Microsoft Releases IE7 Beta 3

Kawahee writes "Microsoft has released IE7 Beta 3 to the public. From TechNet Flash: 'As a result of customer feedback, IE7 Beta 3 contains some feature changes in addition to the planned reliability, compatibility, and security improvements. If you've previously installed a beta of IE7, you should uninstall it before installing this release.' For the first time, the Administrator's Kit for Internet Explorer 7 is also available, which is described as 'the most efficient way to deploy and manage Web-based solutions.'"

Comments

Let's see.

[Inoshiro]

IE 7 still did not correctly implement the box model, positioning, all CSS1, all CSS2, or any CSS3. The same IE-specific parsing bugs for CSS are in place in IE 7.

At this point, you have to ask; is it that the people at Microsoft are incapable of producing a specs-compliant rendering engine (when every one else in the world can?), that they are roped by backwards compatibility, or that they think people will see IE 6 + tabs as "good enough"?

It's to the point where every site I make has 2 code paths: not IE, and the IE-specific overrides (up to an additional 20kb per page!).

Re:Let's see.

[CaymanIslandCarpedie]

Perhaps you mean that most browser makers don't shoot for acid test compliance

I cannot claim to know exactly what the GP meant, but my reading of it thought he was refering to some peoples view that ACID2 is biased against MS. The issue was basically some of the authors of the original ACID test (and other outside parties) have accused the ACID2 of basically being used as a marketing tool against MS. Thier view on this is basically the original ACID test was built by asking "what are the most important features needed" and was then a tool to help all browser companies work toward compliance by meeting that test. They claim ACID2 was basically built by asking "what features is IE worst at and will have the most trouble implementing".

Now I have no idea about any of this but thats what I'm guessing he was talking about. That it was supposedly built more as a challenge to MS than to help ALL browsers work toward a common goal. How true any of this is, I got no idea. Just stuff I've ran across on the web. Here is one reference I was able to quickly find.

Re:Let's see.

[CastrTroy]

I think what he's saying is that just because the browser passes the ACID test doesn't mean that it renders 100% of the pages correctly. As far as I'm aware Safari/Konq is the only browser that passed it fully. Yet, other than IE, it's the browser I hear most about when it comes to rendering errors. You could make a program correctly renders the input of the ACID test. It wouldn't really look at the input, just read the file and output what it's supposed to. The ACID test doesn't test 100% compliance. It's just something for bragging rights. It doesn't show real world results with the billions of pages on the web.

Re:Let's see.

[jonadab]

> So Firefox only scored a 1500 on its SAT. IE is still wearing a helmet
> and drooling on itself as it takes the short bus back to Redmond.

You exagerate. Yes, Firefox handles CSS (especially CSS2) rather a lot better than IE, but the browser drooling on itself on the short bus with regard to CSS is Netscape 4 (remember that frustration?), and the one scoring 1500, when it emerges, will handle a lot more CSS3 than any browser I've yet seen can manage so far.

Perhaps a better analogy would involve grade levels: IE7 is still a couple of grade levels behind Firefox, but in time it can potentially get to where Firefox is today. Sure, it needs to study CSS2, but that's coming up on its academic schedule, I imagine.

> IE is the biggest frustration on the planet right now to anyone who actually
> works in this industry.

Okay, that much is true. (Well, it's true if you define "this industry" as web development specifically, rather than lumping all computer stuff together as is often done.)

XHTML DTD Parsing

[charleste]

But no mention in the FA as to whether or not IE will support the DTD specs from W3C (currently, IE 6 does not ignore entities that have the IGNORE attribute, hence XHTML 1.1 DTDs from W3C are not parsed - they throw errors.

Still the slowest browser around

[drinkypoo]

It seems to be slightly faster than b2 was, but it still takes longer to load than firefox (which is hilarious, because I consider firefox to be one slow-starting mofo) and loads pages about half as fast. Probably it's a checked build or something, but it's unacceptable. I'm uninstalling and going back to IE6 (for those things for which I need IE) as soon as I feel like rebooting.

Does IE7 support the <CANVAS> tag?

[cr0sh]

I find this tag to be one of the more interesting "new" HTML tags. From what I understand, the Safari browser (OSX?) was the first to have it, then (IIRC) Opera, then finally the Firefox browser. IE6 doesn't support it, although people have been able to create some interesting workarounds using SVG.

I just like the possibilities this tag brings to browsers and web applications, as well as (simple) gaming. However, I haven't heard anything about it working (or not) in any of the IE7 betas that have been released yet...

No, it doesn't

[Kelson]

Changes to IE7's rendering engine have been primarily in fixing bugs and catching up to established standards like CSS. came out of WhatWG (or, more precisely, came out of work Apple was doing to make Dashboard widgets possible, then submitted to WhatWG), which, so far, the IE team appears to be ignoring.

Since WhatWG's work does seem to be catching on, with Opera, Firefox and Safari all implementing features and not just talking about it, there might be some pressure on Microsoft to start adding support in IE 7.5 or IE 8.

Re:IE 7 is a Major Improvement

[99BottlesOfBeerInMyF]

It's not perfect, but it's a major improvement in basically every way over IE 6.

Wow, seven different press releases/comments from MS. Well, someone just installed it on a test box, so let me take a look at the HTML I'm outputting. Golly looks just the same as IE6. IE fails to show either the CSS or XHTML formatting it failed to before. Now lets take a look in some other browsers. Firefox works. Opera works. Safari works.

They can talk all they want, but they still haven't managed to do anything. Talk is cheap. Luckily, as this is content that only network security experts will be looking at, nobody cares is it is unformatted for IE users since none of them would touch the bloody thing.

Re:Wishful Thinking

[rainman_bc]

That's a part of life. They need to take a queue from practically every other platform and just BREAK STUFF sometimes for the greater good.

You must be new around here. Every time a new Microsoft OS comes out, someone bitches endlessly about killer app from 1992 not working any more.

Honestly, Microsoft is damned either way around here. On the one hand they break compatibility and a bunch of sysadmins chime and and bich about some custom app not working any more and they will refuse to upgrade their LAN, and if they don't, well you are exactly correct... Some security holes are left open.

CSS is overrated

[nmg196]

I'm a web developer and the main reason I don't use CSS exclusively isn't because of poor browser support or rendering bugs, but CSS itself.

CSS is really really annoying. Sometimes you just need to use tables because even with a good standards compliant browser like Safari, it's just not possible to do what you want with CSS.

Things which use to be REALLY easy with tables in quirks mode (like a 3 column layout, 100% high with a header and footer) are almost impossible to implement using CSS. There are a multitude of websites giving example or template layouts in CSS. Some of these show 100 odd lines of CSS with loads of exceptions for each browser. The same thing can often be achieved with a single table in about 8 lines of code.

I mean just look at the CSS for slashdot - there are pages and pages of it and loads of browser exceptions. It had none of this complexity when it was a basic tables based site. They've just used CSS whereever possible for the sake of it.

The box model is really really annoying. If I tell something to be 200px wide, then it should be 200px wide all the time. However in most standards compliant browsers, it will be wider than 200px because it adds the margins and borders to the outside of it's own width. That makes it very difficult to work with as you have to subtract all those dimensions from the width you're giving it and need to alter about a million dimensions everytime you want to make something a bit wider.

CSS has a long way to go in my opinion.