Hey! No arguing with John Resig around here ;)

But the script tag thing is a templating solution - so the advantage
is obvious: having to maintain a 10kb+ HTML fragment encoded as a
JavaScript variable (or even comprised of jQuery statements) is not
fun.

I'm not a fan of the script type="text/html" method myself, and if i
have to do client-side templates I prefer to have it as a hidden
element or in a hidden textarea. But these get read by screen readers
and Google, unlike the script tag method.

I'm not sure why the poster needs to access that DIV via a script tag,
but as its type is FBML - I'm guessing they got stuck implementing
some dodgy Facebook app... ergh. So, best of luck to him/her!


On Sep 10, 1:59 pm, RobG <robg...@gmail.com> wrote:
> On Sep 10, 9:49 am, Mr Speaker <mrspea...@gmail.com> wrote:
>
> > I think it's perfectly valid to have a div inside a script tag (or at
>
> It is never valid markup in an HTML document, a div element can't be a
> child of a script element.
>
> > least useful), if a div makes sense to the TYPE you defined for the
> > script.
>
> If you want to discuss HTML markup, then an HTML group would be a much
> better place to do that:
>
> <URL:http://groups.google.com/group/comp.infosystems.www.authoring.html/br...
>
>
>
> The content of a script element is not considered markup, it is
> script.
>
> > For example, John Resig uses a script tag with type "text/
> > html" in his micro-templating 
> > solution:http://ejohn.org/blog/javascript-micro-templating/
>
> I think that is a very bad idea for a web page. What advantage does
> that approach have to assigning the string to a variable and using a
> script language that the browser understands?
>
> He demonstrates that a script element can contain random text, and
> sets a trap for any browser that dares to be standards compliant and
> end the script element at the first occurrence of </.
>
> > In this instance though (and in reply to the original author) you add
> > an ID to the SCRIPT tag,
>
> While browsers tolerate that, it is invalid markup.
>
> --
> Rob

Reply via email to