W3C home > Mailing lists > Public > www-html@w3.org > August 1998

RE: include facility in html

From: David Norris <kg9ae@geocities.com>
Date: Fri, 28 Aug 1998 15:53:05 -0500
To: <www-html@w3.org>
Message-ID: <000601bdd2c5$dad19340$1dd628cf@illusionary.dyn.ml.org>
Client side:
	First, I will say, as I always do, that adding include functionality to
HTML would likely be bad at best.  The server is the most efficient place to
do this sort of thing.  The server, hopefully, has a faster connection to
the network to gather the included content.  The server is by far more
'knowledgeable' than the client about the content being served.  What if
your 'included' content isn't available at that time or to that client?
Well, your page would break.  That is not good.  The server would 'know' how
to handle that situation, provided you have some sense about your design.
(If it doesn't know how to handle the situation it can easily inform a human
instantly.)  For instance, the server could use a cached copy of that
content or retrieve the content from some other source.  All of which, the
client couldn't do effectively.

The HTML include would be neat for people with junk servers, but, that is
the extent of it.  Even SSI would be better to generically include content
than using the client to include the content.  And, SSI is terrible.  It
seems that the people asking for HTML include functionality are usually
looking for a way around a bad server.

BTW, NT's Information server does support SSI.  It supports SSI in exactly
the same manner as Apache and nearly every other web server.  In fact, the
Information Server training information refers you, via URL, to the Apache
XSSI documentation for syntax.  (Source: Mastering Internet Development(R)
Part#:91256, Microsoft Press.  MS Internet Information Server interactive
training course.)  Yes, I took the course. ;)

Server side:
	I don't know that a written standard is 'needed' for server scripting.  I
might be nice, to some extent.

The server doesn't have the problem of clients.  The server normally runs on
one system that is tightly controlled by a few people.  There may be
millions of clients for every one server, and, those clients are running on
a plethora of hardware that all have widely variant requirements.

Most servers are internally similar in function.  (A Unix-based server
doesn't even need a keyboard, screen, or many of the other things normally
associated with a computer.)  The same constructs will run on most server
machines.  The popular server tools all work equally well on nearly all
modern OSes.  PERL, PHP, Apache, and most other common server components run
on Unices, Windows, and Macintosh. As well as most other OSes that don't fit
those three descriptions.  In a well designed system, you can drastically
change the hardware and never even touch the server.  You could be running
Linux on a Sparc one day and switch to MacOS 8 the next without breaking the
scripts and content served.

Whereas, even minor changes can drastically effect the client's ability to
use a complex web site.

I think that the reason you don't see a written standard for web servers is
very fundamental.  The people that would consider writing the server
standards see that the standard is already agreed upon.  They didn't write
the standard in plain English.  They wrote it in C.  Projects like Apache,
PERL, PHP, etc are the de facto server standards.

Now, with that said.  What might be good is to write down a standard set of
rules for creating content using these de facto standards.  Many people
blindly start using these technologies without considering whether what
there are creating is compatible with the fifteen other tools they might
use.  Standard object names, for instance.  I have had hair pullers when
converting databases from one medium to another.  Every language can
understand a name like 'tblCoolStuff', but, all languages, except the one
originally designed for, puke on the given names like 'Some.Really
Cool^Stuff!'  These problems are simple to avoid.  If you design everything
using a standard set of rules then transferring between different tools is
easy.

,David Norris

World Wide Web - http://www.geocities.com/CapeCanaveral/Lab/1652/
Illusionary Web - http://illusionary.dyn.ml.org/ <-- 02:00 - 10:00 GMT
Video/Audio Phone - callto:illusionary.dyn.ml.org
Page via mail - 412039@pager.mirabilis.com
ICQ Universal Internet Number - 412039
E-Mail - kg9ae@geocities.com
Received on Friday, 28 August 1998 16:52:45 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:15:37 GMT