xmlhack: XTM in quest for a new home

After the announcement that GCA/IDEAlliance will stop to support XTM (XML Topic Maps), the project is looking for a new hosting organization and the names of OASIS, TEI-c and KnoW have been the first to be mentioned.

Announced by Paul Conn, GCA Vice President and IDEAlliance Chief Technology Officer, the decision appears to be motivated by the independent "business model" of the XTM project, which made it difficult for IDEAlliance to be involved in the project or to benefit from it:

Under the current model, IDEAlliance has none of these responsibilities [project management, administration, budget and marketing] and therefore neither provides nor receives much value, and we believe our resources can be better directed elsewhere

Even a least one observer, Patrick Durusau, has suggested that the decision might have been guided by personal matters:

I will refrain from commenting on the almost certain origin of this arbitrary decision by IDEAlliance and the childish petulance that urged it.

It seems likely that Steve Newcomb and Michel Biezunski, both influential GCA members and founders of the XTM project, were the main cement that bound the XTM project and GCA together -- and that after their departure from XTM the two pieces had no more reasons to stick together.

Other stories:

Re: XTM in quest for a new home (Michel Biezunski and Steven R. Newcomb - 23:14, 6 Apr 2001)

[Mike Champion:]

> I'm confused about topicmaps.org, topicmaps.net, the > original ISO topic maps activity, and this > announcement. Is this a personality clash, disagreement on > principles, disagreement on technology, or what?

[Michel Biezunski and Steve Newcomb:]

None of the above. As far as we know, there is no contradiction between ISO's, Topicmaps.Org's, and Topicmaps.net's views on the nature of topic maps.

- The original ISO Topic maps activity, started in 1995, ultimately resulted in the publication in January 2000 of a standard, called ISO/IEC 13250:2000 Topic Maps, that represents an industry consensus about what topic maps are and how they can be interchanged. Michel originally proposed the project to ISO and worked for about 4 years to build a consensus about it, together with Martin Bryan and Steve Newcomb. This was a long process, but it worked.

- In January 2000, we (Michel and Steve) were persuaded that a subset of the ISO standard should be created that would be adapted to the needs of the Web community. It would be expressed in XML, it would use URIs, etc. Since we wanted it to be created quickly (and we had just concluded a five-year-long ISO process), we created an ad hoc organization, called "TopicMaps.Org". Its mission was to publish a spec within one year. Under our leadership, TopicMaps.Org was created, chartered, supported, and sponsored, and the group actually delivered an XTM 1.0 Specification in December 2000 at the XML 2000 Conference in Washington. Please understand that Topicmaps.Org was not set up to oppose ISO; XTM is an XML-oriented derivative of ISO/IEC 13250:2000, and the XTM DTD says so. As far as we were concerned, the purpose of TopicMaps.Org was to create the XTM Spec *quickly*. ISO is still willing to develop activities related to topic maps, and the high value of the XTM DTD is very well understood by ISO. As far as we know, there is no unhappiness with ISO at TopicMaps.Org, either.

- There is an increasing amount of work going on in Topic Maps Land. There are real-world implementations, product developments, and the need for enriching the standard with add-ons (in a way that is somewhat similar to XML's enrichment via XLink, query languages, etc. -- this list is very incomplete, and no slight is intended by any omission).

- Many people are working to establish a common understanding about the exact nature of topic map information. The dream of global knowledge interchange depends on the idea that, at some fundamental level, the interpretation of interchangeable topic map information is uniform among all applications.

However, the dream of global knowledge interchange and federability does *not* depend on any uniformity with respect to the methodologies used to create, maintain, render and use topic maps, or even on the idea of a uniform API to topic map information (although some agreements about APIs might be very helpful in jump-starting the industry, just as the DOM and SAX have been extremely helpful for the XML industry in general). These are all areas where there must be freedom and incentive to innovate, and where global standards are not, strictly speaking, necessary and may, in some cases, even be inadvisable. In all these areas, diversity should be encouraged and cherished.

- As far as we know, there is at least one monumentally important area of general agreement throughout Topic Map Land: the serialization syntax for interchanging topic maps in XML, i.e., the XTM 1.0 DTD, which we, along with *almost* everyone else we know, firmly believe should be considered as the basis for implementations, for users, and for creating an information interchange industry that relies on topic maps. The existence and adoption of the XTM DTD is one of those rare minor miracles that make sense technically, politically, and economically. We support it and we recommend it unequivocally. The XTM DTD can and should be used *now* as the basis for interchanging topic maps. There is nothing that prevents anyone from doing so. It would be even better if TopicMaps.Org could provide some guarantee that the DTD is not going to change soon, in order to let the industry start. We think this is what users care about now. Other aspects of topic maps will emerge when they are mature. Anyway, a foundation (the XTM DTD) and a sub-foundation (the ISO 13250 standard) already exist. Let's use them.

- www.topicmaps.net is a web site where we, Steve Newcomb and Michel Biezunski, have published the current state of our thinking on the interpretation and meaning of topic map documents. We think it will be useful for making topic maps sharable and federable at large scales. What we are publishing on this web site has no normative status; as it describes itself, it's merely "our work in progress for your information".

On a different "topic" (please excuse the expression):

We don't know what Patrick Durusau meant by his remarks about GCA/IDEAlliance's decisions regarding TopicMaps.Org. However, some have interpreted his remarks as a veiled accusation that we, Michel and/or Steve, exerted our influence at IDEAlliance in such a way as to cause harm to TopicMaps.Org.

Neither of us did any such thing.

Also, contrary to what Eric van der Vlist says in his article, neither of us is currently a GCA member. It is true that we have each worked for GCA from time to time on a contract basis, and we have both found much to admire about the contributions that GCA has made over many years to the information management and publishing industries. Like many others, we have also participated in many GCA conferences, and we have led several tutorials under the GCA banner.

(signed) Michel Biezunski and Steven R. Newcomb

========================================== Michel Biezunski, InfoLoom Tel +33 1 44 59 84 29 Cell +33 6 03 99 25 29 Email: mb@infoloom.com Web: www.infoloom.com ==========================================

Steven R. Newcomb, Consultant srn@coolheads.com voice: +1 972 359 8160 fax: +1 972 359 0270

405 Flagler Court Allen, Texas 75013-2821 USA

I'm confused about topicmaps.org, topicmaps.net, the original ISO topic maps activity, and this announcement. Is this a personality clash, disagreement on principles, disagreement on technology, or what?

Is there any reason to expect that XML topic maps will emerge from this as a viable, standardized technology that real businesses will care about?