Comments on: Web Devout was dugg http://www.webdevout.net/tidings/2006/08/11/web-devout-was-dugg/ Updates on the march of progress. A weblog about web design, standards, web browsers, and the overall health of the Web. Wed, 14 Dec 2011 14:45:24 +0000 http://wordpress.org/?v=2.8.4 hourly 1 By: David Hammond http://www.webdevout.net/tidings/2006/08/11/web-devout-was-dugg/comment-page-1/#comment-1135 David Hammond Tue, 15 Aug 2006 02:16:07 +0000 http://www.webdevout.net/tidings/2006/08/11/web-devout-was-dugg/#comment-1135 Well, you did express that the way "inherit" and invalid property failure issues are included in the tables "makes IE in particular look bad". But yes, I recognize that the Internet Explorer development team is now putting forth as much effort as can be expected from any development team of its size, and even my tables suggest that Internet Explorer is now making progress as fast as its competitors. If the practical benefits of the improvements are in fact underreported due to the way percentages are calculated (currently, features and bugs are not weighted based on real world importance), then it could be that IE7 has made somewhat more progress in CSS 2.1 support than its competition in the same length of time. To be honest, in terms of practical improvements, I do think IE7 made somewhat more progress over IE6 than Firefox 1.5 did over Firefox 1.0. Here's a brief comparison of the improvements: Firefox 1.5: <ul> <li>Counters (direct from no support to full support)</li> <li><code>cursor</code> URLs</li> <li>Outlines (again, from no support to full support)</li> <li>Proper support for the <code>quotes</code> property</li> <li>Various other improvements</li> </ul> Internet Explorer 7: <ul> <li>Fixed an <code>!important</code> bug</li> <li>Much improved selector support (although selector parsing is still full of bugs)</li> <li><code>background-attachment: fixed</code> on all elements</li> <li><code>transparent</code> border colors</li> <li>Fixed some float bugs, but added new ones</li> <li>Fixed some margin bugs</li> <li><code>position: fixed</code> support</li> <li>Improved <code>width</code>/<code>height</code> support (still have some bugs)</li> <li>Added <code>min-height</code>, <code>max-height</code>, <code>min-width</code>, <code>max-width</code> (still have some bugs)</li> <li>Varous other improvements</li> </ul> Probably the main reason IE7's improvements don't show a higher percentage increase than Firefox 1.5's, and arguably a true-to-reality reason, is that most of the improved features still have a lot of bugs in them, so some values are simply "less incomplete" where I can't objectively break down the feature support into a percentage, and the newly supported features often jump from "N" to about 50% rather than straight to "Y" as, for example, Firefox 1.5's outline and counter support did. In the real world, bugs in selectors, margins, and <code>width</code>/<code>height</code> have much more of an impact than <code>quotes</code> and outlines, and so it can definitely be argued that Internet Explorer 7 improved at least somewhat more than Firefox 1.5 or Opera 9 did. That said, I don't think the progress is fast enough to be considered significantly catching up to Firefox and Opera, and even so, it seems that progress would naturally slow down as support starts to near 100% (as it is in Firefox and Opera) when the remaining bugs begin to deal only with relatively picky details here and there. I squarely lay the blame for the current Internet Explorer problems on Microsoft's past actions, not your present actions. Internet Explorer is making progress, but it is currently so far behind that it seems IE won't likely catch up to the competition for at least several versions, if ever. Assuming Internet Explorer continues to have an influential market share for the foreseeable future, I can only hope that the situation continues to improve at a reasonable pace. As I have said before, I would love to see a ground-up rewrite of Trident focused around proper implementations of the standards. Well, you did express that the way “inherit” and invalid property failure issues are included in the tables “makes IE in particular look bad”. But yes, I recognize that the Internet Explorer development team is now putting forth as much effort as can be expected from any development team of its size, and even my tables suggest that Internet Explorer is now making progress as fast as its competitors. If the practical benefits of the improvements are in fact underreported due to the way percentages are calculated (currently, features and bugs are not weighted based on real world importance), then it could be that IE7 has made somewhat more progress in CSS 2.1 support than its competition in the same length of time.

To be honest, in terms of practical improvements, I do think IE7 made somewhat more progress over IE6 than Firefox 1.5 did over Firefox 1.0. Here’s a brief comparison of the improvements:

Firefox 1.5:

  • Counters (direct from no support to full support)
  • cursor URLs
  • Outlines (again, from no support to full support)
  • Proper support for the quotes property
  • Various other improvements

Internet Explorer 7:

  • Fixed an !important bug
  • Much improved selector support (although selector parsing is still full of bugs)
  • background-attachment: fixed on all elements
  • transparent border colors
  • Fixed some float bugs, but added new ones
  • Fixed some margin bugs
  • position: fixed support
  • Improved width/height support (still have some bugs)
  • Added min-height, max-height, min-width, max-width (still have some bugs)
  • Varous other improvements

Probably the main reason IE7’s improvements don’t show a higher percentage increase than Firefox 1.5’s, and arguably a true-to-reality reason, is that most of the improved features still have a lot of bugs in them, so some values are simply “less incomplete” where I can’t objectively break down the feature support into a percentage, and the newly supported features often jump from “N” to about 50% rather than straight to “Y” as, for example, Firefox 1.5’s outline and counter support did.

In the real world, bugs in selectors, margins, and width/height have much more of an impact than quotes and outlines, and so it can definitely be argued that Internet Explorer 7 improved at least somewhat more than Firefox 1.5 or Opera 9 did. That said, I don’t think the progress is fast enough to be considered significantly catching up to Firefox and Opera, and even so, it seems that progress would naturally slow down as support starts to near 100% (as it is in Firefox and Opera) when the remaining bugs begin to deal only with relatively picky details here and there.

I squarely lay the blame for the current Internet Explorer problems on Microsoft’s past actions, not your present actions. Internet Explorer is making progress, but it is currently so far behind that it seems IE won’t likely catch up to the competition for at least several versions, if ever. Assuming Internet Explorer continues to have an influential market share for the foreseeable future, I can only hope that the situation continues to improve at a reasonable pace. As I have said before, I would love to see a ground-up rewrite of Trident focused around proper implementations of the standards.

Posted using Mozilla Firefox 1.5.0.6 on Linux.

]]> By: Chris Wilson http://www.webdevout.net/tidings/2006/08/11/web-devout-was-dugg/comment-page-1/#comment-1134 Chris Wilson Mon, 14 Aug 2006 18:12:41 +0000 http://www.webdevout.net/tidings/2006/08/11/web-devout-was-dugg/#comment-1134 David, I want to be clear that I was not intending to criticize the structure of your tests; I was stating that I don't believe your test suite accurately portrays the improvement in standards support in IE7, and that was being picked up the community as proof that MS isn't interested in improving standards in IE (when nothing could be further than the truth). David, I want to be clear that I was not intending to criticize the structure of your tests; I was stating that I don’t believe your test suite accurately portrays the improvement in standards support in IE7, and that was being picked up the community as proof that MS isn’t interested in improving standards in IE (when nothing could be further than the truth).

Posted using Flock 0.7.4.1 on Windows.

]]>