Re: color text?

H&kon W Lie (howcome@w3.org)
Wed, 3 May 95 09:58:46 EDT

Lou Montulli writes:

> I fully support the idea of allowing sytle information as attributes.
> It gives us progressive layout as well as backwards compatibility.
>
> It is the best of all worlds. We should definately move forward
> and implement this...

Your programmers are probably finished by now, but one should think
and experiment a bit around these issues before deploying them. If
you're really concerned with the reader not having to wait before
content shows up on the screen, why aren't you eagerly defending the
LINK way of referring to a style sheet? If the browser doesn't happen
to have the style sheet cached (I think it will more often than not),
it can fetch it after displaying the content.

On the part of the STYLE attribute: yes, it may potentially get big
(just like META, KEYWORD and whatever), but on average it will be more
compact than per-element definitions (or, per-element overrides in
case you allow for the propagation of styles).

Consider this:

<HEAD>
<TITLE>title</TITLE>
<STYLE notation=css>
h1, h2, h3: font.family = helvetica; font.style = bold; font.color = #000
</STYLE>
</HEAD>
<BODY>
<H1>H1</H1>
<H2>H2</H2>
<H3>H3</H3>
</BODY>

size: 191 bytes

Versus:

<HEAD>
<TITLE>title</TITLE>
</HEAD>
<BODY>
<H1 font.family = helvetica; font.style = bold; font.color = #000>H1</H1>
<H2 font.family = helvetica; font.style = bold; font.color = #000>H2</H2>
<H3 font.family = helvetica; font.style = bold; font.color = #000>H3</H3>
</BODY>

size: 273 bytes

Cheers,

-h&kon

Hakon W Lie, WWW project CERN, CH-1211 Geneva 23
http://www.w3.org/hypertext/WWW/People/howcome/