X-UA-8-Ball
Once again, with apologies to John Gruber…
Q: Do standards-based developers have to use X-UA-Compatible if they care about progressive enhancement for future browsers?
A: As I see it, yes.
Q: Doesn’t Internet Explorer already offer a fine-grained way to target specific versions?
A: Yes, definitely.
Q: So if there was a problem with the IE6 to IE7 transition, doesn’t that mean people weren’t properly using an existing tool which could have future-proofed their sites?
A: You may rely on it.
Q: Given that, is it likely that people will properly use a new tool that could future-proof their sites?
A: Very doubtful.
Q: Chris Wilson says that HTML 5 documents won’t need to use X-UA-Compatible. Since IE8 will be Microsoft’s first implementation of an unfinished specification, isn’t it likely that there will be bugs and that designers and developers will need to rely on them to get the intended rendering?
A: Most likely.
Q: So when IE9 is released, won’t Microsoft be back in the same situation of “breaking the Web”?
A: Signs point to yes.
Q: What will we be asked to do when IE9 comes out, then?
A: Cannot predict now.
Q: But if it goes down this course, IE will need to keep adding new version-targeting mechanisms with every release?
A: It is certain.
Q: So X-UA-Compatible won’t really solve the problem of forward compatibility for IE?
A: Outlook not so good.
Q: Has anybody actually sat down and thought this through logically?
A: Don’t count on it.
Q: Why are so many big names in the standards world so non-chalant about it, or even supporting it, then?
A: Ask again later.
Feed of latest entries
February 19, 2008
#
Apparently Microsoft uses a similar tactic to code IE.
February 20, 2008
#
I keep trying to write up a concise, meaningful thought on this whole X-UA-Compatible thing, but I haven’t been able to stop shaking my head in wonderment long enough to do it.
Wait, I think I’ve got it: WTF
February 20, 2008
#
it’s hard not to think of conspiracy theory things involving cash, really.
does zeldman have a new sports car?
February 20, 2008
#
You’ve done an excellent job breaking down some of the arguments used to promote this strange new system, but there’s one that I’m not quite getting.
You’re promoting conditional comments as an existing solution to future proofing a site (for IE). But as I understand it, X-UA-Compatible is an inversion of this: instead of adding code to specifically support old versions of IE, you can essentially do nothing and get IE7 forever (ignoring the practicality and likelihood of this being true in practice). It seems that they are specifically targeting the uneducated web devs by allowing them to remain gleefully ignorant, while trusting that web professionals will play along.
My feeling is that the WSP is going along with this, not because they think it’s an awesome idea, but because they feel that this is the only way Microsoft will continue to support standards in IE. For many of us web developers, though, we’d rather IE just dried up and went away.
I support the alternate idea of MS releasing standalone versions of specific browsers that can be run alongside each other (without hacks), and enabling a built-in compatibility detection system similar to how the pop-up blocker works (eg. detect known IE6&7 layout workaround and suggest a compatibility mode using the notification bar, then use that mode for every return visit unless the site is updated). MS needs to stop making their problems into our problems.
The worst part is X-UA-Compatible won’t save me any time as a developer. I’ll still have to code and test workarounds for IE6, IE7 and now IE8+. I wish they would just deprecate IE6 and be done with it.
February 20, 2008
#
To see the flaw in your argument, ask yourself which rendering mode an “uneducated” person’s page is most likely to trigger.
February 20, 2008
#
The way I read it, if you didn’t know anything about X-UA-Compatible (or it’s meta equivalent), your site would be “future proofed” against newer releases of IE because IE would always use the IE7 rendering engine for your page. Whereas with conditional comments, I would need to be aware of the differences between IE7 and IE8 in order to craft code that would render my page correctly in both browsers.
Maybe there is something I’m not seeing correctly? It seems to me that conditional comments require active knowledge and development on my part, where X-UA-Compatible doesn’t. Not that this makes X-UA-Compatible a good idea, but I can see where the argument, when applied to “don’t break the web”, might succeed.
February 20, 2008
#
I should say, “when I want to maintain the status-quo”. Obviously, adding X-UA-Compatible does require additional work and knowledge when I want to use whatever features IE8 might provide.
February 20, 2008
#
Someone who’s as “uneducated” as you’re making out isn’t likely to be triggering standards mode in IE, or to be relying on bugs in IE’s standards mode. Such a person is almost certainly working in quirks mode and blissfully unaware of it, and hence in need of no version-targeting mechanisms at all. Have a look at the breakdown of how
X-UA-Compatibleaffects different groups of developers in my earlier post on that subject.February 20, 2008
#
I think Microsoft should use the same technology they use in AntiSpyware, people simply flag sites as not rendering properly in IE8 and choose a rendering mode that the site should use, then this information is shared across all versions of IE8+ within the same zone (ie intranet or internet). I’m not sure how you’d trigger the mode to change when the site gets upgraded though, maybe you could ping the IE server with the UA-Compatible header to tell it to switch to the correct mode (the ping would have to originate from the site in question). Once a ping has been received from a site, the community voted mode is overridden by the official mode.
February 20, 2008
#
I see where you’re coming from. My argument was based on the assumption that many developers were triggering standards-compliance mode whether they intended to or not, possibly due to the tool they were using to generate content — and that it was bugs in standards mode that would require workarounds. I could be off base here, it’s been a while since I used quirks mode ;)
One of the original articles I remember reading (might have been on A List Apart, but it escapes me at the moment) originated the idea that Doctype switching was no longer practical because so many tools and templates triggered standards mode now.
BTW, I don’t use the term “uneducated” as an insult, more as a reference to the developers of the sites that the IE team seems to be trying to protect — those developers that would be unaware of X-UA-Compatible or other triggers and whose sites would break because they relied on the default rendering in IE6 or 7.
February 20, 2008
#
I thought this was a pretty interesting post from Meyer. When I read it, I thought about your articles and thought you might be interested in seeing the other side of the coin.
http://meyerweb.com/eric/thoughts/2008/01/24/almost-target/
February 20, 2008
#
Then their tools should update.
February 20, 2008
#
He did just buy an apartment in NYC, and that doesn’t come cheap. :-)
Seriously though, I don’t understand his position either. Through this whole thing it seems like he’s come a bit unraveled. I think he’s frustrated, but I don’t understand why he hasn’t explored (or talked about having explored) some of the solid alternatives to this whole debacle.