Re: More thoughts

Brian Behlendorf ([email protected])
Wed, 26 Oct 1994 19:44:21 -0700 (PDT)


On Thu, 27 Oct 1994 [email protected] wrote:
> Reading the discussion here, another thought just occured to me: You guys
> are all really good HTML "coders". Showing our products at trade shows, I
> have talked to alot of people who want to create HTML without being
> "coders". This means that our products are going to allow people to create
> HTML without typing a single <h1>.

What is a "coder"? My grandmother was great at taking old recipes and
"hacking" them by adding spices here and there, so I'd call her a "coder"
even though she never touched a computer in her life. I had several
English schoolteachers who were very strict in enforcing writing style and
proper punctuation who also never used computers. It sounds like you're
suggesting that the importance of semantic text is an ivory-tower notion
with no real application or significance in the real world. Several
other posts here earlier today are basically echos of this position.

I've taught many *many* nontechnical people how to use HTML. All seemed
perfectly capable of understanding the importance of semantic markup when
explained to them. They also understand that we're not saying
"presentation is unimportant", it's just that there are no widely
deployed efficient standardized solutions to this. People are trying,
but it's a hard problem, and if it's of utmost concern one can put up
PostScript, Acrobat, even MSWord files.

We are *deluding* ourselves if we think our presentational needs can be
completely fulfilled just by adding tags to HTML that are limited to being
backwards compatible with HTML 2.0 browsers. The HTML 3.0 proposal even
concedes this. Making sure old browsers can read our new documents
shouldn't be a priority, as long as 1) new browsers can read old docs, and
2) some form of down-grading script could be run to turn valid HTML 3.0
docs into valid HTML 2.0 docs. And has been said here before, with a real
SGML engine in the browser, this debate is moot. (Oh yeah, addendum to my
post on Hyper-G - it has a built-in SGML engine.)

> This also means that naive users are
> going to think that whatever they can do to make their document look right
> on the screen is fine. WYSIWYG. I've seen some discussion here about what
> people should and shouldn't do with various types of formatting, but lets
> just assume for a moment that people creating these documents are not going
> to realize the difference.

Implementation decisions based on the self-proscribed needs of
self-admittedly "naive users" by those Who Should Know Better is
inexcusable.

Brian