- From: George Lund <george@lundbooks.co.uk>
- Date: Thu, 19 Sep 2002 11:08:51 +0100
- To: www-style@w3.org
Bert Bos <bert@w3.org> writes >I have a proposal... I have read through fairly carefully, and will read through again. But my initial comment would be that I was looking for something simple for authors to use, and I don't think this is quite it! (The JavaScript would seem more simple.) If this kind of thing could be going on in the background, with a way of triggering it in a style sheet, that would be better. (I think there may be some general principles here: style sheets shouldn't be expected to simulate an entire GUI environment via the mechanisms available *within* them, they should be able to *call upon* elements that are considered common to all GUIs. How individual browsers implement style sheet suggestions shouldn't be a matter for the W3C recommendations, even though in the case of Mozilla it may be that they really are implemented by applying some internal mechanisms that look like CSS. Internally, browsers like Mozilla should, ISTM, be using some kind of replacement mechanism, so that 'simple-looking' CSS can be 'deconstructed' into a practical layout rule involving, say, positioning and, perhaps, 'state' (as per your suggestion). But other browsers should be free to implement things using native widgets, as is done on many browsers with HTML form controls. In this way, there is both a simple language for authors to deal with, and the possibility that the user's browsing experience will be made easier by commonality with the GUI they are using.) I realise there are some difficulties with my original proposal on its own as to what is to be displayed when the element is collapsed. Perhaps the idea would have been better tackled with a new HTML attribute; that could more easily have fitted with the way user-agent CSS works right now. A few small points to do with your examples: implementations I've seen of the collapsible list usually require either a double-click, or a single click on a plus symbol. (He said desperately trying to speak in a non-Windows-centric way!). And I would want the usual method of implementation to be that the first-child of the element was the thing that needed to be double-clicked, so that all the important text was there as content in the document, not lost in attribute values or - worse - in a style sheet. thanks for the response -- George
Received on Thursday, 19 September 2002 06:09:54 UTC