The International Museum Community's Position in the Internet Domain Name System: Background Information

[Important note: The implementation scheme for the creation of new top-level domains provided below is totally obsolete. The current state of affairs is described on the Internet Corporation for Assigned Names and Numbers ICANN Web site.

Basic concepts

The myriad of computers that are interconnected via the Internet keep track of each other through a strictly regulated system that provides each machine with a unique numeric address. These so-called IP addresses (IP = Internet protocol) are based on the physical division of the Internet into its component networks.

The users of Internet services are not expected to keep track of the intricacies of IP addressing. Each computer that provides Internet services is normally assigned a "host name". Here, as well, rigid control is necessary to ensure uniqueness and locatability for each machine in the global name space.

The mechanism currently used for this purpose is the Domain Name System (DNS), first proposed in 1983. The DNS is a sophisticated yet straightforward distributed database system designed to allow for the, "controlled delegation of the authority and responsibility for adding hosts to the system." This authority is delegated to organizations that may operate several networks at different locations. It is, therefore, very important to note that domain subdivision is "based on administrative authority or organization boundaries (not necessarily network boundaries)."

In 1984, it was determined that the domain name system would be maintained as, "a tree-structured global name space that has a few top level domains. The top level domains are subdivided into second level domains. The second level domains may be subdivided into third level domains, and so on." The initial top level domains included the now well-known COM, EDU and ORG. The proposal also allowed for a series of national domains based on the two-letter country codes in ISO-3166, "Codes for the Representation of Names of Countries".

Provision was further made for the delegation of a top level domain to a "multiorganization" that, "cannot be easily classified into one of the categories and is international in scope." An INT was subsequently created for organizations based on intergovernmental treaties. The hosts used for the registration of network-related information and for the operation of network facilities were given a NET domain. A GOV was originally provided for any governmental offices but was subsequently restricted to agencies of the United States Federal government, with State and local agencies being registered in the country domains.

The detailed mechanics of the DNS need not be discussed here. Every computer that is connected to the Internet needs to know the IP address of one or two "name servers". These are machines that are capable of querying the DNS to find an authoritative answer to the question, "what is the IP address of the machine named WHATEVER.DOMAIN" and returning the answer to the computer that needs it in order to proceed with whatever it is doing. Any organization wishing to provide services via the Internet will, however, need some degree of familiarity with current procedures for selecting, registering, and maintaining Internet domains.

Expanding the DNS

The DNS, as initially designed, ran reasonably smoothly for as long as the Internet was maintained primarily for the academic and research communities. As the Net was opened for commercial use, domain designations started to acquire commercial value. Resulting problems were exacerbated by the explosive popularization of the Internet, not least by the Web boom.

The demand for relevant, easily remembered domain designations has for some time far exceeded the supply. The scope of the contention that has arisen because of this is difficult to imagine. In an attempt to modify the system better to accommodate current needs, a procedure for increasing the number of International Top Level Domains was proposed in August 1996 (by Jon Postel, the author of the initial 1983 proposal).

The current state of the developments that this triggered is that seven new international "Generic Top Level Domains" (gTLDs) will be taken into operation alongside the current COM, NET and ORG as soon as procedures for their operation have been finalized. (The other three-letter domains, including EDU, are restricted to use in the USA.)

The new domains are:

• .firm - for businesses, or firms
• .store - for businesses offering goods to purchase
• .web - for entities emphasizing activities related to the World Wide Web
• .arts - for entities emphasizing cultural and entertainment activities
• .rec - for entities emphasizing recreation and entertainment activities
• .info - for entities providing information services
• .nom- for those wishing individual or personal nomenclature, i.e., a personal nom de plume

The accompanying definitions of the existing gTLDs are:

• .net - for entities emphasizing data networking activities, especially with respect to the Internet
• .org - for not-for-profit entities
• .com - for businesses or firms of a commercial nature

The creation of further gTLDs is envisioned once the results of the current extension can be evaluated. The bodies desginated to control this process initiated operations subsequent to the signing of a Memorandum of Understanding on May 1, 1997. The middle of 1998 seems a reasonable present estimate for the establishment of any additional domain designations.

Unresolved problems

A partial list:

• There are no uniform rules for the management of the top level domains. Differences may be particularly acute in the national domains. Although there is a US domain, it is not used in a manner comparable to the other national TLDs.

• The 3-letter gTLDs are currently managed in a manner that renders the operator of a domain entirely vulnerable to a claim made at any time by a trademark owner.

• Organizations commonly register their second level designations in multiple top level domains -- "just so that nobody else gets them." Inappropriate TLDs are also used because, "the one we wanted was already in use." This significantly deflates the utility of a domain designation as a means for locating the network services provided by an organization. It would be entirely possible to encounter, say, a clothing manufacturer operating BAGGYPANTS.NET solely for advertising and sales purposes.

