- From: Olav Junker Kjær <olav@olav.dk>
- Date: Thu, 07 Jul 2005 21:04:21 +0200
Ian Hickson wrote: >>>It isn't a data schema language > The same reason it isn't a spreadsheet language: making it one wasn't one > of the goals of designing the language. :-) Well, it still might have happened by accident! I like to think of the input/select/textarea-elements as semantically describing a data type through constraints. Incidentally, the default rendering of these elements in a web browser is some kind of input widget designed for entering that type of data. (Of course the names of the input types is inspired by the default UI widget (checkbox instead of boolean etc.) but thats just because its easier to remember if you are not educated in formal logic.) > HTML could be used for describing a schema; it just wasn't designed to > do so, and thus would probably not be the best language to use for > that. Perhaps not, but its not that bad either. The types in WF2 is comparable to the basic types in a typical SQL database. Of course its far from as powerful as W3C XML Schema, which supports useful types like NOTATION and ENTITY, sorely missed in WF2. A <form>-element could be considered a description of the interface to a web service. The input elements describes the schema for the input data, while other content is documentation. HTML is the WSDL for simple REST-like web services. While some WSDL tool are able to generate "stub" code to be used by developers when developing client applications, HTML UA's are much more powerful since they are able to generate the whole client UI on the fly, just from the HTML-description. Of course a tool might also be able to generate stub code from a HTML-description, if a developer wanted to develop some custom software to access a service which already has a HTML-described interface. regards Olav Junker Kj?r
Received on Thursday, 7 July 2005 12:04:21 UTC