Working together for standards The Web Standards Project

On 16 February, Web Standards Project Members Faruk Ateş, Porter Glendinning, and I got together with Chris Wilson, Platform Architect for Internet Explorer to talk about IE8′s proposed default version targeting behavior of having to opt-in to the browser’s new standards mode.

As you may recall, the version targeting opt-in requires developers to use the new X-UA-Compatible HTTP header (or the meta equivalent) in order to take advantage of the standards-based layout and scripting improvements in the IE8. Under the current setup, any page/server not opting-in to would continue to render and behave as though it were being viewed in IE7, even if the site was being viewed on IE8, IE9, or IE1000. This doesn’t sit well with many standards-aware developers who think that the IE team got it backwards; many of them believe that you should have to opt-in to keep your site from being interpreted by newer layout and scripting engines. (Folks interested in hearing both sides of that argument should check out the latest issue of A List Apart, where Jeremy Keith and Jeffrey Zeldman square off on the topic.)

In our “round table” discussion, we talked about several proposals from the community that could help bring IE8′s standards mode to the masses, including:

• encouraging Microsoft to ship a patch to IIS that automatically targets sites run on that server to IE7 (in hopes of avoiding the potentially catastrophic impact IE8 may have on intranets and the like);
• shipping the IE8 beta with standards mode on by default just to see how many sites break; and
• making IE8 a standalone browser, capable of being run side-by-side with IE7 to allow users the flexibility of using one broswer for some sites and the other for other sites.

If you are interested in listening to or reading a transcript of the discussion, you can do so by following one of these links:

• WaSP Round Table: IE8′s Default Version Targeting Behavior [mp3]
• WaSP Round Table: IE8′s Default Version Targeting Behavior [transcript]

I’d like to give special thanks to fellow WaSPs Glenda Sims and Steph Troeth for organizing, producing, and transcribing this Round Table.

This buzz has been translated into Polish.

Your Replies

#1 On February 24th, 2008 3:27 pm Jonathan Kahn replied:

Excellent discussion, and amazingly a possible solution seems to have come out of it — strict doctypes (as opposed to transitional) triggering IE8 standards mode, without requiring meta tags.

That would be good enough for most standardistas, wouldn’t it?

#2 On February 24th, 2008 4:29 pm Mark Thompson replied:

I’m relieved to hear this awful metatag idea isn’t set in stone. In my view strict doctypes are the way to go, anybody serious about web standards knows the difference between transitional and strict doctypes. There have also been enough tutorials and education-focused articles on ala etc. for most developers to build to the standards.

For me, inserting a metatag in order to have new features is the wrong way around. Those with legacy content should be requesting historical rendering, leaving valid content to be rendered against whatever doctype is defined.

How about running a poll to see what we web developers feel would be the best way forward? I think a LOT of web professionals would jump at the opportunity to voice their opinion on such an important decision.

Historical issues aside, this is about the future.

#3 On February 24th, 2008 5:10 pm Devon Young replied:

I really like the idea of shipping IE8 in standards mode to see how many sites will break. After all, they might be jumping the gun and getting ahead of themselves. Likely, a lot of sites will break. But hey, they broke in IE7 beta too. By the time is came out, many developers fixed things. I realize MS doesn’t want to “break the web” or force people into fixing their code…but, why doesn’t Microsoft use its 800lb gorilla-ness to “bully” people into fixing their code the better of the internet? They do it when they want to get something done for themselves all the time. It would be great PR for them…all the anti-MS folks who develop, would see them finally throwing their weight around for a good purpose.

#4 On February 24th, 2008 5:13 pm David Naylor replied:

Doh!! I thought CW was referring to strict doctypes when he said that X number of the top 200 sites trigger standards mode…

Well… if that isn’t the case, go for the strict doctypes man!

#5 On February 24th, 2008 9:41 pm Scott replied:

I agree with Mark that if there’s an opt-in, it should be for opting in to old versions of rendering. Make it an easy way out for developers that aren’t interested in fixing their code, or want more time to fix their code. Standards-aware developers should not be penalized.

If strict vs. transitional could replace the meta switch, I would be in favor of that. I suspect it unlikely that anybody is using a strict doctype that doesn’t know what they are doing.

Other browser developers aren’t getting hung up on compatibilty issues, at least that I’ve seen. Breaking the web forces people to clean up their act and their code. As Devon said, it’s good PR for Microsoft, which is often accused of holding the web back because of it’s “outdatedness”.

#6 On February 24th, 2008 11:13 pm Dana Lee Ling replied:

I tend to code using CSS3 and HTML5 counting on “graceful degradation” in older browsers. I have always relied on new versions of browsers gaining new rendering capabilities. The IE8 proposal would stand my logic on its head – newer IE browsers would never provide better rendering because my lack of the appropriate meta tag would cause IE1000 to render using IE7 capabilities. I’m not about to go back to standards compliant pages and add that IE specific meta-tag either – I have hundreds of “hand” coded files on my manually maintained web site.

