RE: [tmql-wg] overview query languages
"Steve Pepper" <[email protected]> Fri, 23 Apr 2004 11:24:18 +0200
| Newsgroups | gmane.text.xml.xtm.general,gmane.text.xml.xtm.tmql |
|---|---|
| Message-ID | <[email protected]> |
* Stefan Henke: | I´m currently working on my master thesis about a "topic grid" | (distributed topic maps). This is very interesting. I would like to hear more about it. There is other work going on in this area, both theoretical and practical. For example, Ontopia and a Norwegian company called Bouvet have been sketching out a remote access protocol for topic maps called TMRAP. A brief overview of this work can be found at http://www.ontopia.net/topicmaps/materials/tmrap-spec.html This protocol has been implemented in ZTM (Zope Topic Maps), the Open Source system used to drive a number of public sector web portals in Norway, and also in the Ontopia Knowledge Suite (OKS), which is used in other projects. The goal, in the first instance, is to provide for seamless navigation from one portal to another on the basis of published subjects. The next goal will be to support the exchange of topic map fragments between such systems, and eventually querying across them. All of this is somehow encapsulated in the notion of "Seamless Knowledge"[1], which we are promoting as a moniker for one of the things that Topic Maps is designed to help us achieve. TMRAP will be implemented in the next version of the Omnigator, which means that your Omnigators will be able to talk to each other. | As I´m new to topic maps, I´m not familiar with all parts of | it. I´m wondering at the moment about the current status of the various | standards around topic maps (HyTM, XTM, SAM and soon) and especially a | standard for a query language (TQML). Thats why some questions, I would be | happy getting some answers to: | | - what is the timeframe for TMQL to get published, so that there will be | implementations which support the specification? Four reasonably complete query languages for topic maps have been put forward and implemented: - AsTMa? (Robert Barta) - TMPath (Dmitry Bogachev) - tolog (Lars Marius Garshol) - Toma (Rani Pinchuk) These languages were presented and evaluated at the meeting of SC34/WG3 (the ISO Topic Maps committee) last week. The committee reached the conclusion that it would like to have a language whose core was based on predicates, but with XPath- and XQuery like extensions for compact addressing and output construction respectively. The editors have been instructed to prepare a first draft of this language by October 1st, in time for consideration at the November meeting of WG3 in Washington DC. A document summing up the consensus will shortly be available at http://www.jtc1sc34.org/repository/0502.pdf (In the meantime, you can find it at http://www.ontopia.net/tmp/tmql-0.1.pdf) The target date for a stable TMQL that is ready to go to DIS (Draft International Standard) stage is May 2005. | - as querying topic maps is a main part of the topic grid, I want to give an | overwiew of currently available query lanuages. What do you think about the | most important languages: tolog, astma, empolis, toma, tmpath? Except for | tolog, I couldn´t find implementations of the languages which are available | either freely or as open source. Did I miss something or is this right? I will let Robert, Dmitry and Rani answer this. Empolis' topic maps query language was not put forward as a proposal for TMQL. My understanding is that the reason for this was that they did not consider it to be a suitable candidate, but perhaps Holger Rath can comment on that. | - I had a look at tm4j. AFAK, tm4j supports tolog as query languages. Is | this right? Are there any restrictions or limitations of tolog support in | tm4j? As far as I know (Kal can correct me if I am wrong), TM4J only implements version 0.1 of tolog, i.e. the version that consists primarily of dynamic association predicates and one or two built-in predicates (instance-of() and /=). Version 1.0 of tolog, which was published (and released in the Ontopia Knowledge Suite) in December 2003 has not been implemented ... and probably will not be, now that we expect TMQL to be finished quite soon. | Sorry, partly the questions don´t directly refer to TMQL, but I think this | is the best "meeting place" for all the guys that are touch with query | languages for topic maps. Don't apologize! This was the best place to ask these questions. We appreciate your interest and would like to welcome you to the community. Please tell us more about your work. Best regards, Steve P.S. I have Cc:ed this answer to TopicMapMail, because I think it will be of interest. If there are further postings in this thread please choose *either* TMQL-WG (if it's about TMQL) *or* TopicMapMail (if it's about Seamless Knowledge or TMRAP), but don't post to both. Thanks. [1] Re. Seamless Knowledge: As far as I'm concerned, this is in no way different from what has been called "Global Knowledge Federation", except that it sounds a bit less lofty, relates more directly to problems that people really face today, and is therefore easier to sell :-) I also say that it's "kinda like the Semantic Web, only not quite..." -- Steve Pepper <[email protected]> Chief Strategy Officer, Ontopia Convenor, ISO/IEC JTC 1/SC 34/WG 3 Editor, XTM (XML Topic Maps 1.0)