• Organizations operate multiple second level domains (2LDs) rather than implement third level domains (3LDs). This is often to their own detriment but the practice is rampant and aggravates the other problems deriving from an overcrowded name space.

• The authority that presently operates the 3-letter domains does have rules that would provide relief from many of these problems but does nothing to enforce them. Indeed, a significant reason for the present action was to reduce the difficulty caused by the lack of clarity in the way the existing gTLDs are administered.

It is difficult to see how the new gTLDs will eliminate many of these difficulties. It is safe to assume that organizations which already feel it necessary to register in multiple TLDs will feel no less pressure to do so in the new gTLDs.

The museum situation

Turning, at long last, attention to the state of museum affairs -- where do we fit into the scheme of things?

The rules for registration in the current three-letter domains clearly state that, "museums register under country domains." Indeed, the US domain makes explicit provision for a third level museum subdomain in the form MUSEUMNAME.MUS.STATE.US, thus putting the State Fishery Museum (SFM) in Albany, New York, at SFM.MUS.NY.US. This would be straightforward enough if it weren't for the fact that museums are permitted to register anywhere they want and do turn up in COM, EDU, NET and ORG, to say nothing of the country domains in which they legitimately abound (with the peculiar exception of US.)

The proposed new gTLDs provide no obvious niche, although it is to be expected that museums which feel comfortable in an ARTS context may wish to register there. Virtual museums might feel at home in WEB. It will also be interesting to see if many museums feel themselves to be in the INFO biz.

The initial version of the proposal that led to the new gTLDs called for the establishment of up to 150 new TLDs to be operated by newly created administrative authorities which would be responsible for up to three new domains, each. The possibility of a specific museum domain being created was not without appeal. If, however, its control were to be assumed by a profit-driven authority, the museum community could easily end up being the target of exploitation on a new front. Nor would there be any mechanism for the "authorization" of any organization claiming to be a museum by virtue of domain registration.

In keeping with its mandate to protect the museum community from ill-treatment, the International Council of Museums, ICOM, considered action to maximize the potential benefits to its sector of the impending changes in top level domain structure. No matter who might end up responsible for operating a museum TLD, ICOM wished to be able to exert a clear influence over both the selection process and domain administration. Given the initial uncertainty about what this might entail, ICOM was prepared to consider making a bid for direct authorization to operate a museum TLD.

The initial proposal did not allow for new authorities to operate individual new domains. Each new authority would be expected to operate up to three domains. At that point, ICOM considered acting jointly with other non-governmental organizations (NGO's) within the heritage sector to prepare for the eventuality of coordinated operation of a single TLD authority. ICOM had previously acted jointly with the following other NGO's in specific contexts: the International Council on Archives (ICA), the International Council on Monuments and Sites (ICOMOS), and the International Federation of Library Associations (IFLA). Since several of these were developing network resources comparable to ICOM's, joint action on the domain issue seemed appropriate.

The ICOM Executive Council discussed this matter in the presence of most, but not all, of the other NGO's. The most immediate result of this was a proposal submitted by the present author to the "International ad hoc Committee" (IAHC) that was charged with identifying the scope of the new gTLDs. This proposal deliniated a few Museum and Heritage Sector Interests in the iTLD Delegation Process (Please note that this text has now expired and is being referred to for historical reasons.)

Current alternatives

As it turned out, the initial gTLD proposal was altered in a manner that rendered the NGO action largely irrelevant. Although they are discussing a concerted response to the current situation, for present purposes, museum alternatives will be discussed without reference to neighboring sectors. It should, however, be noted that preliminary discussions within the museum community suggest that a top level HERITAGE domain may prove more appealing for general museum use than would a top level MUSEUM domain. On the other hand, organizations with a clear nominal identity within the heritage sector feel that the concept of heritage is too diffuse to serve as the basis for a disciplinarily coherent Internet domain.

The most significant question at this point is the extent to which the museum community would perceive benefit in establishing domain-level commonality. Not all that long ago, when domain name space was not terribly densely populated, the most prestigeous addresses were the shortest ones. Even if the organization operating SI.EDU might not be immediately identifiable as the Smithsonian Institution, once having learned that it was, few would be likely to forget it. Postulating the impending creation of a MUSEUM TLD, it is not certain that the operators of SI.EDU would perceive any benefit to moving this to SI.US.MUSEUM, despite the clearly enhanced indication of sectorial identity. Indeed, it may be rather brash even to assume any such sense of identity.

