I understand the sense, because I often want to control the appearance of
the documents I serve, but _this is not a paradox_ because authors specify
the content of their documents.
Clients should have _total_ control over how HTML is rendered. Even if
rendering options were limited (like they are in MacMosaic) there's no way
an author can test his/her HTML on every possible platform. Therefore, the
limitations do nothing but frustrate (me, at least) users who don't like
the hard coded choices.
I think a good graphical browser would define two or three "root" styles --
they would be at least:
"header"
"normal"
All variants would, by default, include the characteristics of their root
style -- that way, I could say "I want all my headers in Helvetica" and not
have to redefine each level; "I want all my body text in palatino" and the
same... You should at least be able to define "character" level styles
(font, size, style...) and maybe even "paragraph" level. (line spacing,
blank line between paragraphs)
My model is the way a good word processor handles "styles" -- MS Word, for
example. If I wanted H1 to be in Zapf Chancery, and H2-5 in Helvetica, I
should be able to choose that without having to set every heading.
Is this just a low priority, or are client authors opposed to giving so
much control to users?
Joe
--
joe germuska * [email protected] * www * res hall net * instruct tech
academic computing & network services * northwestern univ
"Free the people with music..." - Bob Marley