A four-level page model

[Stephen Poley]

Good advice is frequently given in c.i.w.a.* on page structure: matters
such as separation of content from presentation, graceful degradation,
non-dependence on Flash etc. For some while I've wondering whether that
could be transformed into a layered model: something that would help
people to form a mental picture of how a complex page should be put
together.

I've had a go at this:
http://www.xs4all.nl/~sbpoley/webmatters/pagemodel.html

It's not that easy to get the concepts over concisely though. I'd
appreciate feedback, both on the model itself and the way it is
presented. Is it something that could be useful to people trying to
achieve wide accessibility?

--
Stephen Poley

http://www.xs4all.nl/~sbpoley/webmatters/

[Karl Smith]

Stephen Poley wrote:

Good advice is frequently given in c.i.w.a.* on page structure: matters
such as separation of content from presentation, graceful degradation,
non-dependence on Flash etc. For some while I've wondering whether that
could be transformed into a layered model: something that would help
people to form a mental picture of how a complex page should be put
together.
Data layer -->
interpretation (design) layer -->
prettification (de-zyner) layer

I've had a go at this:
http://www.xs4all.nl/~sbpoley/webmatters/pagemodel.html
I think you're too heavily influenced by the CIWAHians and their
constant put down of CSS as "mere presentation". Presentation is about
interpreting data as well as prettifying websites.

Here's one the author could have done without Java, with a very big
and messy HTML table and some clever CSS:

http://physci.org/applet/PToEApplet.htm

And here's one I could have done without XML & CSS2, by throwing a lot
of class="..." at HTML (but that would still *need* to use CSS1), or
by using lots of FONTs:

http://users.tpg.com.au/karl6740/stevewaugh/CSS-EM.xml

Explain how these pages would be better if their data layers were in
HTML, not Java or XML?

It's not that easy to get the concepts over concisely though. I'd
appreciate feedback, both on the model itself and the way it is
presented. Is it something that could be useful to people trying to
achieve wide accessibility?

[Andrew Thompson]

Karl Smith wrote:

Stephen Poley wrote:

Good advice is frequently given in c.i.w.a.* on page structure:

.... Here's one the author could have done without Java, with a very big
and messy HTML table and some clever CSS:

http://physci.org/applet/PToEApplet.htm

(chuckles) Since you mention it.
http://www.1point1c.org/chemistry/

--
Andrew Thompson
* http://www.PhySci.org/ Open-source software suite
* http://www.PhySci.org/codes/ Web & IT Help
* http://www.1point1C.org/ Science & Technology

[Andrew Thompson]

Andrew Thompson wrote:

Karl Smith wrote:

Stephen Poley wrote:

Good advice is frequently given in c.i.w.a.* on page structure:

...

Here's one the author could have done without Java, with a very big
and messy HTML table and some clever CSS:

http://physci.org/applet/PToEApplet.htm

(chuckles) Since you mention it.
http://www.1point1c.org/chemistry/

OH, and I should point out that whatever
cleverness may be in the CSS is due to a
thread from c.i.w.a.s
http://groups.google.com/groups?th=b43a8df8a99412bd
[ though all faults and any lack of
validation I take complete credit for.. ]

--
Andrew Thompson
* http://www.PhySci.org/ Open-source software suite
* http://www.PhySci.org/codes/ Web & IT Help
* http://www.1point1C.org/ Science & Technology

[Alan J. Flavell]

On Mon, 8 Feb 2004, Karl Smith wrote:

I think you're too heavily influenced by the CIWAHians and their
constant put down of CSS as "mere presentation".
I think that's putting an unfair slant on the issue. I don't know
anyone who objects to a visually attractive presentation, but the
usual counsel it not to make interpretation of the content solely
dependent on the CSS, in other words to treat the CSS as maybe
desirable, even highly desirable - but in the ultimate analysis:
optional.
Presentation is about
interpreting data as well as prettifying websites.
Then you'll need to provide some alternative "interpretation" of the
data for the situations where the CSS is inapplicable.
Here's one the author could have done without Java,

Hang on - how did we suddenly jump ship from CSS to Java?

There are some kinds of material that simply cannot be done in HTML,
with or without CSS. It would be unrealistic to pretend otherwise. On
the one hand that's no excuse for making stuff critically dependent on
Java (or Flash, whatever) when they _could_ equally well have been
done in HTML; but on the other hand that's no reason to try to outlaw
Java etc. (for consenting readers) when it's an appropriate solution
(provided, of course, that one fulfils whatever the local requirements
may be for providing an accessible alternative).

