- From: Shelby Moore <shelby@coolpage.com>
- Date: Thu, 02 Jan 2003 17:04:20 -0600
- To: John Lewis <lewi0371@mrs.umn.edu>
- Cc: www-style@w3.org
At 04:19 PM 1/2/2003 -0600, John Lewis wrote: [...] Most of your response was irrelevant to my conceptual point in this thread, only because I _AGREE_ with you. I agree with what you wrote. You just more exactly stated what I was trying to casually summarizing. Your disagreement with me was only in terms of being more exact, _AND_ than exactness actually enforces my point (below)... >> I am just saying basically that class selectors is a mechanism for >> non-semanticallly associating things with the markup. The >> ___selection___ by class does not consider the semantics of the >> elements of markup. > >That's true (if by "things" you mean "style"). The selection is >defined and happens in CSS, not HTML. Note that I _AGREE_ with you. This is exactly my point. My point is that the selection of style happens in CSS, not in HTML. In other words, the mechanism for style binding (non-semantic binding) happens in CSS layer, not in HTML layer. This is precisely my point!! CSS is not doing semantic binding. It is doing style binding, which is orthogonal to (does not depend on) semantics. Also (orthogonal point), you might be interested to note that style is not the _ONLY_ thing which can be non-semantically bound. There are endless possibilities. Reusing CSS selector logic (not necessarily CSS spec or syntax itself, just similar mechanism) one could orthogonally bind for example events, as I have suggested in previous thread: Using CSS for XML Events: http://lists.w3.org/Archives/Public/www-style/2002Dec/0077.html >Do you think it *should* consider the "semantics of the elements of >markup"? (I don't.) Absolutely not!! That has been precisely my point against XBL in this thread. See the subject of this thread! "CSS is wrong W3C layer for semantic behavior *markup*" I am saying that XBL provides the capability to bind (existing or new tags) semantics to new (changed or augmented) interpretations. Thus it also should have bi-directional line from XBL to "Semantics" layer in Ian's diagram of layers earlier in this thread. Once you agree on that bi-directional line (which was missing in Ian's diagram), then you realize that XBL is doing semantic binding. Yet it is also doing CSS and DOM layer features. Once you realize that merging layers in this way is bad (as you agree), then you realize that my point is correct. I have no problem with the CSS features of XBL. I have no problem with the DOM features of XBL. I have no problem with the semantic binding features of XBL. My problem is with combining all those into the same layer (same syntax and specification. Because then when we want to swap at that layer, we are dependent on swapping 3 layers non-independently (non-orthogonally). Watch for my next email to Ian in just a few minutes. It will explain this much better than I think I have in the past. -Shelby Moore > >-- >John
Received on Thursday, 2 January 2003 18:03:16 UTC