New IE7 Information Announced

Brandon writes "Looks like the IE team is trying to catch up to some of the major OS browsers. They have finally added proper PNG support and have fixed numerous CSS bugs. The full post is on The Official IEBlog." From the post: "We're doing a lot more than this in IE7, of course, and we're really excited that the beta release is almost here - we're looking forward to the feedback when we release the first beta of IE7 this summer. Stay tuned for more details as we get closer to beta."

Comments

nuts to -moz-border-radius

[rebug]

Instead of implementing a vendor-specific tag, why not support the proposed CSS3 border-radius property?

Re:nuts to -moz-border-radius

[rebug]

They support CSS3 selectors, which are still candidate recommendations.

I think it's that their support for the border module is terribly incomplete, so they're not going to suggest that they support any of it quite yet.

Re:middle-click for tabbed browsing

[tepples]

Then I probably shouldn't tell you about Maxthon or any of the other IE wrappers that add tabs and retain all the ActiveX holes.

min-width and hacks

[UnConeD]

Internet explorer actually treats width and height as min-width and min-height. Very annoying if you don't want it, but you can use it like this:

select {
min-height: 100px;
_height: 100px;
}

IE will (mysteriously) ignore the underscore prefix and parse the second style, while compliant browsers only recognize the min-height style.

This shows that the important question is in fact not "how many CSS bugs will IE7 fix?" but "how many CSS bugs will IE7 keep?". These bugs are currently needed to make IE6 behave properly. If IE7 fixes the rendering bugs but keeps the parsing bugs, we'll have to figure out new bugs to update the IE6-only hacks with.

Re:min-width and hacks

[Sentry21]

The worst is that web designers used specific parsing bugs in IE5 to trick it into rendering things the way it's supposed to. IE6 then went ahead and fixed the parsing bugs (so the trick doesn't work) without fixing exactly what it was that the trick was needed for in the first place.

The latest trend is to use Javascript to set the styles, which (imho) is a much better idea - never rely on bugs to implement features.

Acid2

[Omniscientist]

Looking at the blog's comments, I saw someone mention the Acid2 test. I did a google and took it, and my Firefox 1.0.2 failed it.

Anyone know more details about this test and what browsers do pass it (I'm guessing IE6 doesn't, I don't have it so can't test it)? I'm surprised Firefox didn't, not because I'm a fan boy or anything, but because I presumed Firefox was in accordance with most of the standards.

This is the test and this is what it should look like. Here's some info about how it works.

Re:Thank you, Microsoft

[mmusson]

In this IE6 world a webdesigner cannot use transparent pngs.

This is not completely true. PNGs with 8-bit alpha channels render correctly. Google maps uses this trick to create the push pin drop shadow.

Re:Thank you, Microsoft

[medeii]

Your assertion is misleading.

Firefox, Opera, Safari, and other "niche" browsers render PNGs correctly, with the use of the 8-bit alpha channel. IE6, on the other hand, ONLY recognizes boolean transparency in PNGs -- in other words, it treats them like GIFs. It is possible to force IE5+ to recognize the full alpha channel, but only with the use of a Direct3D filter command.