Domain: edwards.name
Stories and comments across the archive that link to edwards.name.
Comments · 118
-
Old Slashdot Story
Some sick bastard is trying to fix this using Stylesheets:
http://dean.edwards.name/IE7/intro/
http://developers.slashdot.org/developers/04/03/1
2 /0454228.shtmlHe even included the png hack. Sick sick sick.
-
Re:Great browser, but...
I guess IE7 is this little thingy, it's IE6 but with better CSS handling more standard compliant etcetera.
-
Re:Great browser, but...
Why would you ever be hired with that attitude? The fact is, all the IE moaning is a BIG MYTH.
While I agree with you that one has to make pages work with IE; allow me to say that you are talking out of your ass.
IE is a pain to develop for. It either doesn't comply with the standard, completely breaks it, or causes weird quirks that make no sense.My sites always have the proper DOCTYPE, they have to in order to validate (I validate all pages prior to posting). Sometimes _having_ a doctype makes IE show new bugs.
Frankly, unless you're building some fancy site with the absolute latest CSS and Javascript features, IE will render your page *just fine*.
Your sites are obviously _very_ basic with minimal design. The moment you pass into medium complexity you start running into problems. I don't use any Javascript.
Yet, why exactly do you need to be using 32-bit images on your page? It's quite easy (and beneficial for download speeds) to get by with 8-bit images.
We need 32 bit images in order to have variable alpha. If you don't have variable alpha, then you can't have any image without including the background it's going to be over.
Including the background in your image completely negates the advantage of declaring colors in the CSS. Normally I could change the background of my site by changing one line, now I have to change all images as well.
There's other reasons why you might want a variable alpha, say to have transparent graphics or windows (the navigation on the left). Which are not heavy as the image is just 2 x 2 pixels and weighs in at a hundred bytes or so. In fact there are tons of creative uses that people haven't even thought of.
and tell me why they play such an integral part of your development process that you cannot find a way around them.
You're missing the point entirely. The idea is to allow the designer do her thing (so I can get a beautiful unique website) and then translate that to CSS, compromising wherever needed; not just half-assedly add some design to a text document.
However, when one tries to do anything remotely interesting, Explorer completely dies on you.
The best solution I found so far is using IE conditionals and serving IE a different and sometimes strange CSS file. In the end I usually figure it out reasonably, but it takes ridiculous amounts of effort and the site is never as good as it's Safari/ Mozilla
/Opera counterpart.If you want an example you can look at my current site (I linked it above too), I've only starting the IE fixing, you can see the ugly IE file with filters: and other weird junk so that PNGs show up with alpha: here. So far it's taking a serious effort and even when finished won't look as well anyway. Look at it under both IE and then Firefox.
v -
An IE quirks workaround.
Here's a quick link that may be of some value to you web-heads:
IE7 stylesheet
Include this stylesheet with your code and some of the IE quirks get handled for you. The page contains enough info to get you going. Good luck! -
A solution is at hand
IE7. Check it out.
-
Re:Mozilla Blues
Mozilla's equivalent to 'behaviours', XBL, is actually more powerful than Internet Explorer's.
As a demonstration of this, one elite scripter has implemented CSS behaviours using XBL.
You just pull in the script and most of the IE behaviours will "just work." Demos on-site.
:-) -
Re:Petition google to rank XHTML pages highly too
Or instead of waiting for IE to be fixed, you could cheat.
-
Re:CSS3 Opacity added
The "filter" CSS property is a proprietary extension to IE. It is not now nor ever will be part of the W3C CSS3 spec.
As an aside, Mozilla has supported opacity/transparency as well for a while now through its proprietary "-moz-opacity" CSS property. The main difference here is that while it's clear that you are using a browser-specific extension with Mozilla, the IE variant appears like any other CSS property, and you may not realize that you are in IE-only land.
But if proprietary extensions don't bother you, "-moz-opacity: 0.93" and "-moz-opacity: 0.3" will give the same effect to your post-its as the filter attribute does.
However the real news item here is that the new Firefox code -- like Safari 1.2 before it -- supports the "opacity" property as specified by the W3C CSS3 color module working draft. Perhaps it will be incorporated in the IE7 hack for compatibility. -
Re:here we go again!
