Forum Moderators: not2easy
As of today, CSS3 allows us to do amazing things, no questions about that. But having IE6/7 not able to read it and representing about 35% of the internet users, what is today concensus about using CSS3 ?
I love CSS3 because I spend less time developping my styles and because I don't need to add useless classes to my html elements which makes it better, but what is the point if I have to maintain a second style sheet for IE and even hack my html output to make it compatible ?
As a concrete example, I'd like to use the new display:table feature on divs. Takes me a few seconds to set it up with CSS3, but then I still need to create html table + specific CSS style for IE. (this is mainly to fix the famous equal height divs with dynamically generated content).
So please, tell me what kind of strategy we need to start a new project today ? Stick to the old xhtml/css2 or jump to CSS3 with hacks for IE ?
Only 35% on IE6 and IE7 ? That's not what I see on my sites, I'm, still at 42% of visitors ...
display:table for getting equal hight boxes ... I think there are alternatives that you can feed to legacy browsers via a conditional comment. Not to make it look the nicest possible, but to keep it readable.
CSS3 has many more far nicer enhancement in store for us, and I vote for using it all. And let browsers not capable of handling it deteriorate somewhat gracefully (IE8 does virtually nothing that's new in CSS3 (just what IE always had already)
CSS3 will be held back for a long while if we let IE do that to us, as right now there isn;t a single version of IE that does go for it like all the other browsers are doing at an ever increasing rate.
Hacks: no thanks, conditional comments to work around the shortcomings and bugs to make it acceptable: sure; But making it look the same is just too hard, too much work.
IE and Opera are dead-weights as far as CSS3 is concerned right now. Granted Opera supports opacity and all the CSS3 selectors. They recently added support for downloadable fonts. The real meat with Opera is Presto 2.3 when Opera's native vector graphics engine dubbed Vega debuts. Then we should start seeing respectable CSS3 being deployed like border-radius and multiple-background-image support.
Internet Explorer 9 is subjective. They still have a considerable number of upper-level CSS 2.1 bugs that will no doubt get some attention. The IE team pretty much has indirectly hinted that border-radius will make IE 9. However any one who knows enough about overall clientside/browser related topics knows that the majority of work on IE 9 will be on overhauling the deplorable state of JavaScript stuck at version 1.5 which at best has proprietary implementation via JScript...so it's anyone's guess if standard JavaScript meets version 1.3 or 1.4 without proprietary implementations from JScript.
The Opera team has expressed difficulty with implementing CSS3 properties without a native vector graphics library built in to Opera...so what does that suggest about IE9? What is one without any doubt the biggest missing standard from IE8? SVG. XHTML support is likely as well as the "groundwork" of other supportive technologies has been slowly being introduced. Serve an XHTML 1.1 page as text/xml and you'll get the same XML error reporting as you do in Gecko when the same page is served as application/xhtml+xml.
So IE9 should be pretty exciting all around. I frankly don't expect any major CSS3 additions to IE9, maybe a couple other properties though we'll likely see the CSS3 selectors added. They'll likely implement their own SVG/vector graphics engine first. IE10 will likely see the major push for CSS3 that we all want...and yeah no one is looking forward to not only the wait for that to come along but for all the older versions of IE to evaporate even years after Microsoft jumps on the standards bandwagon.
As far as IE today...conditional comments. Design for standards and fall back to a style sheet explicitly used to correct the rendering of IE.
- John
What Microsoft really needs to start doing is to _FIX_ IE8, (actually: fix the older ones too) and not make yet another new version that their customers fear and don't want to upgrade to and block progress for many years to come.
We do not need IE9, IE10, we need all the functionality in IE8, and we need it "yesterday", not in a decade or two.