Many museums do not currently have the benefit of brief domain names and there is nothing uncommon about URL's such as http://www.supernet.com/museums/downtown/welcome.html or http://www.hypercom.net/~finearts/index.html. Here it more likely that improvement would be perceived in the ability to operate in a manner that more clearly identified the museum in question.

The one extreme case is the museum that, for a variety of entirely legitimate reasons, can see nothing but detriment in a move to a domain name based on sectorial solidarity. The other extreme is the museum that is utterly at the whim of an Internet Service Provider and can see no means for appearing in any degree to being an autonomous network entity, to say nothing of having an understandable and memorable domain designation.

Common ground

As it happens, it is entirely possible for an organization to operate in several network domains. Although this statement is abhorent to some basic networking principles, there are also circumstances under which it can be entirely justified. It is suggested here that the aggregate benefit to the community, of the museum sector establishing a coherent domain identity requires careful evaluation before being dismissed. Within the context of the requisite dialog the following details may be worth considering:

• Host computers on the Internet are commonly referred to by alias designations. For example, although the Swedish Museum of Natural History (NRM) might have no reason to discontinue the operation of WWW.NRM.SE, it would be unlikely to suffer any disadvantage from allowing this machine also to be located as WWW.NRM.SE.MUSEUM or any other name configuration that the museum community might end up adopting.

• Repeated reference is made here to the form, INSTITUTION.COUNTRY.MUSEUM, despite this requiring the creation of a new top level domain. An obvious alternative would be the form INSTITUTION.MUSEUM.COUNTRY. Since there are already top level domain designations for individual countries, this form would not require any expansion of TLD space. The problem here is that the close to 200 national domains all have different rules for the creation of 2LDs. Some use a predefined second level structure that may be patterned on the current gTLDs, while others forbid any generic second level designations. Coordinating the changes that would be necessary to allow for a MUSEUM 2LD in each national TLD would require an effort of staggeringly greater proportion than would be necessary to set up a single new TLD. Similar problems would result from an attempt to extend the model used in the US domain, as few other national TLDs have exclusively regional 2LDs. Although a museum gTLD could be under the control of a centralized authority, each national museum 2LD (or 3LD) would be operated by a separate sovereign authority.

• On an ever more crowded network, the choice of the address at which a resource is located is of increasing significance. The brevity of a domain name has lost its value and current trends favor the acquisition of the most explicit possible designation. Coordinated multi-institutional domain operation within a clearly defined sector has obvious potential for enhancing the locatability of all participants.

  Many would counter that the Net is in desperate need of mechanisms for the location of resources by automated searching rather than fortuitous URL-oriented browsing. It is further often suggested that domain designations will, in the process, lose their significance altogether. Whatever their ultimate fate might be, explicit domain names are demonstrably still the focus of keen attention and it is difficult to see what benefit museums might derive from ignoring that fact.

And then what?

Arguments in support of a coordinated effort at establishing a sector-wide domain identity have been provided here. It is up to the community to decide if any steps should be made towards its implementation. In order to avoid any risk of the process being impeded by competition among previous initiatives within the community, means will be needed for all interested parties to share operation of the domain resource. Fortunately, the new TLD process also requires the development of means for the sharing of its administration. With any luck, the museum microcosm can adopt whatever this sharing mechanism turns out to be and, in the interim, need do little more than observe its development.

Should the matter be regarded as worth pursuing, there are presently a few bases for action:

• There are several 2LDs within the present gTLDs that could serve as a basis for a consolidated museum domain space. Several candidates are already controlled by agencies within the museum community. Since the transferral of domain ownership is commonplace, it should also be possible to approach the owner of an appealing domain that is presently controlled outside our community with a thought towards its acquisition. Museums can, with no further ado, be assigned primary or alias domain identities on third and fourth levels within any pre-existing 2LD.

• A new 2LD could be registered for the same purpose by whatever coordinating body the present initiative might create. This 2LD could either be within the present gTLD structure or anchored within the proposed new gTLDs assuming, as is probably unlikely, that consensus could be reached about which of the new gTLDs might suit the purpose. (At this point it is not even being assumed that consensus will be reached about the desirability of coordinated action.)

• Efforts could be made within the terms of the new gTLD proposal to create a MUSEUM.INT or to create a new MUSEUM TLD. Note, however, that the requirements for becoming an actual TLD registry probably lie outside the scope of what might realistically be undertaken by any museum coalition. Action towards the creation of a new TLD would require the cooperation of one of the new administrative agencies and, as no such agency has as yet been appointed, it is too early to tell if such collaborative action might be reasonable.

• Various current initiatives are designed to increase the locatability of museum network resources. The scope of one or several of these might usefully be expanded to include domain issues.

http://www.remunere.net/musdns.html
Latest update to main text: 3 July 1997

Introductory note added 16 June 2000