Thursday, March 19th, 2009

IE 8 Launches at MIX09, probably without one of your features

Category: Browsers, IE

You may have noticed that day 2 at MIX was IE8 day (compared to the Silverlight 3 day one) in that the puppy launched!

This is good news in that IE 8 is better than IE 7, which is better than IE 6. Maybe, with 6 being two versions back, we have more weight to get rid of the chap. However, even though Microsoft has recently claimed that IE 8 is the fastest browser, its legacy weight still shows for now, and the other browsers are sprinting ahead.

I have a sneaky suspicion that IE 9 is going to be different and a lot better (new team? got the old greats back? or are they on Azure?), but that is wild speculation on my part.

We had a lot of comments on the news today, and one person who will rename anonymous purely because I don’t know if he would like to be attached or not said this:

When you do your IE8 post, don’t forget to toss out a line indicating the writing-mode bug was not fixed. Their only CSS3 feature and they’ve now effectively prevented the entire web for using it for the next x years. Hilariously, they also tried to cover up the bug by marking it fixed and forcing me to open a new one with the same description.

I also count two regressions in IE7 compatibility mode that make it different from IE7. Kind of a headache for those counting on sending that header.

So, thanks for the better browser guys, and we can’t wait to see what happens next. Shortly we will have IE 8, Firefox 3.5, Safari 4, Chrome 2, Opera 10, and others….. and the Web will have nicer cars to drive around on it. Let’s build applications that take these cars on a decent joy ride instead of boring them in the slow lane.

Posted by Dion Almaer at 10:29 pm
22 Comments

++---
2 rating from 51 votes

22 Comments »

Comments feed TrackBack URI

My plan is to simply not support IE at all for web applications. I’m not sure if I’ll be allowed to do that, but that’s my plan.

Notice I don’t say “for websites“. There’s a difference.

I’ve noticed many people writing about how one “should” support users without JavaScript, without flash, use graceful degradation, etc. I’m pretty sure they are talking about web sites, not web applications, but sometimes, its hard to tell. For web sites, it is obviously not acceptable to not support IE, and especially unacceptable to not support IE7 and even IE8.

On the other hand, I think it is perfectly acceptable to expect a proper browser (and even the latest version of one) for a web application, as the application is meant to do something, and the browser is simply a platform for that.

Much like how some software developers decided to only support Mac OS X Leopard when it came out, I think web applications should be supported only on platforms most suited for them.

That is why I am not going to try too hard to make stuff in my web applications work in Internet Explorer. It will just take away time that I could use to make the application better.

Comment by ialexi — March 19, 2009

The only good thing I see from this is that IE6 may finally die.

Comment by ajaxery — March 20, 2009

IE8 feels very slow. I have a page that loads an HTML table, then does an ajax call to get the data and populate the table, maybe 18 rows x 10 columns. Chrome executes it instantaneously, you can’t even notice the ajax call. In IE8 it takes about 3/4 of a second, it’s terrible. In compatibility mode it’s much faster. What gives?

Comment by rubley — March 20, 2009

IE6 will *certainly* not die: it will survive, in some form, in Windows Mobile 7. See http://blogs.msdn.com/windowsmobile/archive/2008/11/11/internet-explorer-mobile-6.aspx

Comment by bguijt — March 20, 2009

Good job in the anonymous person spotting this bug. Microsoft seem like a bunch of charlatons at times.

Unfortunately though, your attempt to make the person anonymous failed because you link to two pages with his name displayed for all to see.

Comment by demwunz — March 20, 2009

@ialexi, it would have been nice to define your understanding of web applications versus web sites. I, for one, don’t get your point.

Comment by igstan — March 20, 2009

I get the impression this was rushed out the door. I quickly bumped into the “missing content” bug in my web app, which according to their bug database was supposedly fixed. From reading around the web it seems a lot of the known issues managed to get shipped in the final release.

*sigh* another set of bugs to support.

Comment by Joeri — March 20, 2009

Sorry I mean Windows Mobile 7

I’ve tried IE 6 on Windows Mobile and is the same browser, including the same visual and DHTML bugs, only the JavaScript engine is different because is the same as IE v8 desktop.

Long live to IE 6 :)

Comment by jmarranz — March 20, 2009

@igstan Web applications are applications online. Bespin, for example, is a web application. They are meant to allow you to do tasks. You can’t browse through them; back/forward make little sense; they are not websites.

In comparison, Ajaxian is a website.

Comment by ialexi — March 20, 2009

The author’s not very anonymous when his name appears on the MS tickets you linked to. :)

I also noticed that there are still some issues with IE8’s Compatibility View and using X-UA-Compatible to set emulation mode–they don’t behave the same.

I’ve blogged about the issue here:
http://blog.pengoworks.com/index.cfm/2009/3/19/Internet-Explorer-8-issues-with-Compatibility-View-and-EmulateIE7

I’m planning on posting a test case today.

Comment by DanSwitzer2 — March 20, 2009

It looks like IE is still not put up with superior Mozilla, and tries to win back the first position. :)

Comment by Iflexion — March 20, 2009

Having to deal with 3 browsers from one vendor just isn’t worth the money, let’s just drop in an EmulateIE7 tag and be done with IE8.

Comment by Jadet — March 20, 2009

IE9 will be much better because it’ll adopt webkit or gecko… at least that’s what some naive part of me is still hoping. We may be able to wear Microsoft down enough by then that they finally get it. Unfortunately they have a long history of, well, not getting it no matter what the public says, so it’s going to be a challenge.

