are you tired of this? then return to latest posts
(pardon me for the string of negativity)
As soon as I started reading Aaron Gustafson's article on ALA #251 Beyond DOCTYPE: Web Standards, Forward Compatibility and IE8 I knew this wasn't just another "best practice article". This was going to throw the entire web-constructing community into an uproar.
Ever since I got into the web standards groove, circa 2003, I remember finding experts blogs (for example simplebits, mezzoblue, meyerweb, etc.) great places to find insightful comments on best practices and real world code, embedded on their own sites. A List Apart, a magazine, has always been a the beacon of excellency throughout the years. That was why today's issue has created - and will continue to create - such an echo!
Aaron announces a feature that will be a part of IE8 called version targeting that basically allows developers to "freeze" their websites in a given version of a given browser - they expect to have other vendors implement this. This was driven by the fact that many websites "broke" when IE7 saw the light of day. Not because IE7 was worse, but because websites were custom-made for IE6, which included several bugs. When these were corrected in IE7, websites broke.
Now, Microsoft doesn't want the same to happen again with IE8, so their "vision" is to give the option to the developer to choose which render engine his/her website should be rendered with. By spitting an HTTP header (or meta tag) with X-UA-Compatible: IE=7, IE8 will load the IE7 engine and ignore the improvements made in the newest version of their browser. At least this way, websites won't break.
Despite some criticism, it's basically the reverse of the hated browser-sniffing of the late 90s. It's actually forcing the browser to adapt to the code in the page and not the other way around. Despite all this, I honestly think that's not the answer. Freezing browser engines in time cannot be the answer.
For the sake of keeping my sanity, I won't even touch the subject of IE8 behaving like IE7 if you don't opt-in using the HTTP header or the meta tag, I'll let you follow that link and read what Jeremy had to say about that. Please, do.
I just think we could use what we have instead of creating more mechanisms that will generate more entropy. We already have Conditional Comments and to be honest, out in the field, at work, I've frequently felt we could use the same mechanism for other browsers (Firefox, Opera...). Why not use that for fixing the "broken" websites by newer versions? Use a stylesheet specific for IE8 and fix your mess! Pronto.
I just can't get behind this concept. But I'm open to discussion, if you think you can persuade me, please do so in the comments. Even if Eric Meyer comes out in Aaron's defense, I see this as a huge blow to the web standards effort. It's like saying ok, we lost the war, let's just find a way of saving everyone some time and be done with it
. It's making each render engine their own little standard and allowing lazy developers to get away with it. Is this the web we want for our children??
Also, I can't see a future where every browser out there has to carry their old render engines on their backs.
Can you?
If you want to know more, read this:
Daniel Barradas on tue, 5 feb 2008 19:15
I guess that will work if you have one or two sites maybe even ten. Let's say you sold 50 and they suddenly stop working. A few things will happen:
1) as a developer you will look bad until you fix it to the general public
2) as a company you will look bad until you fix it for your clients
3) 50 sites? 100 sites? 1000 sites? Will you charge your clients? Or will you donate the (how many work?) hours? Will you have enough manpower to do it?
I think you might have overlooked this side of the question. So I do think they're saying "we can't do it" but on the other hand they might be "helping" end users, companies and clients. It might not be the best approach but this kind of mechanism would save a lot of headaches.
Just my two cents.
(Of course we wouldn't have this discussion if everyone followed standards... )
andr3 on wed, 6 feb 2008 07:41
Ok ok, I'm joking on a serious matter. Sorry.
I can relate to what you're saying, I truly can. The thing is that this is a very dangerous path. It's like they're covering their tracks. It's not Microsoft's purposed intent, that you so well explained, that scares me; it's what this mechanism will allow. It will allow websites to stay frozen in time and make the browsers incrementally bulkier. And since I do not believe nor want Mozilla and other browsers to implement this mechanism, this would make sites even more IE-specific! Back to the 1990s!
I know that in making a browser goes more stuff than meets the eye, I do. This is a very difficult matter, by all means. However, this whole thing has grown on me a bit, but still, their proposal that IE8 will work as IE7 if no version is targeted, is just plain wrong.
Apparently, if you use HTML5 doctype but use nothing but HTML4 elements, you're good. None of this mambo-jumbo. Could be a way out... (argh! More "hacks"!)
Billy on fri, 15 feb 2008 13:39
Stop! Use Mozilla! Why would you choose a donkey when you can ride a horse for free?
brotha-hug _=)
P.S.: Here's a free idea: let's all use the same base rendering engine from now on?
Comments have been disabled
Sorry about that.
Feel free to engage with me via twitter.
Tweet to @andr3