Z39.50/RDF Integration ProjectThursday September 16th, 1999A project documented at Mozilla.org aims to integrate Mozilla's RDF capability with the ANSI/NISO Z39.50 specification for information retrieval in a client/server environment. The Z39.50 protocol is used in many bibliographic and digital library applications, and the project aims to make it possible to query the thousands of Z39.50 connected servers from Mozilla, with the results appearing in Mozilla's sidebar. This is another great project brought to you by that supposedly non-existent group of third-party Mozilla developers. One of the project's contacts works for the Z39.50 Maintenance Agency. Good qualification, I would think. Thanks to Joerg Beierle for the news. You know, at some point people should stop adding cute things like library access and, well, make mozilla _useable_! There are tons of features that are simply nonexistant, or unfinished. All of these extras are nice, but at some point someone needs to yell "feature freeze!" And just get the thing out the door. If we keep throwing more junk on Mozilla, it will get big and ugly. I agree with this in principle, but I don't think there are any changes in policy or in action which need to be taken by the core Moz devel team. The Netscape engineers are pretty much solely working on bugs and features outlined by the milestone plan, and IIRC a feature freeze is due up when we hit M12. Also, you have to understand, there is a very steep learning curve to become conversant in even a small chunk of the Moz code -- this means that external developers will find it significantly easier to create extras than to create the core. That said, you do have a point: if we keep throwing more junk on Mozilla, it could very well get big and ugly. That is what the component architecture is all about. The more modules written for the component architecture there are, the more popular Moz ought to be; and for those who like their applications separate, their needs are met as well. Point is, progress on components does not affect progress on the core, but we could indeed use more people working on the core, which is where the lack of external developers lies.... I was under the impression, and I'm wondering if this is true, that Mozilla would be able to have dynamically refreshing content in Web pages without having to manually refresh the page. Like new posts on this page, for instance, without having to reload. Am I wrong on this? Several comments on the Z39.50/RDF Integration Project express concerns that efforts should be directed to completing the core and debugging activities. Although I am as eager as most of the people visiting this site for a usable version of Mozilla, I think Z39.50 support is extremely important. I give two reasons for this: 1) support for technologies like Z39.50 and MathML is the only way to convince the world that Mozilla is more than just "a better mousetrap". 2) Z39.50/RDF Integration has an importance in itself: it merges two fundamental ways to access information. Mozilla with Z39.50 support allows access by navigation on the one hand and search and retrieval on the other. Navigation is the preferred way on the web and it is mostly used for human visual consumption. Z39.50 is about data access on the semantic level. Richard Philips rphilips@lib.ua.ac.be The app developers are hard at work at getting Mozilla feature complete. How would these third-party developers help with that, when they might have a core competency in a different area? Why not allow people to contribute in the way they are best fit? I thought that was the whole point of Open Source development. You got to understand that independent developers will work with things they find interesting. It may not drive Mozilla faster towards release. But it will involve more people in the Mozilla source code and they will discover bugs and enhance the parts they use. You don't see anyone complaining about GNOME why you just made a new application based upon the GNOME libraries. So who's working on the mozilla SSH client? ;) W As the initiator of the Z39.50/RDF project, let me just pop up here to say a few words about this work. Firstly, there is very little the Z39.50 and broader metadata community can do to help stabilise core Mozilla, other than submit bugfixes. Secondly, I personally believe that this and similar projects play an important role for Mozilla. More important than the additional functionality that we're working towards is the sanity check involved with having external developers understand the work of the Netscape engineers. By showing how easy (or hard) it is for outsiders to pick this stuff up and run with it, we aim to tap into the expertise of several wider communities. The RDF 'Enabling Inference' page I put together is a similar example. Within 7 days of proposing this project, someone had created a logic-rules aware plugin for the RDF system in Mozilla. This does several things: it sanity checks the documentation, shows the feasibility of non-netscape developers building interesting and genuinely new information management tools on the RDF/Mozilla core, and brings the open standards (eg. RDF, Dublin Core, Z39.50) and open source communities a little bit closer together. I appreciate the 'why bloat out Mozilla when it still is pre-beta' aspect; please remember that these plugins are optional and (for now) exploratory, and are funded by outside effort. The Z39.50 work is at an early stage but has prompted a fair bit of discussion offlist, which should be moved into a more visible forum. There is a huge amount of programming talent in the Z39.50 and digital library community, and an unrivalled wealth of information locked up in Z39.50 datasources. The Z39.50 project page was the first of three that I've added to the RDF section of the Mozilla site (http://www.mozilla.org/rdf/doc/). All of which (Z39.50, Dublin Core, Inference) are exploring the idea that we can tap into external expertise through using RDF as a common information marketplace. Mozilla right now reflects all RDF datasource (bookmarks, history etc) via an RDF API into Javascript. Things seem to be getting stable enough for other groups to make significant contributions... cheers, Dan -- daniel.brickley@bristol.ac.uk http://www.mozilla.org/rdf/doc/ |