Posted by Joel Shepherd on 11/14/06 16:27

Toby Inkster <usenet200611@tobyinkster.co.uk> wrote:

> If you allowed includes in HTML, then you could have this situation:

From the example to follow it seems you're talking about client-side
includes, right?

> <div>
> I am in a DIV.
> <include file="foo.html">
> I am in the same DIV.
> </div>
>
> Where "foo.html" is:
>
> </div>
> <div>
>
> Then the meaning, and the generated tree of the first document is
> changed when the second file loads up. For this reason, it has always
> been that when an HTML file is sent to a browser, it is always the whole
> thing in one file.

I'm confused. Why can't exactly the same problem happen with SSI (which
strictly speaking, is not part of HTML, but is used closely with it)?
I.e., what difference can it make whether the include functionality is
client-side or server-side? With respect to generating broken or
unintended constructs, it looks like they have equal potential.

--
Joel.

[Back to original message]