- From: Wingnut <wingnut@winternet.com>
- Date: Wed, 29 Jun 2005 13:19:46 -0500
- To: www-style@w3.org
Paul Duncan wrote: > <section removed> > > It very frustrating as someone who has to work with these standards every > day. They are not intuitive and overly complex. I came here to try and > understand. > For myself as well... and likely many others. One thing I find I need to do often... in my hobby-hacking... is try to get the CURRENT COMPLETE STYLE SETTINGS from some html element in some webpage somewhere. I work around "nodeserving"... where modified webservers (built-in content management systems) "transclude" html elements from webpages, and send out JUST the asked-for node/nodetree. What I need, is the CMS/webserver to be able to "peek" the 122 CSS-1 stylerule properties FOR said transcluded node... and mail THAT out WITH the node. But cascading stands in the way of this, I believe. The CMS/webserver must actually render-up the doc in a dom-tree of its own, then do the "gathering" of the css properties FOR that node, then package THAT, and send it along with the node itself. In brief... I could use... get namespace.node.markup and get namespace.node.stylerule and maybe... get namespace.node ...automatically returns both... even with metadata maybe. Does a 'stylerule query' such as this... retrieve the styles that apply to CHILDNODES under this top-node? yep. It has potential to return an xml doc that has tags/data for up to 122 css properties for EACH element within/beneath. Hoggy, eh? What I do in my time wasting, is haul around short articles written by colorful authors. They don't mind me borrowing the comments from their webpage/webserver... but they'd like me to present it to MY readers... with the same LOOK (style) as THEY have used in its initial presentation. So, I need webservers/content management systems that can deliver style goods as wanted... and if we don't want CMS's to grow dom trees each time a node is to be "translcuded with style"... then we have to say goodbye to cascading. *shrug* I'm more mouth than brain, as the folks on this list already know. :) Just thought I'd comment. <section removed> > Regards > Paul > Ya made my brain stir, Paul. Hope I didn't contaiminate the thread by subject wandering. I've done yappings about box models as well, talking about making borders a "secondary layer" and thus they don't affect the battle between margins and paddings in box models. I've talked of circular and u-shape-it box models, and the dreaded 8-url picture-border nightmare. Box models are odd, 1920's like creatures, eh? My group is into VRML and making simulated palm trees out of xml files, and navving via Huey helicopter. These mired-in-unix kids 'round here got their cars stuck in Quakerville errr sumthin'. :) Wingnut
Received on Wednesday, 29 June 2005 18:23:50 UTC