Although getting this implemented would be a nightmare, I would think that XSLT's approach to variables (declarative rather than imperative - you can't set a variable's value twice) would be feasible in CSS.
But then again, one could argue that you should just use XSLT for your layout logic in the first place.:P
Please don't do this. Please please please, for the love of god, don't do this. Why? The content inside the expression() tag is evaluate upon every (yes, every) window event that could possibly alter the value of the content. Load? Evaluate. Resize? Re-evaluate. Click? Re-evaluate. Mouseover? Re-evaluate.
I've seen a single window, consisting of nothing but a dynamically sized popup menu using an expression() on each element, balloon from 100MB usage to over 400MB memory usage (and lock up the system), simply from grabbing a border of the window and shaking it around a few times.
Expressions and behaviors in CSS are bad. If the CSS box model doesn't work for you, or if you want the sizes tied to an event, then tie them to the damn event already. Using Javascript.
Do you actually read these articles, or do you just scan through them to find the part where you can lash out against Microsoft? A lot of what you're saying is true... as of 1998, but then again, they were working on creating the features that are used everywhere now. Regardless, I take it you've never touched XSLT or MSXML. If you had, you would have realized that their XML processor is one of the fastest and most robust for XSLT 1.0 out there today... because they were making it against a standard that made 100% complete logical sense, and they didn't have to support tens of thousands of corporate-level developers' broken code. But that is the case with IE and HTML, so you're better off bitching at the developers who write the broken code in the first place.
Because the legitimate owner will only see it as a minor irritation and will continue on happily afterward, whereas the pirate will be stuck working around it, finding a key, finding a patch, and so on... and will be thinking the entire time that this should cost $300 or such. Installs a sense of guilt, I suppose - if they come around to thinking "OK, I like this OS", the very next thought in their mind will be "I should pay for it already."
Not surprising that they're fully aware of it, surprising that they're willing to say so in public.
Slashdot Mirror
But on a serious note, the W3C needs to get the (numerous) loose ends of XHTML2 wrapped up so the web can move on.
Although getting this implemented would be a nightmare, I would think that XSLT's approach to variables (declarative rather than imperative - you can't set a variable's value twice) would be feasible in CSS. But then again, one could argue that you should just use XSLT for your layout logic in the first place. :P
Please don't do this. Please please please, for the love of god, don't do this. Why? The content inside the expression() tag is evaluate upon every (yes, every) window event that could possibly alter the value of the content. Load? Evaluate. Resize? Re-evaluate. Click? Re-evaluate. Mouseover? Re-evaluate. I've seen a single window, consisting of nothing but a dynamically sized popup menu using an expression() on each element, balloon from 100MB usage to over 400MB memory usage (and lock up the system), simply from grabbing a border of the window and shaking it around a few times. Expressions and behaviors in CSS are bad. If the CSS box model doesn't work for you, or if you want the sizes tied to an event, then tie them to the damn event already. Using Javascript.
Do you actually read these articles, or do you just scan through them to find the part where you can lash out against Microsoft? A lot of what you're saying is true... as of 1998, but then again, they were working on creating the features that are used everywhere now. Regardless, I take it you've never touched XSLT or MSXML. If you had, you would have realized that their XML processor is one of the fastest and most robust for XSLT 1.0 out there today... because they were making it against a standard that made 100% complete logical sense, and they didn't have to support tens of thousands of corporate-level developers' broken code. But that is the case with IE and HTML, so you're better off bitching at the developers who write the broken code in the first place.
Because the legitimate owner will only see it as a minor irritation and will continue on happily afterward, whereas the pirate will be stuck working around it, finding a key, finding a patch, and so on... and will be thinking the entire time that this should cost $300 or such. Installs a sense of guilt, I suppose - if they come around to thinking "OK, I like this OS", the very next thought in their mind will be "I should pay for it already." Not surprising that they're fully aware of it, surprising that they're willing to say so in public.