You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@shindig.apache.org by Kevin Brown <et...@google.com> on 2008/03/06 11:01:48 UTC
IE8 testing
Hey everyone,
I'm doing some testing of our current implementations with IE8 beta 1. So
far it looks like we'll have to make at least one significant change due to
quirks mode default differences, but otherwise it seems to work pretty well.
--
~Kevin
Re: IE8 testing
Posted by Kevin Brown <et...@google.com>.
On Thu, Mar 6, 2008 at 4:03 PM, Bruno Bowden <br...@google.com> wrote:
> That sounds good to me.
>
> I still think we should change the spec so that quirks mode is up to the
> discretion of the container. With future caja inlining, gadgets will be
> forced to go into standards mode, so it's better for us to avoid caja
> being
> opt-in.... which would dramatically lower adoption rates.
A container page that was in standards mode (almost all containers) would
have to use an iframe for gadgets where quirks=true, even if caja was used.
It can be on the same domain, though, which avoids most of the performance
issues anyway. The same would be true if the container was quirks and the
gadget was standards.
I find it unlikely that the iframes themselves will actually go away, even
if caja is perfected. We'll still have rendering and performance issues to
address, and iframes (on the same domain as the parent) offer a pretty good
solution for that.
>
> On Thu, Mar 6, 2008 at 2:54 PM, Kevin Brown <et...@google.com> wrote:
>
> > IE8 doesn't use a DOCTYPE to distinguish between quirks and standards
> mode
> > (except possibly for HTML 4 transitional) Instead, they use an HTTP
> header
> > either from the server, or specified in <meta http-equiv>
> >
> > The former is better since it avoids parsing the document twice.
> >
> > Solution:
> >
> > if gadget == quirks: always add the header (other browsers will ignore
> it)
> >
> > if gadget == standards, always output the correct doctype and never
> output
> > the header (works on all browsers)
> >
> > On Thu, Mar 6, 2008 at 2:50 PM, Bruno Bowden <br...@google.com> wrote:
> >
> > > IE8 defaults to standards mode, so I presume that you're proposing
> > adding
> > > a
> > > DOCTYPE to make IE8 explicitly use quirks mode? Can you apply this to
> > all
> > > renders, so that we're not doing useragent detection on the server.
> What
> > > DOCTYPE will that be?
> > >
> > >
> > >
> > > On Thu, Mar 6, 2008 at 2:01 AM, Kevin Brown <et...@google.com> wrote:
> > >
> > > > Hey everyone,
> > > >
> > > > I'm doing some testing of our current implementations with IE8 beta
> 1.
> > > So
> > > > far it looks like we'll have to make at least one significant change
> > due
> > > > to
> > > > quirks mode default differences, but otherwise it seems to work
> pretty
> > > > well.
> > > >
> > > > --
> > > > ~Kevin
> > > >
> > >
> >
> >
> >
> > --
> > ~Kevin
> >
>
--
~Kevin
Re: IE8 testing
Posted by Bruno Bowden <br...@google.com>.
That sounds good to me.
I still think we should change the spec so that quirks mode is up to the
discretion of the container. With future caja inlining, gadgets will be
forced to go into standards mode, so it's better for us to avoid caja being
opt-in.... which would dramatically lower adoption rates.
On Thu, Mar 6, 2008 at 2:54 PM, Kevin Brown <et...@google.com> wrote:
> IE8 doesn't use a DOCTYPE to distinguish between quirks and standards mode
> (except possibly for HTML 4 transitional) Instead, they use an HTTP header
> either from the server, or specified in <meta http-equiv>
>
> The former is better since it avoids parsing the document twice.
>
> Solution:
>
> if gadget == quirks: always add the header (other browsers will ignore it)
>
> if gadget == standards, always output the correct doctype and never output
> the header (works on all browsers)
>
> On Thu, Mar 6, 2008 at 2:50 PM, Bruno Bowden <br...@google.com> wrote:
>
> > IE8 defaults to standards mode, so I presume that you're proposing
> adding
> > a
> > DOCTYPE to make IE8 explicitly use quirks mode? Can you apply this to
> all
> > renders, so that we're not doing useragent detection on the server. What
> > DOCTYPE will that be?
> >
> >
> >
> > On Thu, Mar 6, 2008 at 2:01 AM, Kevin Brown <et...@google.com> wrote:
> >
> > > Hey everyone,
> > >
> > > I'm doing some testing of our current implementations with IE8 beta 1.
> > So
> > > far it looks like we'll have to make at least one significant change
> due
> > > to
> > > quirks mode default differences, but otherwise it seems to work pretty
> > > well.
> > >
> > > --
> > > ~Kevin
> > >
> >
>
>
>
> --
> ~Kevin
>
Re: IE8 testing
Posted by Kevin Brown <et...@google.com>.
IE8 doesn't use a DOCTYPE to distinguish between quirks and standards mode
(except possibly for HTML 4 transitional) Instead, they use an HTTP header
either from the server, or specified in <meta http-equiv>
The former is better since it avoids parsing the document twice.
Solution:
if gadget == quirks: always add the header (other browsers will ignore it)
if gadget == standards, always output the correct doctype and never output
the header (works on all browsers)
On Thu, Mar 6, 2008 at 2:50 PM, Bruno Bowden <br...@google.com> wrote:
> IE8 defaults to standards mode, so I presume that you're proposing adding
> a
> DOCTYPE to make IE8 explicitly use quirks mode? Can you apply this to all
> renders, so that we're not doing useragent detection on the server. What
> DOCTYPE will that be?
>
>
>
> On Thu, Mar 6, 2008 at 2:01 AM, Kevin Brown <et...@google.com> wrote:
>
> > Hey everyone,
> >
> > I'm doing some testing of our current implementations with IE8 beta 1.
> So
> > far it looks like we'll have to make at least one significant change due
> > to
> > quirks mode default differences, but otherwise it seems to work pretty
> > well.
> >
> > --
> > ~Kevin
> >
>
--
~Kevin
Re: IE8 testing
Posted by Bruno Bowden <br...@google.com>.
IE8 defaults to standards mode, so I presume that you're proposing adding a
DOCTYPE to make IE8 explicitly use quirks mode? Can you apply this to all
renders, so that we're not doing useragent detection on the server. What
DOCTYPE will that be?
On Thu, Mar 6, 2008 at 2:01 AM, Kevin Brown <et...@google.com> wrote:
> Hey everyone,
>
> I'm doing some testing of our current implementations with IE8 beta 1. So
> far it looks like we'll have to make at least one significant change due
> to
> quirks mode default differences, but otherwise it seems to work pretty
> well.
>
> --
> ~Kevin
>