Comment by pendensproditor — March 20, 2009

I came across the same writing-mode bug therefore I know the secret anonymous identity. So far the only other issue I’ve come across was that they’ve appeared to fix the ginormous button width problem, so I had to special-case the IE hack I used to fix it. Also I have some browser detection that appeared to be broken, then after instrumenting it to alert() it wasn’t broken, then wasn’t broken no matter what. There seems to be a race-condition or something similar that makes IE8 not have MSIE in its userAgent, or not have executed global javascript before executing other javascript. So far, everything else worked surprisingly well, but its still early and writing-mode is a doozy.

Comment by JonathanLeech — March 20, 2009

@ialexi: I’ve been preaching that distinction for a couple of years now, and for a change I can actually prove when I first said something: July 2006, in the first chapter of my first published book (gratuitous plug: apress.com/book/search?searchterm=zammetti&act=search)… and I know I was saying it even before then.

I agree with you 100%, and it’s a very important distinction to make because it informs how you build what you’re building greatly. It’s not, as some people think, as simple as the difference between an internal-facing product and a public-facing product, it’s what the product *does*, what it’s purpose is and who it’s for, among other considerations. I wish more people understood the difference and designed accordingly.

Then again, if they did, people like me wouldn’t have anything to write books about :)

Comment by fzammetti — March 20, 2009

Mmmm interesting, can’t wait to check if IE8 is faster than other browsers.. too bad that IE6 is not going to “die” anytime soon.

Comment by hquinn — March 20, 2009

I found that you can no longer pass 0 instead of false to an xmlHttpRequest’s open method to specify a synchronous call. It works in IE6 and 7, safari, and firefox but not in IE8 or IE8 with IE7 compatibility mode on. It makes me think there could be lots of other minor incompatibilities between IE7 and comptibility mode.

Comment by paulsidekick — March 20, 2009

I just felt like my soul died a little bit after reading this.
Also, is it just me or is that microsoft page rendering funny in FF? Figures.

Comment by cdude — March 20, 2009

if(navigator.userAgent.indexOf(‘MSIE’)!=-1){window.close()}

Comment by jgeerdes — March 20, 2009

@jgeerdes : hahah, good one ! :P

Comment by vsync — March 21, 2009

@ialexi, amazing post!

However I have some comments on your writings.

Alex: My plan is to simply not support IE at all for web applications. I’m not sure if I’ll be allowed to do that, but that’s my plan.
Sergey: That is a good plan. Looks like you spot the trend! And of course, your plan will help users of your application (the end-users).

Alex: Notice I don’t say “for websites“. There’s a difference. I’ve noticed many people writing about how one “should” support users without JavaScript, without flash, use graceful degradation, etc. I’m pretty sure they are talking about web sites, not web applications, but sometimes, its hard to tell.
Sergey: Probably the border between web applications and web pages lay somewhere where you decide not to support certain browser?

Alex: For web sites, it is obviously not acceptable to not support IE, and especially unacceptable to not support IE7 and even IE8.
Sergey: Agree.

Alex: On the other hand, I think it is perfectly acceptable to expect a proper browser (and even the latest version of one) for a web application, as the application is meant to do something, and the browser is simply a platform for that.
Sergey: Yes, “proper browser” said well. By the way, what kind of “web aplications” do you have in mind? Something like “Google Docs” and “GMail”? If so, these applications can run in all “proper” browsers, where proper is roughly defined as:
– Internet Explorer (released in 2001 or so)
– Opera 9 (released in 2006 or so)
– Safari 3 (released 2007 or so)
– Chrome (released 2008)
Take a note, when first “proper” browser was released (didn’t WebKit (Safari, Chrome) or Presto (Opera) exist at that time? How long did it take them to catch up?)

Alex: Much like how some software developers decided to only support Mac OS X Leopard when it came out, I think web applications should be supported only on platforms most suited for them.
Sergey: Right, but what platforms? Do you mean the “web as a platform”? If so, then go ahead and support it (indeed the runtimes of such a platform include all browsers), your task is then even simplier – that is one platform.

Alex: That is why I am not going to try too hard to make stuff in my web applications work in Internet Explorer. It will just take away time that I could use to make the application better.
Sergey: Actually there is no need to try hard to make “your web applications” work in Internet Explorer. It is just a little bit different. That’s all. Again if you in your “web applications” could use great libraries like jQuery, it would be pretty much the same.

And just a litle be more to collaborate. Alex, the Internet Explorer browser is the farther of the platform we discussed earlier. Not only it first enabled the platform, it is still ahead there! For example:
– it has native support for an essential HTML Componentization model, that is missing from most of of other browsers
– it has Data-binding model builtin, that is only being developed as part of HTML5 standard and will come in other browsers later on
– it has Drag-And-Drop model builtin, recently adopted by other browsers and being discussed as part of HTML5
– it has declarative 2D vector graphics and animation, neccessity of which was also acqknowledge by HTML5 intiative etc.
Not to say that the data communication API, called XMLHttpRequest was also first enabled in IE

And it got that all back in 2001 or earlier.

So, stop blaming IE for being too long and too far ahead. Start using its power instead.

Cheers,
Sergey.

Comment by SergeyIlinsky — March 22, 2009

While comparing with IE6 and IE7. I think IE8 is some what better and has so many bugs debugged..

Comment by cirjegan — November 18, 2009

Leave a comment

You must be logged in to post a comment.