Re: Formatting Options in HTML? (Michael Grobe)
From: (Michael Grobe)
Message-id: <>
Subject: Re: Formatting Options in HTML?
Date: Fri, 14 May 93 14:15:58 CDT
X-Mailer: ELM [version 2.3 PL2]

in regard to the question of supporting centering and right justification
and compact in html (and with apologies to anyone who gets
redundant copies):

i am a new user of html, and an old user of roff.  my first reaction
to html was "what a weak formatting language."  as i thought more
about displaying documents on a wide variety of clients, however, i
began to understand better what markup without formatting was about
and thought "what a clever idea to let clients use whatever capabilities
they have to display what needs to be displayed."  as i prepared
my first document (which is a quick reference to html by the way), i
was frequently frustrated by not being able to get things to look
quite right.  in particular, <dl> didn't seem to want to cooperate
with my aesthetic (nor did <dl compact>.)  finally, i realized that
it was the renderer, not html, that was not cooperating; html provided
a suitably defined <dl> structure, but the renderer was making 
decisions i didn't like.

extrapolating just a bit from this line of experience, i concluded that
an html renderer (or www client) should allow the READER to
define how to render html constructs.  for example, the reader should
be able to define the typeface, indentation, line spacing, and characters
used for bullets (if any) at each level of a nested <ul>. in addition,
the reader should be able to change this rendering while viewing a particular
page.  in this context, the centering and right justification questions
seem to become a matter of reader choice. the reader may choose to have
the client center headers and right justify text, or not.

having prepared documents for years, i have spent a lot of
time trying to pick display formats that convey information clearly.
it is something of a relief to imagine moving the display format
responsibility away from the information provider to the information 
consumer (not the client program, but the reader herself), who should
actually be able to do a better job of it, since each reader knows
or can discover her own preferences (provided, of course, that the
client gives them the capability). 

from this perspective the "compact" of <dl compact> is even something
of an encroachment into the responsibilities of the renderer/reader.

as i mentioned, i am new to html, so this may be old hat to the www
community, but it may also serve to remind veteran users of the
learning process experienced by new users.

:michael grobe
academic computing services
the university of kansas