...they would if it was indeed a standard...I actually hope you are right. They do have a large install base (i'll give them credit for that). However i realize that there is already a long list of standards that they have not implemented.
Not having those standards implemented (or only having some of the standards implemented) is part of what (for me) makes html programming a pain. If you don't know what I'm talking about try and use some CSS2 stuff like
:before and :after in IE. (there are workarounds like IE7, but they should't be needed). -
The worst part is IE development has stalled.
Microsoft business practices aside, I really, really hate the fact that IE development is at a standstill. Microsoft has said they won't release a new IE until Longhorn.
Meanwhile we have to kiss web standards goodbye to please 90% of the public using IE.
Amusingly there's a work-around under development called IE7, mentioned on Slashdot a while back.
But the fact is Microsoft is keeping us from adopting things like CSS2, PNG and SVG more than anything else. -
IE hacks for PNG alpha blending / CSS fixes (IE7)
I'm surprised that nobody (at least who's been modded up enough) seems to have said anything about two particular projects that attempt to deal with a couple of the main problems that developers tend to have with MSIE.
"IE7" is an Internet Explorer hack that parses standards-based CSS that you provide in a page, and mangles it so that earlier versions of IE display it how it's supposed to be displayed.
"PNG in Windows IE" is a hack that tells IE to use a separate ActiveX control to load any PNG's in the page, instead of the internal image display code. This causes it to get alpha blending right. (I think there are a few variations of this hack around the web besides the one I've linked to.)
Both are javascript hacks that you can include at the top of a page and add the appropriate construct around them so that only IE will see them. Clearly they're not perfect, and I'd be edgey about using them in important websites without a lot more testing.
But has anyone actually used them effectively? How useful are they?
I've managed to get the PNG hack working, but I still haven't been able to get IE7 going. (Possibly something to do with the server sending the wrong MIME type.)
-
Re:Well...
It's that the standards are positively byzantine, and so complicated to implement that it's simply not worth the effort.
No, most are actually pretty simple. Granted you have to first get through the language barrier (since these are specs, they have to use precise wording). However, the concepts aren't that difficult. HTML, CSS, XML, XSLT, and RDF are pretty easy to author. XML and CSS is also easy to parse (for the latter, see IE7 or my demo's source). Sometimes there are errors, but the W3C is pretty good at clarification and fixing errors, such as in the specifications: CSS level 2 revision 1, XML 1.0 third edition, HTML 4.01, and more. I don't understand why they are so confusing to you.
-
Re:Avoiding Piracy
If you want to save some of the hassle with bits of CSS missing from IE, you might want to check out this "IE7" thing. You put it in a stylesheet and it makes various things work which didn't already work.
-
IE CSS "compatilibity patch"
This is an attempt at providing a kind of "compatibility layer" for IE in order to essentially "emulate" CSS compatibility on IE. He calls it a "compliance patch" or "IE7"
:) The idea is that web developers include this file that uses DHTML to provide a compatibility layer for IE. If the user is using a standards-compliant browser (e.g. Mozilla) then the page renders as normal. So web page developers can just develop straight to standards. -
Re:What in God's name...?
I can't imagine a serious book on CSS talking about IE CSS extensions.
Why not? It is important to know what extensions are aviable for the major HTML engines (Opera, Gecko, KHTML, Tasman, and Trident) for several reasons. If you're writing software specific for a particular web browser, then it becomes extremely important to know about such engines. Or when most browsers support part of the standard, and a simple hack can add the other one to the list. Or when you're writing software to add some CSS support to IE (like Dean Edwards does in IE7). When is a little more knowledge ever a bad thing?
-
Re:Still doesn't work well for me
-
Re:IE works just fine
Um...in fairness, the grandparent post was talking about IE5 for Mac. Seeing as how MS has abandoned IE for the Mac at that version...there isn't really any Microsoft alternative he can upgrade TO.
And yes, even the most recent version of IE for windows STILL does not handle CSS fully or appropriately in all cases. To wit -- if it were, why would this project exist?
-
Re:I was expecting another kind of patch
I've done this too: http://dean.edwards.name/moz-behaviors/ (when my site is back up!)