[Karl Smith]

> Stephen Poley wrote:

... a layered model: something that would help
people to form a mental picture of how a complex page should be put
together.

I've had a go at this:
http://www.xs4all.nl/~sbpoley/webmatters/pagemodel.html

Here's the annotated version:
http://users.tpg.com.au/karl6740/temp/pagemodel.html
There should be enough abrasive comments there for you.

--
Karl Smith.

[Karl Smith]

"Andrew Thompson" <Se********@www.invalid> wrote:

Karl Smith wrote:

...

Here's one the author could have done without Java, with a very big
and messy HTML table and some clever CSS:

http://physci.org/applet/PToEApplet.htm

(chuckles) Since you mention it.
http://www.1point1c.org/chemistry/

OH, and I should point out that whatever
cleverness may be in the CSS is due to a
thread from c.i.w.a.s
http://groups.google.com/groups?th=b43a8df8a99412bd

Never saw that at the time, I've been away from this NG for a while :-)
I much prefer the Java version (and I hate Java).

[Andrew Thompson]

Karl Smith wrote:

"Andrew Thompson" <Se********@www.invalid> wrote:

Karl Smith wrote: ...

Here's one the author could have done without Java, with a very big
and messy HTML table and some clever CSS:

http://physci.org/applet/PToEApplet.htm

(chuckles) Since you mention it.
http://www.1point1c.org/chemistry/

.... ...I much prefer the Java version (and I hate Java).

That indicates to me that you _detest_ the
HTML version (I am in the process of hiding
most of my Java behind the scenes on the
server in favor of HTML/CSS where practical).

So, is it the 'messy HTML', or the (non?) clever
CSS you dislike more?

[ ...I suppose you cannot please all
the people all the time. ;-) ]

--
Andrew Thompson
* http://www.PhySci.org/ Open-source software suite
* http://www.PhySci.org/codes/ Web & IT Help
* http://www.1point1C.org/ Science & Technology

[Stephen Poley]

On 9 Feb 2004 05:31:35 -0800, go************@kjsmith.com (Karl Smith)
wrote:

Here's the annotated version:
http://users.tpg.com.au/karl6740/temp/pagemodel.html
There should be enough abrasive comments there for you.

Yes, I get that impression! I'll mull over them for a while.

--
Stephen Poley

http://www.xs4all.nl/~sbpoley/webmatters/

[Firas D.]

Karl Smith wrote:

Stephen Poley wrote:

... a layered model: something that would help
people to form a mental picture of how a complex page should be put
together.

I've had a go at this:
http://www.xs4all.nl/~sbpoley/webmatters/pagemodel.html

Here's the annotated version:
http://users.tpg.com.au/karl6740/temp/pagemodel.html
There should be enough abrasive comments there for you.

When did Nielsen say that scrolling is evil? Must be a loong time ago,
if ever.

[Eric Bohlman]

"Firas D." <fd********@firasd.org> wrote in news:c0b5q3$13bm5l$1@ID-
214165.news.uni-berlin.de:

When did Nielsen say that scrolling is evil? Must be a loong time ago,
if ever.

He's got a very annoying habit of sometimes expressing himself in sound
bites that give a misleading impression of what he really has to say in the
details. Maybe he feels it's the only way to get the attention of PWPH
(People With Pointy Hair). He's well known for saying "users don't scroll"
but it's clear from his actual writing that what he means is that a user
who comes to a page and doesn't immediately see what he's looking for is
unlikely to scroll down in search of it, *not* that he won't scroll to
continue reading something he's interested in. He also says "users don't
read" but again his actual writing indicates that he means that users won't
read through lengthy prose unless they *already know* that it's what they
were looking for. But a lot of people, particularly of the pointy-haired
sort, just remember the sound bites and take them literally.

In the case of magazine-article type sites, I suspect that the chief
motivation for breaking the article down into little pieces is to maximize
the number of ads the user sees while reading the article.

[Barry Pearson]

Stephen Poley wrote:

Good advice is frequently given in c.i.w.a.* on page structure:
matters such as separation of content from presentation, graceful
degradation, non-dependence on Flash etc. For some while I've
wondering whether that could be transformed into a layered model:
something that would help people to form a mental picture of how a
complex page should be put together.

I've had a go at this:
http://www.xs4all.nl/~sbpoley/webmatters/pagemodel.html

It's not that easy to get the concepts over concisely though. I'd
appreciate feedback, both on the model itself and the way it is
presented. Is it something that could be useful to people trying to
achieve wide accessibility?

