On 17 March 2011 11:47, Alan Gresley <[email protected]> wrote: > IE7 > compatibility modes in both IE8 and IE9 do not show the true IE7.
There's a lot of FUD surrounding this topic, and a fair bit of conflation of terms. I assert that, according to intended use, and as far as I'm aware: • 'Compatibility mode' (no version number) is a half-way-house kludge, triggered by the user via a specific button, forced by the developer via X-UA-Compatible meta value, or triggered in the DOM inspection tools, developed by Microsoft to emulate certain aspects of the previous browser's rendering properties while still allowing for other unspecified improvements of the current (JS processing, less 'controversial' rendering 'improvements', etc) to come through. • What I'll refer to as 'Act-as' modes (I'm going to avoid specifically avoid the term 'emulation' for this description, because an X-UA-Compatibility value of 'Emulate=IE8', for example, is the equivalent of IE9's compatibility mode, and distinct from this feature), forced by the developer via X-UA-Compatible meta value, triggered in the DOM inspector via the Browser Mode dropdown in the method I described earlier — which is a way of behaving, at least according to intention and as far as I'm aware, in practice, EXACTLY as the browser described — including JS processing, and generally the complete lack of any future version's properties. Microsoft describe X-UA-Compatible in detail here: http://blogs.msdn.com/b/askie/archive/2009/03/23/understanding-compatibility-modes-in-internet-explorer-8.aspx I can prove that the two things above are distinct: uk.msn.com (and the vast majority of sites with any kind of complex layout) renders visibly differently in IE9's compatibility mode than it does in IE9 running IE8 Act-as/browser mode. Screenshots here: barneycarroll.com/IEbrowserModes.html Now I have always used the latter feature (specified browser mode via the DOM tools) in IE8 (and now IE9) to test site performance for lower versions (excluding IE6 and lower, for which I use a virtual machine), believing that the /exact same experience/ can be perfectly emulated. David says he has seen discrepancies between standalone IE7 and IE8 under IE7 browser mode. Are you saying the same, Alan? Can you provide test cases, or recollect what kind of discrepancies these were? (or are you simply saying that compatibility mode does not accurately reflect any specific version of IE, which I hold to be true?) > I would say that if it works in recent versions of WebKit, Presto or Gecko, > it should work in IE9. Please remember that IE period is not like any other > browser since it has distinct versions. In saying this, it is not a bad > thing. It has allowed IE9 to be on par with other browsers in respect to > CSS2.1 and with large support of CSS3. I agree. At the very least we can be grateful that it's easily possible to target the distinct versions of IE and supply different code or force different behaviour, whereas the incremental rolling-update format of other browsers means any quirks in a particular version are impossible to target and very difficult to isolate. Regards, Barney Carroll [email protected] 07594 506 381 ______________________________________________________________________ css-discuss [[email protected]] http://www.css-discuss.org/mailman/listinfo/css-d List wiki/FAQ -- http://css-discuss.incutio.com/ List policies -- http://css-discuss.org/policies.html Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
