- From: Boris Zbarsky <bzbarsky@MIT.EDU>
- Date: Wed, 20 Apr 2005 16:37:58 -0500
- To: Anne van Kesteren <fora@annevankesteren.nl>
- CC: www-style@w3.org
Anne van Kesteren wrote: > ... did I miss something or is this undefined? Should CSS2.1 define this? I think it's pretty clear that we want to avoid hysteresis in our rendering (that is, the rendering of a particular DOM with particular styles applied must be independent of the path taken to get to that DOM and those styles). This may indeed be worth stating as a basic principle of CSS, but I thought the formulation of the specification already made that clear: when the rules applied to the DOM change or the DOM changes, the CSS processing model should be run through again and new results computed, since CSS assumes a given set of rules that apply and a given DOM. Given that, I feel that anything other than just completely rebuilding the rendering of the page from scratch is just a UA optimization that should have the same effect as said full rebuild would have. -Boris
Received on Wednesday, 20 April 2005 21:48:42 UTC