#7 On February 25th, 2008 9:21 am Shawn Lauriat replied:

So, did they actually answer the “can we get a standalone IE beta” question asked more than once? I missed it if they did, and testing against my current web applications would give me a lot more than anything else at this point, and I’ve grown very, very tired of running multiple VM images just in order to test IE.

For the record, I second Mark Thompson’s entire comment. Though it does pose the problem of getting IE to display the fixed rendering in existing applications that would benefit from the new behaviors, but can’t switch doctypes at the drop of a hat.

#8 On February 25th, 2008 10:23 am Shelley replied:

What on earth did you people hear that made you think that Microsoft is not going ahead with the meta tag, just as it’s been described?

I’m also very disappointed that one person didn’t ask in this discussion, “Is IE8 going to support the XHTML MIME type?”

In fact, the whole thing sounded like an infomercial.

#9 On February 25th, 2008 7:23 pm WaSP Member agustafson replied:

I’m also very disappointed that one person didn’t ask in this discussion, “Is IE8 going to support the XHTML MIME type?”

I think that’s an excellent question, and one we can bring up in our next Round Table with Microsoft (tentatively scheduled for some time around SXSW). This Round Table was very focused on the default behavior and ways we could ease the lives of standards-aware developers vis a vis version targeting.

Taking it a step farther, I think we should also ask if the XHTML mime-type could trigger IE8′s standards mode… I will make sure I do so either in that call or via email directly to Chris.

#10 On February 26th, 2008 4:58 am Robin replied:

I think we should also ask if the XHTML mime-type could trigger IE8’s standards mode

Isn’t that a given? Why are we even having to question that?

#11 On February 26th, 2008 12:15 pm Dave H replied:

If someone is going to encourage Microsoft to ship IIS with X-UA-Compatible HTTP headers already configured, possibly they could ask them to add a few file extension/MIME type mappings while they at it? (.xhtml, .xht, .svg, .svgz, .json… :-)

#12 On February 26th, 2008 12:54 pm John Arthur replied:

I just want to know how long we’ll have to deal with version-targeting…

If DOCTYPEs such as HTML5, which aren’t widely used today, will NEVER trigger version-tracking, MS has engineered a fantastic solution, and I’d be happy if current DOCTYPEs trigger IE7 by default. If they actually think this will work forever without more problems than solutions, I’m going to take up my pitchfork and torch.

Give us an exit plan for version-tracking (and please spell it out), and I’m all yours (more or less). Of course, you’d need to stick to it, but I’m sure that’s not too much.

Anyone with me?

JA

#13 On February 26th, 2008 9:21 pm Dominic Shiells replied:

What do they mean by breaking the web,
What pages will be breaking?, early frontpage and dreamweaver pages.
We need to almost do a study on the web, what pages will actually break.
Couldnt the Browser detect whether it is a non standard web page and then go to quirks mode.
If it ships in standards mode, the difference is layout engines, also what size would ie8 be because if it has to include Trident which is the old layout engine and then the New Layout Engine it would be extremly slow.
If both are running at once.

What we also have to realise that older webpages always change, so it may not break the web, how many website will be broken. Statistics!!!
Microsoft have a powerful marketing scheme could not they advertise about changing to standards mode. So that the release of ie8 people would be prepared so that they can change their sites.

#14 On February 27th, 2008 12:01 am Shelley replied:

I think we should also ask if the XHTML mime-type could trigger IE8’s standards mode

Isn’t that a given? Why are we even having to question that?

Not a given. We’ve asked Microsoft to confirm whether it will finally support application/xhtml+xml. And we asked if IE8 will then use standards mode when finding this mime-type.

No answer.

You would this would the simplest answer is the world to make, wouldn’t you?

#15 On February 27th, 2008 4:30 am Robin replied:

Well, yes, you would. But it’s been a long time since the IE team were open about upcoming features :(

#16 On February 28th, 2008 2:49 am Nathan replied:

I seriously hope this is still a table issue and they haven’t made the final decision yet, we all know there are better ways to deal with this issue, lets just hope that our opinions can alter the outcome to suit both parties, because if it only suits us we may as well give up now…

I’m all for XHTML mime-type triggering standards mode, I’m sure 99.9% of websites that would break under “full standards compliance” in IE8 will either have no DOCTYPE or they will have a HTML 4.0 loose DTD. This surly leaves the only broken sites to be those that were defined as XHTML/Strict because the author had no idea what they were doing, hence not our fault and not Microsoft’s fault.

Besides that, do MS truly believe that everyone using IE^ will instantly move to IE8 as soon as its released? Ther will be plenty of time for people to fix these issues if the Strict DOCTYPE switch doesn’t already do the job for them.

#17 On February 28th, 2008 12:59 pm Aleksey V lazar replied:

Every time I validate an XHTML 1.1 document using the W3C a warning about the document “being served with the text/html Mime Type which is not a registered media type for the XHTML 1.1 Document Type. The recommended media type for this document is: application/xhtml+xml”. I think that using application/xhtml+xml mime type as a trigger for standards mode is the best (and obvious) way to go. This will not force any unexpected behaviour for users or developers.

