Re: Frames Are Improving

Matthew James Marnell (
Fri, 23 Aug 1996 15:53:53 -0400

Message-Id: <>
Subject: Re: Frames Are Improving 
In-reply-to: Your message of "Fri, 23 Aug 1996 11:04:39 PDT."
Date: Fri, 23 Aug 1996 15:53:53 -0400
From: Matthew James Marnell <>

:>Matthew James Marnell <> wrote:
:>|Screen size isn't an HTML or an HTTP thing, it is a Style Sheet thing.
:>Screen size is a property of the GUI user agent, just as is the assertion "I
:>accept image/{jpg,gif}."  Asserting that you only can handle a certain
:>resolution of image or feature subset of HTML is appropriate for content
:>The rules for rendering whatever you negotiate are appropriate for style

Okay, assuming that screen size is what is negotiated and documents
are distributed based on screen size.  How much of the pages content
is likely to change?  How much of the presentation is likely to change?
How likely are people to write 10 different documents for 10 different
screen sizes and resolutions?  How likely is that same person to have all
those available to him to view the different documents before serving
them?  How likely are web page designers to start putting up single
pages that state, this page looks best when viewed at X x X resolution
instead of creating multiple pages?  How likely are the non-graphical
browsers to get the short end of the stick because designers don't
consider them to be worth the extra trouble?  How likely am I with
a 21" monitor to devote the entire screen to my web browser?  Once
I have my browser up how likely am I to resize it as need be as I

If you're truly going to negotiate different content, not presentation,
for different screen resolutions who gets the what content?

:>Style is supposed to separate presentation from structure.  How the GUI browser
:>as we now know it handles resizing is a style issue.  But why would you want to
:>prevent a user agent from excluding insane variants (resolutions or dimensions
:>absurd for a given platform) from the set of acceptable variants?

I don't see how this would be the case if it's done in style sheets and
the presentation is dynamically changed by the browser using whatever
internal rules it sees fit.  Right now my browser is set to 788x709
to view a decidedly crappy page that uses absolute pixel widths for tables
instead of relative widths.  Because I default to a decidely small
font, the text is spread way out, and yet if it used relative widths
I'd be able to keep the same width I normally do, and that I'm
comfortable with, the text would be far more readable, to me at least.
Just to give you a hint, it's one of the large browser manufacturers,
and it's not Netscape.

:>1.  All user agents are NOT GUI browsers.

Don't I know it.  I use lynx about 50% of the time.

:>2.  All GUI browsers do NOT run on 24bit 21" monitors.

I don't own a single monitor over 17" and doubt that I'll ever really
need to.  I have 5 monitors on my desk at the moment, giving me views of
3 different OS's from 6 different machines.

:>3.  There will be classes of user agents that we haven't even conceived of yet.

And plenty that have that are being ignored by all but a relative
handful.  We'll have to make sure that robots get a different
negotiated "screen size" and lynx is able to tell the web server
how many rows and columns of characters it's set to, and of course,
what would braille and speech devices do, and all manner of strange
screen configurations for PDA's and Network devices.  In the end,
with "content" negotiated presentation for upwards of 100 different
devices configured to the users specs, what a hell the content
developer will live in.

:>I'm not sure I agree with everything in the mutz draft...just trying to

I realize that, and my comments were not so much directed toward you,
as what I look at as a decidedly backwards draft.