I recently prototyped a site using CSS that only Safari 3+ and FireFox 3+ could understand. I never loaded IE up once while doing it. Boy was it a pleasure being able to have a CSS stylesheet declare rounded borders with one declaration. The problem with supporting IE at all is that so few (almost none) of the CSS functionality that eliminates hacks we've been working with are available. Even IE8 isn't enough. So choosing to support IE at all means intentionally refusing to adopt features supported by Safari and FireFox. That is, unless you want an inconsistent experience for your users.
But why do we have to provide a perfectly consistent experience between desktop browsers?I'm starting to think that this is precisely what we should do to force MS to get with the times. Let your site degrade for IE in ways that do not break core functionality or fundamental design requirements. When users start figuring out that they can't see a prettier version of NYTimes because IE doesn't support @font-face, rounded corners, CSS transformations, and other advanced CSS functionality, many will decide to switch.
The difference between this approach now and how it was implemented in the 90s during the Netscape/IE wars, is that it's a best practice to degrade gracefully but retain important functionality now. The criticism of this strategy back then is that many sites simply did not work at all in one browser or the other. That would be unacceptable.
"That is, unless you want an inconsistent experience for your users."
I'm not so sure why people fight so hard for an IDENTICAL experience between browsers. People running IE6 probably won't even notice your square corners. This is why I keep proposing serving a mobile version to IE6 users. The next time something is off by one row of pixels in IE6, completely ignore it. Chances are, IE6 users will too.
It's not like the IE6 people are also going to be running Firefox to do a side-by-side comparison of how your page renders. Unless they have some new and fascinating psychological abnormality, of course.
Fonts are a really bad example for the (otherwise valid) "fuck IE" mindset -- Internet Explorer was the first to support downloadable fonts, by a margin of at least 5 years!
The only problem is that you have to convert the font into an arbitrarily different format, ans Adobe and the foundries would have screamed bloody murder (as they are doing now with @font-face) if you could just use plain TrueType and OpenType fonts.
Webkit has supported @font-face for quite a while, Mozilla is the real straggler here -- they repeatedly refused to implement it for a decade, and when they finally did they added a bullshit same-origin policy. At least they didn't introduce a second stupid webfont format!
Yeah but MSIE decided to go with EOT which is arguably why no-one else bothered. If they'd used TTF then other browsers would have implemented it earlier because it would actually be a usable solution.
I don't think the format was ever really the problem; people are willing to put up with all sorts of conversion shite to get things up on the web (FLV, for example.) IE's problem was entirely that it was ahead-of-its-time in this regard: fonts were too big on yesterday's internet to download at a decent clip for instant display, and there weren't any open fonts that anyone was interested in using.
It's generally not possible at a business-level to treat IE users as second-class citizens. So while your idea appeals to anyone who has ever done webapp development, it's not conducive to business interests.
After all, there's always another website looking to take the users that you're not willing to fully support.
But it's not about refusing to support them completely. Also, the difference between trying this strategy today and even 2 years ago is that CSS now has features that would be compelling to end users rather than just developers. Users couldn't care less if developers want to use Safari/FireFox/IE8 because they support display: table. But users are becoming more demanding in terms of user experience and as they realize that the alternative browsers let them have nicer typography and cool desktop-like animations/transitions they will begin to demand this kind of thing.
For a while, some die-hards will probably try to use Flash to mimic the experience that HTML5 and these new CSS features offer. But I think the uptake of Flash amongst new developers is less than that of standards-based practices. Simply by virtue of having more developers supporting the standards based approach, I think Flash will eventually be eclipsed until it is relegated to a niche, kind of like how Google Maps is everywhere now and MapQuest is dying.
There is this big assumption that users are resistant to change. I'm inclined to believe they're resistant to unnecessary change. We just need to make them feel compelled to change and users will do so. For the first time in a while, Safari and FireFox actually have compelling features to users that IE does not. That changes the landscape quite a bit.
So, when you argue with your client over this specification for hours, and they decide to give the job to someone who will make it consistent, then what?
From a business point of view, especially in the marketing world, you'll have a hard time compelling the every day firm to ditch IE support. And as I said in the previous thread about this same exact thing, the elites can do it and still get plenty of business, but is everyone in the position to be able to turn down a job because of this?
The answer is probably no.
EDIT:
I guess the real issue is more along the lines of an education issue, and people who don't grasp it, getting put in charge of whatever new micro-marketing site you're releasing this quarter.
Don't argue. Present your case as "You can have IE6 support or you can have features X, Y, and Z. Make the call." Sometimes they'll say they want IE6, and sometimes they'll want the features. But it should be framed as a tradeoff: you can't have both, and supporting obsolete browsers means you can't deliver functionality that may get you more users in the long run.
> But users are becoming more demanding in terms of user experience and as they realize that the alternative browsers let them have nicer typography and cool desktop-like animations/transitions they will begin to demand this kind of thing.
While many users may like these things, "demand" seems a bit strong.
And then there are the folks who don't like "cool desktop-line animations/transitions".
Google Search already does this in a few places where the cost of supporting IE was prohibitive, eg. if you do Search Options -> Related Searches and scroll down, the search suggestions stick at the top of the page on every browser except IE6, with the results scrolling underneath them.
But why do we have to provide a perfectly consistent experience between desktop browsers?I'm starting to think that this is precisely what we should do to force MS to get with the times. Let your site degrade for IE in ways that do not break core functionality or fundamental design requirements. When users start figuring out that they can't see a prettier version of NYTimes because IE doesn't support @font-face, rounded corners, CSS transformations, and other advanced CSS functionality, many will decide to switch.
The difference between this approach now and how it was implemented in the 90s during the Netscape/IE wars, is that it's a best practice to degrade gracefully but retain important functionality now. The criticism of this strategy back then is that many sites simply did not work at all in one browser or the other. That would be unacceptable.