#18 On February 29th, 2008 9:51 am Alan Gresley replied:

@Aleksey V lazar

I’m already using application/xhtml+xml in my header on all my pages. This is only required for off-line testing anyway. The server itself determines how the page is served. I currently have the XML prolog proceeding the DOCTYPE so when IE supports true XML then my pages coded with XHTML will just need a change to the file extension. Instead of htm or html I will use xml or xhtml instead. I don’t have to change the source code, just the file extension.

So yes, I am in full agreement with you, pages with

content=”application/xhtml+xml; charset=utf-8″

should by default render in IE8 standard mode.

#19 On February 29th, 2008 10:36 pm Gérard Talbot replied:

I agree that strict doctypes should trigger IE8 standards mode, without requiring a meta tag.
Opting-in should be for IE 7 “standards” mode; default rendering mode should be super-standards IE8 mode.

yes, a layout breaks; it can be detrimental, in most cases, not so horrible. It may not look quite right, or something may be out of alignment, or something like that.

Aaron G.

By itself, the IE bugs do not indicate how and by how much a layout will break in a given webpage. If there is a IE bug with border and my bordered box is 1px wide, I may not even notice the layout breakage or the bug. But the problem is that sometimes/often, people do all kinds of weird things with borders (like a faux-column) and then their CSS columnar webpage template breaks because of such bug.

Another issue. Many sites are over-constraining, over-coding, over-declaring, over-defining (with their CSS code and markup code), overly rigid and very much pixel-precision focused, oriented. So, then, a small layout breakage caused by a bug or bug fix brings shouts, cries, howls, unfair finger-pointing to IE 8 for breaking up a page.

Sometimes, the breakage may be just a few pixels but it nevertheless ruins the whole design. I can show you such cases where most web designers would go mad. Just a few pixels off can break a drop-down menu list, a CSS columnar webpage template, etc.

What is still unacceptable is the number of very damaging bugs that cause important, severe, grave problems in IE 7:

- Application hangs exists in IE 7
- Crash bugs exists in IE 7
- Peekaboo types of bugs still exists in IE 7
- Guillotine types of bugs still exists in IE 7
- Various content disappear bugs

and I know I’m not exaggerating or over-reacting here. What would be illogical/crazy would be to not fix those severe, incapacitating bugs in IE7 standards mode. It would be irresponsible to do that in the name of backward-compatibility or because of a absolute, blind commitment/prime directive to “not break the web”. I can document and substantiate my claims here.

The discussion is still disappointing in some aspects. Mid-term and long-term planning from Microsoft. How is Microsoft going to avoid having another new browser version release crisis later… with IE 8.5 or IE 9? regarding the same problem we have today? We have no idea. The discussion may be happening again for IE 9 (or IE 8.5) because ordinary/amateur web authors can not upgrade their web design skills, don’t read or write abandonware websites…

Web standards advocacy groups and web standards gurus – in articles, in tutorials, in web authoring forum discussion newsgroups – have always claim and recommended to learn, to read, to follow best coding practices. None of this – no plan of tech evangelization whatsoever – actually seems to be in any part of Microsoft’s agenda. You see at MSDN articles promoting user-agent exact string detection, not object/method support detection: a blatant-obvious error in my opinion. You see at MSDN invalid markup code, invalid CSS code, deprecated practices, wrong, incorrect practices almost everywhere, in articles and in code examples and code excerpts. And I’m not even mentioning IE dev. people’s own blogs or IE blog.

It’s one thing to say that ordinary/amateur web authors do not or can not upgrade their web design skills, don’t read or write abandonware websites. It’s also completely true to say that MSDN articles are abandon-blah-blah, are spectacular examples of invalid markup code, blatant demonstration of web standards incoherence, blatant examples of web standards inconsequence.

Once you fix the browser and once users upgrade the browser (or browser version) they use, then, the remaining thing to do is to upgrade the webpage code. Web authors are the weakest link in the web standards circle principle: they were 3 years ago, they are today and they will be 3 3 years from now.

Gérard

#20 On March 3rd, 2008 8:14 am Jens Grochtdreis replied:

Well, it is hard for me to take WaSP seriously in the fight for better webstandards-support. It seems, you are more and more flirting with Microsoft rather than discussing the best way to improve their products. That’s a shame. I can say for myself, that the German Webkrauts won’t fall in love with one browser vendor. We are in favor of webstandards. I hope, you will soon get to the right track back again. It would be a shame if we would loose WaSP in our struggle for webstandards.

#21 On March 3rd, 2008 10:45 am Goodwinz replied:

This sounds like a sure fire way to make designing seem easier. However, by IE 1000 would this not become an extremely bulky application with supporting for the previous browsers?

#22 On April 3rd, 2008 10:39 pm Manoj replied:

MS should release two browsers,

1. IE 8 which has backward compatibility with IE 7
2. A new browser that is fully standard compliant and is like Opera, Firefox and other browsers.

This will address the issues of both IE only (intranet) sites/applications and standards compliant sites/applications.

Post a Reply

Comments are closed.