From: joe rockmore ([email protected])
Date: 08/08/01
it seems to me that the approaches described in the article referenced below are all dealing with the idea that when producing a document, you want the author to tag certain elements of the doc in XML without explicitly seeing the tags. since the user interface is just another surface syntax for the underlying XML, i don't think that there is enough of a gain. (i do think there is some gain, as most people don't want to deal with the awkward (to a human) syntax of XML any more than they do with that of html.) i am reminded of the software diagramming wars of a few years ago, when everybody and their cousin had a different diagramming method that they were pushing (this predates UML). all "good" methods had concepts like abstraction and information hiding, as needed productivity enhancers. old-style flowcharting was seen as poor since it was just a different surface syntax for the code, not fundamentally different (i.e., not at a different level of abstraction). in a like manner, tagless editors or the like still are just doing XML tagging, albeit in a nicer human-accessible way. by the way, by now i have seen several examples of similar approaches to generating XML markup in an intelligence document during production without the author seeing any XML. a typical one (done by Office of Naval Intelligence based on JIVA tools) allows the user to highlight a word and then simply pick a tag from a pull-down menu. obviously, this is limited to few tags with little structure (no ontology!), but it is seen by the development organization as an 80/20 solution (80% of the value for 20% of the work). i can get anyone appropriately cleared more info if you want to see it. however, when we start applying this approach to DAML, at least the more featureful aspects of DAML, things fall apart. it is not difficult to get accross to people the idea of a tag, and then to provide a nice user experience to establish the tags. but to get the full power of DAML, you will be doing much more than simply tagging elements of a document. making assertions, that are not specifically associated with any single piece of text (and thus don't fall neatly into the tagging paradigm), is going to need some different user experience models, which can then be supported by nice interfaces. i'd like to hear from those of you who thinking along these lines (like the teknowledge (LA) folks) how you see the "tagless editor" growing into the "DAML-less assertion maker" or some such. ...joe +=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-==-=+ a. joseph rockmore, phd -|- cyladian technology consulting 873 santa cruz avenue, suite 204, menlo park, ca 94025-4629 voice 650/614-3791 -|- fax 650/326-1699 -|- cell/pager 650/714-5696 short email to cell/pager -|- <[email protected]> web site url -|- http://www.cyladian.com +=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=+ At 11:16 AM -0400 7/21/01, John Flynn wrote: > >One of the candidates discussed during the Killer Applications breakout >sessions was a plug-in to support creating/editing ontologies/instances >directly in Word. This email is in the spirit of continuing that general >discussion. There are several Word plug-in approaches for direct XML >editing. I 'm not suggesting the one below is any better than others; >however, the article does an excellent job in describing why this approach >is important. > >If XML is to have the same transforming effect on document-oriented >applications that it has had on transactional, business-to-business >applications, it will have to be easily applied by authors who are experts >in the subject matter, but have little knowledge of markup and less desire >to learn it. This article describes an XML editor that does exactly that: >it allows the rest of us to create valid XML in a familiar context without >extensive knowledge of the rules of XML or the DTD behind the document. > >Full article: > >http://www.sys-con.com/xml/article.cfm?id=219 > >We will set up an email list ([email protected]) on Monday to facilitate >continued discussion leading to a consensus on a DAML Killer Application. > > > Thanks, > > >John >
This archive was generated by hypermail 2.1.4 : 03/26/02 EST