How to detect table width or height?

[Mikko Rantalainen]

Re: How to detect table width or height?

Dennis wrote:[color=blue]
> On Mon, 20 Oct 2003 12:46:51 +1000, Mark Parnell
> <webmaster@clar kecomputers.com .au> wrote:
>[color=green][color=darkred]
>>>Without Flash (or a more mature CSS) I have to aim at the
>>>middle of the bell curve (what, 800 by 600?) and let automatic
>>>scrolling and huge empty borders do the rest.[/color]
>>
>>Not at all. The web is fluid by nature. If you don't specify fixed sizes,
>>then the page will flow automatically.[/color]
>
> While that is true, you must admit that most all the top websites opt
> for an "800-wide non-horizontal flowing" format (typically aligned[/color]

It's because "others are doing it too". I think the 800 pixels wide rule
comes from early Windows versions often running 800x600 resolution on
low end hardware and with such a small screen, the only sane thing is to
always maximize the browser window.

The 800 pixels wide rule was OK when the "only" way to access it was
with a desktop computer but nowadays we have much more diversity in the
devices used to access the net. We have pretty much anything from 1 bit
101x86 pixel screen (some cellurar phone with XHTML browser) to
2000x1500 pixel stereo display with 30 bit color (high end 3D
workstation) and *the same* document or site should adjust to both
displays. This is not going to happen, if you have *any* width defined
in pixels. Highly probably the layout is going to fail on smaller
screens if you have anything big side-by-side.

There's some CSS based solutions in the future (you can apply rules
based on viewport size, for example) but we aren't there yet.

Another reason for the 800 pixels wide rule is that some page authors
haven't yet understood that design for a fixed size page (paper) isn't
the same thing as designing for the web. For a paper, you can fix some
things relative to each other and to the whole page. For the web, any
assumptions about the page "size" can prove wrong.

If your page is usable with a single line braille display, I guess it's
usable with anything. Sometimes you have to make a compromise and target
to something "better" than a single line braille display but even in
that case you can do much better that using the 800 rule.

--
Mikko

[Jim Ley]

Re: How to detect table width or height?

On Mon, 27 Oct 2003 19:58:22 +0200, Mikko Rantalainen <mira@st.jyu.fi >
wrote:
[color=blue]
>Dennis wrote:
>We have pretty much anything from 1 bit
>101x86 pixel screen (some cellurar phone with XHTML browser) to
>2000x1500 pixel stereo display with 30 bit color (high end 3D
>workstation) and *the same* document or site should adjust to both
>displays.[/color]

I simply do not agree with such a blanket statement, reading across
more than 800 pixels or so I find impossible (regardless of text size
it's the actual scanning of the eye which is my problem), it's just
too difficult, a maximum width is very important to the usability of
your content IMO.

This is the reason I feel most people suggest widths, small screens do
not suffer as they generally ignore widths anyway, the only people to
suffer is the people with a slightly smaller than expected width
available to them - hence the recommendations on size.
[color=blue]
>If your page is usable with a single line braille display, I guess it's
>usable with anything.[/color]

Then you would guess incorrectly.

Jim.
--
comp.lang.javas cript FAQ - http://jibbering.com/faq/

[Steve Pugh]

Re: How to detect table width or height?

jim@jibbering.c om (Jim Ley) wrote:[color=blue]
>On Mon, 27 Oct 2003 19:58:22 +0200, Mikko Rantalainen <mira@st.jyu.fi >
>wrote:[color=green]
>>Dennis wrote:
>>We have pretty much anything from 1 bit
>>101x86 pixel screen (some cellurar phone with XHTML browser) to
>>2000x1500 pixel stereo display with 30 bit color (high end 3D
>>workstation ) and *the same* document or site should adjust to both
>>displays.[/color]
>
>I simply do not agree with such a blanket statement, reading across
>more than 800 pixels or so I find impossible (regardless of text size
>it's the actual scanning of the eye which is my problem),[/color]

As you say, it's your problem. Other people may experience things
differently.

You choose to have your browser window at a certain width.

If the scanning of the eye (i.e. the angle through which it must move
to see both end of a line of text) is the problem then that doesn't
depend simply on the number of pixels, rather it depends on the
physical size of the monitor and the resolution (in pixels per inch
terms) and the distance from monitor to eyeball.

An author has no way of knowing any of those variables.

If I lean back in my chair I can read long lines much more easily. How
is a web site to react to that and make it's lines longer? It can't
but I can increase the size of my browser window and a fluid website
will react accordingly.

max-width in a user stylesheet would overcome your problem as you
describe it above.

Or indeed simply setting your browser window to be the widthj you find
most comfortable.
[color=blue]
>This is the reason I feel most people suggest widths,[/color]

I feel that "most people" design their web sites in Photoshop or
whatever and then translate a static image into a static web page.

