Anyone care to write a few sentences summarizing the fork? I'm interested, but it's just a possible subject to start some discussion. Why it happened, etc., just a summary. I know the archives are available. Touchy subject? Too technical? Sorry, skip it then.
I author an open-source Java project, DocBook XSL Configurator, which is used to create DocBook XSL customization layers. I'm investigating the possibilities for integrating XML storage, retrieval, searching, etc., into the app. I've just started learning about Xindice. It was easy to set up, and following the IBM developerWorks article by Arun Gaikwad (September 2002) made getting started pretty easy too. I've never written code to use a database before, but I think the task will be doable for me. I'm not convinced my app needs it, but I think I'm going to do this anyway just to get some experience with XND technology. Perhaps I should integrate XND access into Epic Editor instead of DocBook XSL Configurator, or both. I'm mostly a tech writer, partly a Java programmer. I've got DocBook XSL Configurator somewhat integrated with Epic Editor already. I have not yet committed to using Xindice, but the fact that it is an Apache project and the fact that it was easy to set up makes Xindice perhaps the one I'll choose. I've investigated only Berkeley DBXML (lots of compiling and setup effort, done now though) and Xindice. I'm downloading eXit and dbxml right now. I didn't even know of those two before this last poster started the fracas. So, some good out of that. I'm open to advice. Any suggestions or reasoning as to what my plans should be? I guess I could try all four of these XNDs and make some decision, and then perhaps I could include support for all of these NXDs at the same time. I see that there is a berkeley dbxml adapter. I don't readily see the value in using an adapter with Xindice. Makes things easier? Am I missing something? Right now, I'm just sort of blundering forward. I guess I need to learn an overview of XNDs. Any links? Anyone care to offer some explanation as to what is different between these four XNDs (Xindice, Berkeley DBXML, dbxml, eXit) and why I would want to use Xindice instead of the others. So far, Xindice is certainly much easier than Berkeley DBXML, but maybe that should not be the only criteria I consider. I think I should use a XND that does what my users would want it to do, but what might that be? I don't really know. I think an XND should handle a high number of documents (thousands or hundreds of thousands), perhaps even large documents (up to about one or two hundred MB, I'm guessing), provide XQuery/XPath searching, check-in/check-out, maybe some maybe some tracking-history-accountability-blame mechanism, maybe some multiple versioning scheme, not fail under heavy load. Maybe that all is asking too much. Here's a question I'm hesitant to ask. Maybe it's a stupid question, but here goes. How might I provide the same facilities to authors for graphics? If I've got the XML and the XSL all nicely managed with an XND, I've still got the graphics left needing the same care. Must I use some other database for the graphics? I see that there is the db4objects database available. Should I somehow force all graphics, even rasterized images, to be SVG, and then stick the SVG into the XND? I'm not sure that can even be done. Just thinking outloud. Thanks, Steve Whitlatch
