Re: draft-ietf-html-style-00.txt & class as a general selector

Paul Prescod (papresco@calum.csclub.uwaterloo.ca)
Thu, 7 Dec 1995 18:07:03 -0500


At 03:42 PM 12/7/95 -0600, Scott E. Preece wrote:
>Nonsense. It is *less* readable. When you read the text for the
>indicated paragraph you have *NO IDEA* that there is styling magic
>affecting it somewhere else, unless you have memorized the STYLE element
>or refer to it for each paragraph you read.

That is _exactly_ the point. If the paragraph is stylistically affected for
some _reason_ then you should use the CLASS attribute. If it is
stylistically affected for NO reason, then you can use either the STYLE or
ID attributes.

If it is affected for no reason, then you don't really care whether a user
reading the document can figure your reasoning out.

>| As for other advantages, (1) it preserves separation of content and
>| presentation data;

>(1) is an arguable goal (hence the arguing we've been doing).

(1) has always been a goal of the HTML language. If you want a language
that does not support that goal, then you are discussing a different
language. For early references, please see [1] by Tim BL and [2] by Marc A.

>I don't really care about the typing; tools really should take care of
>that. I do care about the author's mental model.

Great. We're finally at the heart of it.

A "poor mental model"? What's wrong with a mental model that holds device
independance as extremely important? And why shouldn't we promote this
mental model? That was the model that HTML was supposed to embody.

>Information,
>including styling information, that is specifically local to an element
>should be with that element.

If the "style" represents "information" it should be a CLASS!! Please
explain why it would be inappropriate for you to make clear to visually
impaired people that you have three examples of class "first example
column", "second example column" and "third example column".

Furthermore, I would suggest that you would be better off to make that a
CLASS, even ignoring our blind friends, because there are other things you
can do with a CLASS that you can't do with directly applied styles.

#1.You can write a sentence like this: The <SPAN CLASS="first example
column">first</SPAN> example is indicative of..... The <SPAN CLASS=second
example column"> is indicative of ...

If you aren't going to discuss these three columns individually anywhere
then why in the world are they different colours??? That can only serve to
confuse!

#2.You can remap a "red" class to gray number 32 everywhere it appears,
because your software does poor grayscaling.

#3.I can read your source to see "why is that column red....oh, just because
it is the first example, not because it represents a deficit. Okay."

If your document source explains your reason for the coloration, then your
reader can configure their software to compensate for its poor display on
their platform. This is the better mental model I want to encourage, and it
is the mental model that HTML (usually) encourages in its design.

Paul Prescod

[1]http://www.w3.org/hypertext/WWW/Provider/Style/DeviceIndependent.html
[2]http://www.ncsa.uiuc.edu/demoweb/html-primer.html