W3C home > Mailing lists > Public > public-html@w3.org > November 2008

Re: Almost Standards Mode still needed?

From: Daniel Schattenkirchner <schattenkirchner.daniel@gmx.de>
Date: Sun, 09 Nov 2008 17:21:34 +0100
Message-ID: <49170E0E.3060802@gmx.de>
To: public-html@w3.org

Henri Sivonen wrote:

> Is this the IE8 CSS formatter with the single images-in-tables quirk 
> like in the Almost Standards Modes of the other browsers? Could a 
> Microsoft representative please confirm?

We talk about the same thing.

> That simple (and I might say unintuitive) CSS is something the *site* 
> needs to provide in the (Full) Standards Mode. Therefore, there's a very 
> real compat problem with sites that don't provide it. That is, if 
> someone ships a browser that implements the CSS2 line box model fully 
> for images in tables for sites that now trigger Almost Standars in 
> Gecko/Opera/WebKit, layouts will break in an ugly way, which probably 
> won't cause a positive user opinion of the browser exhibiting such 
> breakage.

What sites will actually break in an ugly way? Are these gaps too much 
for websites to bear? We have a conflict here, but do we want to stay it 
the same eternally?
Websites that predate table sliced images have no actual problem in 
IE8's Standards Mode btw.

> Yes, we are. Most standards-aware new commercial Web design seems to 
> happen in the Almost Standards Mode--not in the Full Standards Mode 
> these days.

That's true. Most standards-aware new commercial Web design also seems 
to be tableless, though. The actual reason is that authors want @target 
and the likes, not because they're choosing Almost Standards Mode at will.

> If you consider the Quirks Mode, Almost Standards Mode and Standards 
> Mode in Gecko/Opera/WebKit, the one whose elimination would cause the 
> least disruption would be the Standards Mode (if made behave as Almost 
> Standards). However, at this stage, there's reluctance to changing the 
> modes.

An (unwritten) common aim seems to be a one or two mode enviroment (no 
one likes the mode mess). That's the conflict. We can't aim for one sole 
mode without breaking backwards-compat and vice versa. I just think the 
latter is slightly less important because the content should be what 
needs to be available across browsers. Not necessarily the design (for 
websites not created with current standards in mind, that is).

That's actually happening already. Mozilla encountered a problem some 
time ago: when they implemented the CSS transparent keyword for colors, 
some sites broke. [1]

As said, that actually makes me wonder, why Microsoft, breaking the web 
(oh noes) with IE8, but also providing mitigation, introduces this. It's 
not like the other's wouldn't follow, which in turn would cause negative 
user opinion for anyone else as well.

Well, I'm starting to ramble around. Just one more thing: It's 
interesting how one time it's the browser's fault and one time it's the 
website's fault, when something doesn't work. Just as it's needed to 
argue for or against something.


[1] http://www.squarefree.com/2008/07/18/transparent/
Received on Sunday, 9 November 2008 16:22:13 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:38:59 UTC