[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message] RE: Syntax + object model
joshuaa@m... (Joshua Allen) writes: >The real issue here is self-declared defenders of markup complaining >about the need for data models. That's certainly what XML-DEV's >permathread centers on, and leads to all sorts of bizarre comments like >"if you want a data model, don't go near my beautiful XML! Use ASN.1!" Data models aren't necessarily evil. Some data models are intrusive, kludgy, and seem to lack any understanding of markup, but it's very possible for data modeling and syntax-based approaches to live in harmony. For a positive example, I have to say that I'm very impressed by RELAX NG and its underpinnings. I don't claim to understand the math behind it, but I find working with it far more pleasurable than either DTDs or W3C XML Schema. RELAX NG seems to have combined an understanding of what markup does well and the ways people use markup with a model that makes validation reasonable and useful. Because it doesn't change the document content at all, I can combine it easily with syntactic tools to get the results I want. DTDs emerge from an understanding of what markup does, but both do too much (infoset augmentation) and too little (modularization is an interesting challenge.) W3C XML Schema doesn't seem all that connected to what markup does - instead, it seems to be an attempt to capture data in objects or databases and export that vision of data in markup. (XQuery attempts to clean this up, but has the same fundamental excesses.) Both W3C XML Schema and DTDs have modeling problems - W3C XML Schema's model does far too much in an ad hoc tangle, while DTDs do too little, and extending them requires a lot of ad hoc work. The "real issue here" is not so much data models per se - there are perfectly good and useful ways to model marked-up content, depending on your needs. The issue seems more the problem described by the subject line of this message - "Syntax + object model". The original query asked: >imagine XML like a standardized object model PLUS a standardized >syntax. Why would such a standard have less chance to survive? "Object model" is a much stronger phrase than data model, loaded with cultural baggage. If there's a trigger for defense of markup as markup, it's object model, since far too many of the ugliest works created using markup [1,2] are object serializations performed by applications whose developers only care about the data. The problem there lies not so much in the choice of data as in the gross mismatch between best practices for objects (or relational databases) and the best practices for markup. I can and do work with such messes, but it's very difficult to retain even a modicum of respect for their creators. When people claim that it's all just data, pushback seems like the only rational response. [1] - http://lists.xml.org/archives/xml-dev/200302/msg01071.html [2] - http://lists.xml.org/archives/xml-dev/200303/msg00615.html -- Simon St.Laurent Ring around the content, a pocket full of brackets Errors, errors, all fall down! http://simonstl.com -- http://monasticxml.org
|
PURCHASE STYLUS STUDIO ONLINE TODAY!Purchasing Stylus Studio from our online shop is Easy, Secure and Value Priced! Download The World's Best XML IDE!Accelerate XML development with our award-winning XML IDE - Download a free trial today! Subscribe in XML format
|