2010-05-01 [17:01:00.0000] Lachy: I don't think XBL is at all essential to styling or [17:01:01.0000] Lachy: pseudo-elements should cut it [17:02:00.0000] certainly for and also for many common uses of [17:02:01.0000] othermaciej, pseudo-elements can handle simple styling. But if you want more advanced styles, like what canvas or svg can provide, then XBL is better [17:03:00.0000] CSS with pseudos can basically handle colours, backgrounds, borders and sizes. I would consider much beyond that to be more advanced styles. [17:05:00.0000] I wonder how CSS transitions would interact with and psuedo-elements, if you could use them to specify how to transition between state changes. [17:08:00.0000] othermaciej: i replied to julian [17:26:00.0000] woot, i reached zero unread e-mails for the first time since monday! [17:26:01.0000] so, where was i wish timed tracks [17:27:00.0000] Lachy, progress and meter *pseudo* elements? [17:28:00.0000] theMadness: Pseudos within the progress/meter. [17:28:01.0000] So like progress::complete or something. [17:28:02.0000] Got it. [17:29:00.0000] theMadness, to be clear, pseudo-element selectors for use in CSS [17:29:01.0000] Somthing a la meter::stepping allowing us to determine how to transition between states might do it. [17:29:02.0000] Something a la meter::stepping allowing us to determine how to transition between states might actually prove quite useful. [17:32:00.0000] Actually I said meter but I was thinking progress. [17:32:01.0000] theMadness, would probably be done more like progress::complete { transition: width 1s linear; } [17:33:00.0000] I'm not arguing the syntax, I'll let you experts fight it out. :) But all in all it can (could) be very useful. [17:34:00.0000] And with a "width: calc( attr(value,number) / attr(max,number) * 100% )" rule or similar. [17:35:00.0000] tabatkins, I assume the width would be handled by the UA stylesheet [17:35:01.0000] or other mechanism. [17:35:02.0000] Well, yeah. But if it's in CSS, it would be with a rule like that in the UA stylesheet. [17:35:03.0000] Or else magic, yeah. [17:36:00.0000] Doesn't make much sense to allow authors to overriide that with CSS themselves, cause then they could use that to represent different states, rather than changing the element's attributes [17:36:01.0000] Yet we should be allowed to mess with it, isn't that the purpose of the whole thing? Like having a meter that fills using a background image positioning, or an opacity thing... [17:36:02.0000] They can do that with the XBL bindings to canvas. You won't really be stopping anything. [17:36:03.0000] Or font size... [17:37:00.0000] And nod, there are decent cases for using more than a changing width to represent completion. [17:38:00.0000] if it's being represented as a bar, as it would by default, then I can't see anything but the width being used. [17:38:01.0000] Vertical bar is the simplest counter-example. [17:39:00.0000] If it were altered to be a circular gauge, like a pressure gauge with a needle, then it would obviously not use width either. But that's something to be handled with SVG/canvas/XBL [17:39:01.0000] Pfft. You can totally do that with transforms. [17:40:00.0000] But it would lose all the semantic weight. [17:40:01.0000] meter and progress should probably inherit the same magic as for switching between horizontal and vertical bars [17:40:02.0000] theMadness: No, doing the rendering with canvas/svg/etc doesn't harm the meaning of the itself. [17:40:03.0000] Uhm, what would the syntax edn up being? [17:41:00.0000] *end [17:41:01.0000] theMadness, the semantics would be on the element in the HTML markup. XBL is a technology that allows you do bind a template to an element, like meter, which can then have the appearance of replacing it with something more customised [17:42:00.0000] With

title

…