David Hucklesby wrote:
> I agree with Rick here. While a suitable DOCTYPE will put IE 6 and 7
> into so-called standards mode, if your visitor comes from Google's
> cached version of your page they will see it in "quirks mode" anyway.
> Possibly one reason that Georg regularly uses the XML declaration to
> trigger quirks mode at all times. (?)
A very minor reason, but yes.
I level out most differences between "standard mode" and "quirks mode"
to a point where the resulting "mode" doesn't make a layout appear to
break in any regular browser. I have more "serious" reasons for this
approach than what google-caching and other "disturbances" may
introduce, but having my work survive through most of that too is a nice
I observed years ago that all work on future standards ((x)html5, CSS3
etc.) tried to "undo" the split between modes - make them one, as if the
split had been more of a "work-accident" than the result of in-depth
reasoning by previous spec-WGs. Thus, I saw no reason to promote a split
- promoting "standard mode" over "quirks mode", especially since I've
found that there's really _nothing_ to gain.
Most standard-compliant browsers didn't really have a mode-switch to
begin with, simply because they didn't need one. Standard-based markup
and CSS was just treated more consistently than non-standard one, since
browsers had a common standard to fall back on for the former. Those who
created IE/Mac apparently couldn't make that "one standard" formula
work, so they introduced the mode-switch (that spec-WGs now are trying
to reduce the effect of).
After the arrival of a "standard-rendering" IE/Mac, standardistas
started to promote mode-switching and mode-hacking to a point where it
became almost a religion in many camps. I've felt the heat at my end
too, but since I'm pretty agnostic in this respect I have never been
able to see the logic behind their arguments.
Only IE (on both platforms) has ever needed a mode-switch, and now one
(IE/Mac) is dead and the other is slowly closing the mode-gap and we
only have to live with its "old sins".
Now I can test both IE-modes in the latest IE-version, which means I can
pretty much leave out testing in older IE-versions. Minor breakage in
IE5.x is rarely relevant anymore.
All pages I create comes out pretty much identical in all regular
browsers / versions / through google-cache etc., and only those pages
that I have deliberately created to test mode-differences (that I
otherwise don't rely on) show any differences - which they should.
So, what comes out in the end-user's browser is entirely up to me, which
I have to say "suits me just fine" :-)
Relevant addition to the above:
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/