[XSL-LIST Mailing List Archive Home]
[By Thread]
[By Date]
Re: CSS + behavior vs. XSL (was: EcmaScript, gone?)
Subject: Re: CSS + behavior vs. XSL (was: EcmaScript, gone?) From: Andy Dent <dent@xxxxxxxxxxxxxxx> Date: Thu, 3 Sep 1998 03:40:13 +0800 |
>Our only disagreement is whether XSL >should live in the realm of the ad hocery, waiting for a fully declarative >replacement, or live in the realm of the declarative and survive for the >long term. Either way, you can do what you need to do through scripting. Another perspective on the loss of scripting comes from a recent painful lesson of our own. Our OOFILE report-writer is a (ahem) clean c++ design, using several classical OO design patterns. It has renderers for Mac and Window (screen plus printing) and RTF and HTML. Early this year we designed an extension mechanism into the report-writer using the common pattern of 'adorners' (combination of Strategy and Command, for the Design Pattern aficionados). It was in many ways comparable to the use of scripting in XSL. What we swiftly found was that people readily used adorners to customise behaviour in Mac and Windows, but where the rendering paradigm was significantly different they failed to implement them for RTF and HTML. Indeed, because of the paradigm shift in rendering, the adorner model made some things almost impossible to implement. Even for relatively simple jobs like varying the text colour of a cell depending on a value, the model failed to be extended. Our current redesign, apart from some XML focus, is on a vastly more declarative approach. Even though some user code may still be involved, it is making almost trivial value judgements and so in future can be replaced by simple 'behaviour sheets' of some kind. One benefit in the long term will be that a simple declarative model for behaviour should be user customisable at runtime. I do not regard any variety of scripting as an acceptable customising mechanism you should inflict on the average user. (I'm a part-time UI designer with a lot of study time in human interface design, as well as having taught many beginner classes in Excel. I'm not just being a snob.) Andy Dent BSc MACS AACM, Software Designer, A.D. Software, Western Australia OOFILE - Database, Reports, Graphs, GUI for c++ on Mac, Unix & Windows PP2MFC - PowerPlant->MFC portability http://www.highway1.com.au/adsoftware/crossplatform.html XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
Re: CSS + behavior vs. XSL (was: E, Paul Prescod | Thread | Re: CSS + behavior vs. XSL (was: E, Chris Lilley |
Who can have qualifiers in XSL?, Steve Dahl | Date | Re: Who can have qualifiers in XSL?, Henry S. Thompson |
Month |