Your article was one of the most thought-provoking I've seen for some time.
Devising such a model is an excellent idea. (I wish I had thought of it!)

One thought was that it lacked a context. We probably feel we know what the
context is, but it might be worth stating it. Let's see if I can express it.

A web site exists (presumably) to provide the user with a set of valuable
transactions/interactions. So, what is seen on the screen, or rendered some
other way, is the *result* of one transaction/interaction, and provides a *set
of opportunities* for more transactions/interactions.

Perhaps everything you talk about can be judged in these terms. What is the
value to the user of any particular transaction/interaction? What is the value
of the opportunities presented to the user for further
transactions/interactions? How well are those opportunities for further
transactions/interactions conveyed to the user?

This means that everything seen by the user must convince the user that it was
worth doing the last transaction/interaction, because of the value. And that
it is worth staying with this web site because of the value of future
transactions/interactions. So if there are few further
transactions/interactions on the page, the immediate value must be high. And
if there is no immediate value on the page, the conveyed value for further
transactions/interactions must be high. The site contents, the site
navigation, and the individual pages, must all be designed to maximise those
factors for all of the target audience.

All the technologies used (Flash, frames, images, animated GIFs, Javascript,
and all the rest), and other factors (layout & presentation, etc) should
contribute to those values. Any that don't - for the target audience - should
be abandoned. Does "X" make a user feel that s/he was right to link to that
page? Does "X" help a user link to other pages? If not, what is it there for?

The elements of your model need to fit into the above. One way is to increase
the value. (All sorts of layout & presentation factors increase the value).
Another way is to reduce the cost of delivering the value. (Effective use of
technology can do this). So anything that improves the immediate value, or the
opportunities for more links, or reduces the cost of supplying these, is good.
That might be where "separation of content & presentation" fits -
cost-reduction. While style-enhancement increases value.

I'll have to think more about this. But here are 4 pages I've written about
the subject "content versus presentation". I don't believe that this is really
understood, nor that we have the technology we need.

"Separation of concerns"
http://www.barry.pearson.name/articl..._presentation/

--
Barry Pearson
http://www.Barry.Pearson.name/photography/
http://www.BirdsAndAnimals.info/
http://www.ChildSupportAnalysis.co.uk/

[Stephen Poley]

On 9 Feb 2004 05:31:35 -0800, go************@kjsmith.com (Karl Smith)
wrote:

Stephen Poley wrote:

> ... a layered model: something that would help
> people to form a mental picture of how a complex page should be put
> together.
>
> I've had a go at this:
> http://www.xs4all.nl/~sbpoley/webmatters/pagemodel.html

Here's the annotated version:
http://users.tpg.com.au/karl6740/temp/pagemodel.html
There should be enough abrasive comments there for you.

Well, I think the abrasions have healed sufficiently to return to this.
(Actually I've just been distracted by other things.) A few reactions
here:

http://www.xs4all.nl/~sbpoley/misc/karlspagemodel.html
--
Stephen Poley

http://www.xs4all.nl/~sbpoley/webmatters/

[Stephen Poley]

Coming back to this rather late ...

On Tue, 10 Feb 2004 20:33:12 -0000, "Barry Pearson"
<ne**@childsupportanalysis.co.uk> wrote:

Stephen Poley wrote:

Good advice is frequently given in c.i.w.a.* on page structure:
matters such as separation of content from presentation, graceful
degradation, non-dependence on Flash etc. For some while I've
wondering whether that could be transformed into a layered model:
something that would help people to form a mental picture of how a
complex page should be put together.

I've had a go at this:
http://www.xs4all.nl/~sbpoley/webmatters/pagemodel.html

Your article was one of the most thought-provoking I've seen for some time.
Devising such a model is an excellent idea. (I wish I had thought of it!)
Thanks. :-)
One thought was that it lacked a context. We probably feel we know what the
context is, but it might be worth stating it. Let's see if I can express it.

A web site exists (presumably) to provide the user with a set of valuable
transactions/interactions. So, what is seen on the screen, or rendered some
other way, is the *result* of one transaction/interaction, and provides a *set
of opportunities* for more transactions/interactions.

<snip>

I've been wondering whether to try to incorporate some of this, but at
present I don't feel quite comfortable with it, though I think I see
what you're getting at. Perhaps it should be the subject of another
page. I've saved your post for future reference anyway.

--
Stephen Poley

http://www.xs4all.nl/~sbpoley/webmatters/