything in SomeStupidGimmickName.com just won't work as a long-term scalable solution, and trying to educate people about the fuller structure of the system? Companies and organizations of all sizes can help by putting up their sites under subdomains where appropriate -- every subdomain name that's advertised to the public helps educate them that such things exist. A company that took the high road and used logical subdomains of its main domain for all of its sites could then make a point of this in their advertising by saying to "Accept no imitations -- Only sites of the form Sitename.Foobar.com are official sites of the FooBar Corporation!" Once this point has been driven across, FooBar would have little to fear from cybersquatters adopting names with FooBar as a substring.
The addition of new TLDs will also help, by presenting the public with more names ending in things other than .com, making them think a little more about how the names are structured. It's desirable to adopt new TLDs with clear meanings, not just generic substitutes for .com like .biz -- adding new generic names will simply result in the same group of trademark owners registering them in addition to the other TLDs they already have, or filing challenges against others who get there first, but won't expand the namespace in any meaningful way. Thus, I actually like .museum and .aero better than .biz and .info, out of the current group of new names to be added -- though their application is very limited, at least it's clearly defined. Some more well-defined TLDs, hopefully with broader application, would be desirable. Ones I'd like to see are .fan for fan sites (e.g., about celebrities and genres) and .sucks for protest sites. Some TLDs can be explicitly defined as being for noncommercial commentary, where the presence of a name in them does not imply endorsement by the entity having that name as a trade name or trademark. I don't know if the lawyers can be kept at bay by this in the present climate, but at least it could be tried...
This doesn't mean that the UDRP ought to be repealed. There's still a valid function for a dipute resolution process in cases where somebody intentionally and abusively registers a misleading name to try to profit from somebody else's trademark. But this process should be limited to a narrow category of clearly abusive registrations, not for every case where two entities both claim to have rights to some string of characters. The policy as now written covers the relevant cases very well, if it's interpreted as written (which, unfortunately, the panelists haven't always done; sometimes, they've stretched points very far to achieve their desired result). There is no need to expand it to cover cases outside the realm of trademark rights, as the current proposals do.
More on the specific things that are being proposed to regulate:
International Nonproprietary Names for Pharmaceutical Substances:
Well, these are by *definition* nonproprietary... duh! Thus, they belong simultaneously to everybody and nobody, just like any other generic word in English or any other language. Therefore, "first come, first serve" is the only rational way to deal with them. Just as whoever registered "pets.com" first has the right to keep that name, develop a pet-related site there or sell it to somebody else who wants to do so (and the fact that the current owner of that name just went bankrupt is beside the point...), whoever gets one of these nonproprietary drug names first owns it in that particular TLD (but doesn't gain trademark rights to it in any other context, including in other TLDs). If he wants to use it to sell his version of that drug, or to provide generic information about the drug, or to warn people of the dangers of the drug, or to run an avant-garde artsy site having nothing to do with that drug just because the site developer happens to like the sound of the name, that's his own business. Maybe that'll give the registrant an "unfair" advantage over other sellers of the drug, but them's the breaks. Life ain't always fair. Imposing a heavy- handed exclusion over all of these names in all TLDs is an example of the "nuclear flyswatter" approach, dealing with a real or imagined problem with vastly excessive force. And why do generic names of drugs deserve more protection than generic names of any other kind of object or substance? Maybe all words in the unabridged dictionaries of all human languages should be excluded too?
Names of International Intergovernmental Organizations:
Why are you limiting it to that, anyhow? Even *intra*national *intra*governmental organizations seem to want to control "their" name in all global TLDs these days. I say, screw 'em if they didn't get the name they wanted first. Let them use the properly structured name in .int, if they're an international treaty organization, or in their own country code if they're an agency of a particular nation's government. If more such groups do so, the public will gradually learn where to find these sites instead of stupidly expecting them all to be in .com or .org.
Personal Names:
The new .name TLD should be helpful for this, if it's not abused by corporate trademark owners trying to preclude anything they think is an "infringement" -- it would be asinine if McDonalds could stop all people named McDonald from putting their personal site appropriately in .name. Some use of the UDRP in .name would be desirable in the case of attempts to hoard or speculate in sites with names other than the registrant's actual name or nickname.
In other TLDs not specifically for personal names, no special protections for such are needed or desired. If a personal name is being used as a trademark or service mark (whether registered or unregistered), as is the case with many celebrities who have merchandise using their name, then they should have the same rights as any other trademark owner, but shouldn't be able to prevail against a less-famous person who is also named the same thing (as the musician Don Henley has been trying to do against a different Don Henley who has his personal site at don- henley.com).
When two people are named John Smith, or Don Henley, then first come, first served should always rule regardless of the relative fame of the people, excepting only highly abusive cases where the less famous party actually used the domain name to intentionally mislead people into thinking he was the famous person of that name.
Some consideration also needs to be given to noncommercial fair use of celebrity names for the purpose of fan sites or commentary sites. Perhaps, as I mentioned earlier, new TLDs like .fan and .sucks should be created for positive and negative independent sites about a celebrity. But for now, I regard .org as the most sensible place for noncommercial fan sites, and think that any such sites should be allowed to continue, especially if they contain disclaimers that they are not the celebrity's official site. There's more justification to challenging the use of a .com domain by an unauthorized fan, as that TLD implies commercial use, something which should not be done with a celebrity's name without permission (other than in limited cases such as journalistic use).
Geographical Indications:
These should be treated like any other generic word -- whoever gets them first should be allowed to keep them. They shouldn't be regarded as proprietary. Of course, within country code domains, the laws of the country in question apply, and maybe in some countries place names are proprietary or excluded from domain registration. But in gTLDs, the first-come, first-served rule should be maintained. There are plenty of sites named after cities, states, countries, etc. which are being used in a very reasonable manner to provide information about that place (either commercially or noncommercially), or as the site of a person, company, or organization which happens to have the same name as a place. Others, however, are held passively by speculative cybersquatters, but imposing heavyhanded regulation on this would be another case of a nuclear flyswatter. Once again, the appropriate governmental authorities of the place should be encouraged to use the properly structured country code domains for their official site, like ci.miami.fl.us for the official site of Miami, and not worry about who else happened to grab miami.com, miami.net, miami.org, and miami.WhateverElse.
Trade Names:
Actually, domains have more rational correspondence with trade names than they do with trademarks, as, in the original structure of the system, they were intended to represent the organizations on the Internet, not their products and services -- any sites for particular products and services ought to be subdomains of their owning company's site.
A trade name should have some protection against abusive domain registration by others, but not to any stronger extent than is currently true of trademarks. Somebody registering another company's name as a domain name with intent to profit from this association and lacking any rights to the name themselves deserves to be challenged under the UDRP, but it's an abuse of the system for one company with a given trade name to initiate a UDRP case against another company which also has a similar name -- trade names are not globally unique. Once again, first- come, first-served should rule.
For more of my comments and links regarding the domain name system and its structure and conflicts, see my site at: http://domains.dantobias.com/
Daniel R. Tobias
Boca Raton, Florida
April 15, 2001
--Dan
Dan's Web Tips: http://www.dantobias.com/webtips/