Brian Behlendorf (
Tue, 2 May 1995 16:52:39 +0500

On Sat, 29 Apr 1995, Martian wrote:
> Once upon a time you, Ian Higgs +44 171 510 8595, wrote:
> ++
> ++ What I wanted to see is the contents of the object as it would be
> ++ displayed if I had asked for it directly. If the object is HTML
> ++ then it should be parsed in context just like #include for C code.
> ++
> ++ JE] SGML has a mechanism for doing this, namely entity references:
> ++
> ++ Perfect! (NETSCAPE: Netscape 1.1 b4 please ...)
> In my opinion, there is no reason to have an <include> tag in html,
> where the <include> includes another file, combining both files to
> one html file. That will only lead to more network traffic and longer
> display times (since 2 requests have to be made.)

Unless of course the file is already there in the cache...

There are a few
> alternatives:
> 1) Read the manual of your editor and learn how to use macros.

Yes, ccp and #include are useful for now, but not a long term
solution. What if I want to include objects from other sites that don't
necessarily have control over or are updated frequently? Or I don't want
to recreate a tree of 100,000 documents simply because I've fixed a
typo on the copyright message?

> 2) Use server-side includes. Though that has the disadvantages of
> fooling proxies and caches.

Not "fool", "break".

> 3) Use something like m4 macros to generate the files, including the
> included fragments.

See 1.

> 4) Use <a>.

Exactly how it was originally supposed to work: <A REL="embed"
HREF="copyright.html">. And then along came <IMG> and the rest is history.
But I think you were suggesting something like

<a href="/copyright.html">Copyright</a>

which isn't quite there either.


--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-- http://www.[hyperreal,organic].com/