3 ms·

CSS is bad and hindsight is not required. Consider all the obvious page layout things that everyone wants, such as centering, columnar text, non-moving navigati

CSS is bad and hindsight is not required. Consider all the obvious page layout things that everyone wants, such as centering, columnar text, non-moving navigation elements, footers, and so forth that are either unsupported or require rocket science tricks to get half-working. And that's without getting into the inobvious requirements that they've had at least ten years to grapple with. To be fair, some of CSS's shortcomings follow from HTML design errors that are equally egregious (and in some cases have been magnified by HTML5). E.g. It ahould have been possible to create decent looking navigation elements in HTML with no CSS or Javascript, but still isn't.

matthewmacleod 13y ago
Consider all the obvious page layout things that everyone wants, such as centering, columnar text, non-moving navigation elements, footers, and so forth that are either unsupported or require rocket science tricks to get half-working. These features are all now available in the latest CSS specs. Yes, it's taken a while to get there, but I absolutely dispute that these features are "obvious." To be fair, some of CSS's shortcomings follow from HTML design errors that are equally egregious (and in some cases have been magnified by HTML5) I don't in most cases agree that problems in HTML are "design errors," so much as "features it turned out we wanted that we didn't think we did." E.g. It ahould have been possible to create decent looking navigation elements in HTML with no CSS or Javascript, but still isn't. Again, I don't agree that this is something HTML should be involved with.
- Tloewald 13y ago
  All the layout features used in publishing are obvious requirements for a document layout technology, especially when folks from Adobe, et al, are on the relevant committees. All the layout features used in applications are obvious requirements for web applications. It's perfectly reasonable to leave out a lot of stuff from early versions of a product or standard, but we're a long, long way past that. > I don't agree that this is something HTML should be involved with. HTML is involved with it, it's just badly involved. Either don't provide presentation controls or do it properly. Your opinion doesn't matter because there are facts on the ground.
  - matthewmacleod 13y ago
    All the layout features used in publishing are obvious requirements for a document layout technology, especially when folks from Adobe, et al, are on the relevant committees. All the layout features used in applications are obvious requirements for web applications. It's perfectly reasonable to leave out a lot of stuff from early versions of a product or standard, but we're a long, long way past that. Okay, but CSS was a document layout spec for web documents, and has become to be used as a layout spec for interfaces. There are differences there - in other words, CSS is pretty unique in terms of the kinds of areas it has been trying to tackle. Say columnar text - I don't at all think that many people would have considered that it would be required back in the early days of the web. It turned out people did need it, and now we mostly have it. So no, I don't think many of the things you talk about were obvious, and in any case we've mostly sorted the situation - it's taking a long time, but CSS is also special in that there are lots of implementations of it, and many of them are legacy. So don't expect overnight change. HTML is involved with it, it's just badly involved. Either don't provide presentation controls or do it properly. HTML was historically involved, because we didn't realise how important separation of structure and presentation would actually turn out to be. Now we're got a big legacy to deal with. To be honest, I'd have preferred XHTML2 won, but I don't really think it's a massive deal either way - you're arguing for perfection, which exists in literally no software platform. Your opinion doesn't matter because there are facts on the ground. I'll leave you to think about the irony of that statement, and just point out that facts are open to interpretation.
    - Tloewald 13y ago
      A bald opinion backed by nothing does not contribute to the discussion. If you gave some kind of reason for disagreeing -- an interpretation if you will -- that would be something to discuss.

matthewmacleod 13y ago

Consider all the obvious page layout things that everyone wants, such as centering, columnar text, non-moving navigation elements, footers, and so forth that are either unsupported or require rocket science tricks to get half-working. These features are all now available in the latest CSS specs. Yes, it's taken a while to get there, but I absolutely dispute that these features are "obvious." To be fair, some of CSS's shortcomings follow from HTML design errors that are equally egregious (and in some cases have been magnified by HTML5) I don't in most cases agree that problems in HTML are "design errors," so much as "features it turned out we wanted that we didn't think we did." E.g. It ahould have been possible to create decent looking navigation elements in HTML with no CSS or Javascript, but still isn't. Again, I don't agree that this is something HTML should be involved with.

Tloewald 13y ago
All the layout features used in publishing are obvious requirements for a document layout technology, especially when folks from Adobe, et al, are on the relevant committees. All the layout features used in applications are obvious requirements for web applications. It's perfectly reasonable to leave out a lot of stuff from early versions of a product or standard, but we're a long, long way past that. > I don't agree that this is something HTML should be involved with. HTML is involved with it, it's just badly involved. Either don't provide presentation controls or do it properly. Your opinion doesn't matter because there are facts on the ground.
- matthewmacleod 13y ago
  All the layout features used in publishing are obvious requirements for a document layout technology, especially when folks from Adobe, et al, are on the relevant committees. All the layout features used in applications are obvious requirements for web applications. It's perfectly reasonable to leave out a lot of stuff from early versions of a product or standard, but we're a long, long way past that. Okay, but CSS was a document layout spec for web documents, and has become to be used as a layout spec for interfaces. There are differences there - in other words, CSS is pretty unique in terms of the kinds of areas it has been trying to tackle. Say columnar text - I don't at all think that many people would have considered that it would be required back in the early days of the web. It turned out people did need it, and now we mostly have it. So no, I don't think many of the things you talk about were obvious, and in any case we've mostly sorted the situation - it's taking a long time, but CSS is also special in that there are lots of implementations of it, and many of them are legacy. So don't expect overnight change. HTML is involved with it, it's just badly involved. Either don't provide presentation controls or do it properly. HTML was historically involved, because we didn't realise how important separation of structure and presentation would actually turn out to be. Now we're got a big legacy to deal with. To be honest, I'd have preferred XHTML2 won, but I don't really think it's a massive deal either way - you're arguing for perfection, which exists in literally no software platform. Your opinion doesn't matter because there are facts on the ground. I'll leave you to think about the irony of that statement, and just point out that facts are open to interpretation.
  - Tloewald 13y ago
    A bald opinion backed by nothing does not contribute to the discussion. If you gave some kind of reason for disagreeing -- an interpretation if you will -- that would be something to discuss.