Jeff Zeitlin wrote: > On Mon, 8 Mar 2010 18:19:08 +1100, Alan Gresley <a...@css-class.com> > wrote: > >> You can still support IE6- but some authors just don't want to bother >> understanding why IE6- has shocking CSS2.1 support or why a document >> is rendered broken in IE6-. I agree with Andy Clarke by sending >> IE6/Win un-styled pages but their is also that user agent IE5/Mac. > > I'm not making myself clear. He is not serving IE6- an *unstyled* page, > he is serving it a DIFFERENTLY styled page, and a style that is just as > complex - but different. Different graphical accents. Black-and-white > photographs instead of color photographs. Differences like that. IMO, > the same kind of stupid penalization of IE that it suffered when > Netscape users would use the (unsupported-by-Netscape) FONT tag to set > the Symbol font (making the page unreadable in IE) when the first > version of IE came out. He wants to treat IE6 like the same kind of > pariah that Netscape 4 has been for years, with less excuse in the case > of IE6, because IE6 DOES support most of what he wants to do.
Ok, I see. He want IE6- to render a page like it antiquated or ugly. I totally agree but this has to be a top down banishment of IE6-. The governments and the corporations of this world have to inform the masses (by whatever medium) that there are 100s of thousands of web designers pleading for users to stop using IE6-. This does require large scale activism. I see a further dilemma with the fact that governments and corporations still serve pages in quirks mode which IE6 handles brilliantly. Please direct IE8 to this page. <http://css-class.com/test/bugs/ie/ie-chaos2.htm> >> I also disagree with that backwards philosophy. I attempt to suggest >> of philosophy of styling a page with future support in mind. My own >> test pages shows this philosophy. > > You are saying that it is a 'backwards philosophy' to design for a > consistent visitor experience? Are we seriously - and intentionally - > going back to the days of 'Best when viewed with <preferred-browser>', > but based on CSS support instead of support for HTML extensions? Are > you seriously suggesting that if I like a two-column print layout, I > sould write my print stylesheet for it, and to hell with any browser > that doesn't support it, letting it have whatever the defaults are? Yes, especially since all user agents now use the same default styles. Please again look at this page. <http://css-class.com/test/css/defaults/UA-style-sheet-defaults.htm> Note the last column with _Final Styles_. For authors to allow the user agent defaults to flow into the cascade of CSS, one person in particular (hint, list chaperon) has to stop promoting _CSS resets_. > Or > that I should have different print stylesheets for Trident vs. Gecko vs. > Webkit vs. whatever? That I should use generated content in the CSS > where it will make the page whizbang, even if it's not supported by all > browsers, and a failure to generate the content makes the page > meaningless in that browser? That's certainly what it sounds like > you're saying. And if it is, then I *cannot* support that OBSCENE > philosophy; I'll stick with 'graceful degradation'. There is no graceful degradation. You have large steps of degradation. BTW, my test page use one CSS file. This works perfectly well in all browsers accept IE6- but each version of each browser does show changes in level of support for CSS3. IE7 looks fine but I do have a bug to simulate what display:table achieves in other browsers. > Or maybe it's > called 'progressive enhancement'. But what it IS, is ACCEPTING that the > Web is NOT the printed page, that all browsers and computers and > displays are NOT created equal, and while not designing for an > inarguably badly broken (from the CSS support viewpoint) browser like > Netscape 4, at least making an effort to ensure that the results in > arguably usable browsers such as IE5.5 and later aren't unusable. I reject this. If you know how CSS works or which CSS bugs particular browser have, then you can send basically the same style to each page. Note, some HTML can not be styled and function correctly in IE7. Sometimes the HTML must be altered to not send IE7 buggy. >> I have been using CSS3 for over 2 years. IE9 will have support for >> rounded corner so nested divs could be seeing their last days. Here is >> a page of mine (style is basically from 2008) that uses CSS3 and give >> some examples. >> >> <http://css-class.com/test/css/> > > Yay for rounded corners. But note that I was talking about RADICAL > differences in the original post. The difference between rounded and > pointy corners is noticeable from an æsthetic point of view, but it is > not RADICAL, and I don't see how it justifies serving up a completely > different design for browsers that don't support them. Radical, mmm..... How about creating faux columns by background-size or source ordering by display:table. Or even using CSS like this. div+div+div { display:table; height:120px; width:81%; background:url(images/header3.jpg) 60% top; } div+div+div+div { clear:both; width:100%; } > BTW, thanks for the link to CSS3.info on that site; the selectors test > there is another useful and interesting resource (and reveals just how > badly Trident/IE7 is lagging in CSS support as compared with the > versions of Webkit and Gecko that ship with LunaScape). I more concerned with IE7 lack of correct support of the CSS2.1 that it does understand. To be somewhat fair, I should mention that IE7 was the first ever browser to support attribute selectors correctly (including those in CSS3). <http://css-class.com/test/css/selectors/attribute-match.htm> But empty substrings are still a problem. IE7+ and Opera shows them one way and FF and WebKit shows them incorrectly (or correctly according to the CSS WG). <http://css-class.com/test/css/selectors/att/attribute-substring-match-empty.htm> -- Alan http://css-class.com/ Armies Cannot Stop An Idea Whose Time Has Come. - Victor Hugo ______________________________________________________________________ css-discuss [cs...@lists.css-discuss.org] http://www.css-discuss.org/mailman/listinfo/css-d List wiki/FAQ -- http://css-discuss.incutio.com/ List policies -- http://css-discuss.org/policies.html Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
