Re: Portrait vs. Landscape (was Re: THEAD & TFOOT for columns)

E. Stephen Mack (
Fri, 15 Aug 1997 17:19:35 -0700

Message-Id: <>
Date: Fri, 15 Aug 1997 17:19:35 -0700
From: "E. Stephen Mack" <>
In-Reply-To: <v0310280db01a8fb259c8@[]>
Subject: Re: Portrait vs. Landscape (was Re: THEAD & TFOOT for columns)

The so-called "portrait vs. landscape" scrolling issue is not just
about visual presentation.  It invokes larger issues of the context
of information on a page.

However, the discussion currently applies only to visual user agents.
(But it may be applied to auditory browsers whenever there is
information that is presented in parallel: for example, two
people speaking at once, or music playing in a background).

Walter Ian Kaye <> wrote:
>In a mode where
>the document were to wrap vertically and scroll horizontally, how would
>one read lines of text? It just doesn't work that way (as far as humans
>reading is concerned), thus the text wraps horizontally and scrolls
>vertically. This is also the way all Word Processors behave -- the words
>are wrapped horizontally and the document scrolls vertically. This was
>not a paradigm invented just for HTML -- it's simply the way *we* read.

It doesn't have to be so.  Text doesn't have to wrap at all --
it can just extend horizontally.  That allows different ideas
to be presented within the vertical space.  For example, Wired
magazine presents a feature in the front of their magazine where a few
lines in the middle scroll "horizontally" from page to page.  
(Annnoyingly so, I must admit.)

Computer screens, as Peter Flynn pointed out, are typically
"landscape" in that they are wider than they are tall, while
pieces of paper are typically "portrait" by default.

Consider a screen with some information (apologies for ASCII art):
/               \
| Information 1 |
|               |
| Information 2 |

Normally we think of the screen as a porthole of a longer vertical
page, with a vertical scroll bar to move us "up" and "down" on the
|                     |
|   Information 1     |
|  _________________  |
| /               |^\ |
| | Information 2 |X| |
| |               |X| |
| | Information 3 |V| |
| \_______________|_/ |
|                     |

But there's no reason WHY information can't be presented in
parallel, horizontally scrolled:

|                 _______________                 |
|                /               \                |
| Information 1  | Information 2 |  Information 3 |
|                |               |                |
| Information 4  | Information 5 |  Information 6 |
|                |_______________|                |
|                |<XXXXXXXXXXXXX>|                |
|                \_______________/                |

Regardless of personal opinions regarding the effectivness
or desirability of this way of presenting information, there
are times when it may be necessary or desirable.

(Currently, this could be accomplished using tables or
a PRE element with a lot of spaces.  Neither recourse
is particularly desirable.)

HTML already has plenty of implicit assumptions regarding the
the presentation of information.  Many of these assumptions
are in the areas of elements that depend on visual presentation,
such as the deprecated formatting elements.  But while we
are thinking about what HTML elements are visual and explicitly
oriented, we should consider the nature of a "paragraph" of
information.  The implicit assumption of that paragraph is
that it is related to the previous paragraphs in a vertical

I believe that it's vital to expose the assumptions regarding the
presentation of paragraphs, and see what options are available to
HTML authors for changing implicit models.

I also think it's important to expose all of HTML's assumptions and
question whether or not other models are possible, desirable, and

(Probably it's more important to concentrate on the concrete
elements of the draft HTML 4.0 so that the W3C can confidently
move it forward towards being a recommendation, though.  But I've
always been interested in horizontally-scrolling pages so I find
this discussion area to be of interest.)
E. Stephen Mack <>