![[RSS]](http://tweakimg.net/g/if/v2/icons/rss_small.gif){kind=icon}
X-UA-InCompatible
And Microsoft does it again by making their proprietary X-UA-Compatible tag completely incompatible with current browser behaviour...
Please read the Microsofts blogpost about the huge fuck-up they made with their X-UA-Compatible implementation. Not only do they go beyond all logic to make a <meta http-equiv> override actual HTTP-headers, they also made it override the doctype-switching logic that has more or less been reverse-engineered and used by most browser-vendors.
This is somehow unexpected. I would have expected f.i. a IE=7 directive to mean: tread this document as IE7 would have treated it (using doctype switching to determine quirksmode). It actually means: tread this document in IE7 "standards" mode - regardless the doctype (or lack thereof)...
So now MS backed up and introduced yet /another/ switch to get the more logical behaviour (IE=EmulateIE7)... Now that is completely backwards! For ages lack of a (complete) doctype (simply put) has always triggered quirksmode, but now some proprietary switch can change this unless you supply the /right/ switch.
What does this mean for quirksmode documents (by doctype matching heuristics) that are being served with X-UA-Compatible="IE=8" or X-UA-Compatible="IE=edge"? Will these trigger standards-compliant mode in IE but quirksmode in other browsers? Is this then some kind of trick to force other browser vendors to also start supporting this tag? It must be because else Microsoft would not have put the importance of their mode-switching tag above that of doctype switching...
Please read the Microsofts blogpost about the huge fuck-up they made with their X-UA-Compatible implementation. Not only do they go beyond all logic to make a <meta http-equiv> override actual HTTP-headers, they also made it override the doctype-switching logic that has more or less been reverse-engineered and used by most browser-vendors.
This is somehow unexpected. I would have expected f.i. a IE=7 directive to mean: tread this document as IE7 would have treated it (using doctype switching to determine quirksmode). It actually means: tread this document in IE7 "standards" mode - regardless the doctype (or lack thereof)...
So now MS backed up and introduced yet /another/ switch to get the more logical behaviour (IE=EmulateIE7)... Now that is completely backwards! For ages lack of a (complete) doctype (simply put) has always triggered quirksmode, but now some proprietary switch can change this unless you supply the /right/ switch.
What does this mean for quirksmode documents (by doctype matching heuristics) that are being served with X-UA-Compatible="IE=8" or X-UA-Compatible="IE=edge"? Will these trigger standards-compliant mode in IE but quirksmode in other browsers? Is this then some kind of trick to force other browser vendors to also start supporting this tag? It must be because else Microsoft would not have put the importance of their mode-switching tag above that of doctype switching...
 31-05
|
IE6 fucking up BASE (again)? |
Comments
By
SchizoDuckie,
Wednesday 18 June 2008 08:15
Ermmm.. Didn't we mention this just 2 seconds *after* MS announced this 'feature' ? 
By
Thyraon,
Wednesday 18 June 2008 09:49
You know, all these IE fuck-ups would be rather funny if you would forget the fact that things like this are likely going to cost me half a day of hacking my stylesheets in some future project .
.
By
DOT,
Wednesday 18 June 2008 13:29
The basic idea is to never use this tag. It should never be used in the wild. It's only sensible in unmaintained intranet applications that have used the IE6 rendering engine for the past 500 years.
But then you wouldn't necessarily want to (or be able to) go in and change the code in order to add the tag. You would want to set either a HTTP header in the server, or a company-wide browser setting that renders that web-app with the IE6 rendering engine.
But then you wouldn't necessarily want to (or be able to) go in and change the code in order to add the tag. You would want to set either a HTTP header in the server, or a company-wide browser setting that renders that web-app with the IE6 rendering engine.
By
Boeboe,
Wednesday 18 June 2008 20:39
Indeed, just don't use it.
Every time you'd think they would have learned by now, they remind us how much they still want to control the browsermarket.
Every time you'd think they would have learned by now, they remind us how much they still want to control the browsermarket.