Steve

--
"My theories appal you, my heresies outrage you,
I never answer letters and you don't like my tie." - The Doctor

Steve Pugh <steve@pugh.net > <http://steve.pugh.net/>

[Jim Ley]

Re: How to detect table width or height?

On Mon, 27 Oct 2003 19:40:06 +0000, Steve Pugh <steve@pugh.net > wrote:
[color=blue]
>jim@jibbering. com (Jim Ley) wrote:[color=green]
>>On Mon, 27 Oct 2003 19:58:22 +0200, Mikko Rantalainen <mira@st.jyu.fi >
>>wrote:
>>I simply do not agree with such a blanket statement, reading across
>>more than 800 pixels or so I find impossible (regardless of text size
>>it's the actual scanning of the eye which is my problem),[/color]
>
>As you say, it's your problem. Other people may experience things
>differently.[/color]

Sure, but I believe it's common enough for authors to choose it as a
default suggestion.
[color=blue]
>If the scanning of the eye [...] and the distance from monitor to eyeball.[/color]

I understoof the pixel was defined with reference to the distance from
the eyeball, please correct me if I'm wrong.

Jim.
--
comp.lang.javas cript FAQ - http://jibbering.com/faq/

[Christoph Paeper]

Re: How to detect table width or height?

*Jim Ley* <jim@jibbering. com>:[color=blue]
> On Mon, 27 Oct 2003 19:58:22 +0200, Mikko Rantalainen <mira@st.jyu.fi >
>
> reading across more than 800 pixels or so I find impossible
> (regardless of text size[/color]

You're assuming a 43cm (17in) display with 1024px * 768px here, right?
(With real world, not theoretical CSS pixels that is.)
[color=blue]
> a maximum width is very important to the usability of
> your content IMO.[/color]

Yes, it aids more than it harms, IMHO. The value should be selected quite
large, though, maybe 45-50em.
[color=blue]
> This is the reason I feel most people suggest widths,[/color]

No, they do to remain in control of the presentation, which they never were
anyway.

--
"Not only does God play dice with the universe,
but sometimes he throws them where they cannot be seen."
Stephen Hawking

[Steve Pugh]

Re: How to detect table width or height?

jim@jibbering.c om (Jim Ley) wrote:[color=blue]
>On Mon, 27 Oct 2003 19:40:06 +0000, Steve Pugh <steve@pugh.net > wrote:
>[color=green]
>>If the scanning of the eye [...] and the distance from monitor to eyeball.[/color]
>
>I understoof the pixel was defined with reference to the distance from
>the eyeball, please correct me if I'm wrong.[/color]

That's how it's defined in the CSS spec. But no browser that actually
uses that definition has ever been sighted, much less caught and
brought back for study.

Steve

--
"My theories appal you, my heresies outrage you,
I never answer letters and you don't like my tie." - The Doctor

Steve Pugh <steve@pugh.net > <http://steve.pugh.net/>

[Barry Pearson]

Re: How to detect table width or height?

Jim Ley wrote:[color=blue]
> On Mon, 27 Oct 2003 19:58:22 +0200, Mikko Rantalainen <mira@st.jyu.fi >
> wrote:
>[color=green]
>>Dennis wrote:
>>We have pretty much anything from 1 bit
>>101x86 pixel screen (some cellurar phone with XHTML browser) to
>>2000x1500 pixel stereo display with 30 bit color (high end 3D
>>workstation ) and *the same* document or site should adjust to both
>>displays.[/color]
>
> I simply do not agree with such a blanket statement, reading across
> more than 800 pixels or so I find impossible (regardless of text size
> it's the actual scanning of the eye which is my problem), it's just
> too difficult, a maximum width is very important to the usability of
> your content IMO.[/color]
[snip]

I agree, but I also believe that there is much more to this than the total
size of the display.

Content often has inherent size, too. A GIF, a PNG, a JPEG, or a form for
input, is often designed for a particular width. The designers of the web site
had a vision in mind, in which this element occupied X% of the width, and that
element occupied Y% of the width. People who can only change the HTML & CSS
are ultimately limited.

The designers envisioned the overall viewport. They had a view of some X-box
Y-column layout. They said "let's have this form fitting in here; let's have
this image fitting in there". So people deliver such content to be authored.
And authoring has to make do with what it has.

If authoring allows the viewport width to dramatically change without
horizontal scrolling, the relationship between the inherent sizes of that
content and the viewport width can become silly. Perhaps that form designed to
occupy X% of the width of the viewport suddenly occupies 110% of it, or 1% of
it.

In order to accommodate support of very different viewport widths, we also
need much more scalability of content. SVG to replace lots of image content.
Perhaps JPEG2000 to support much more scability of photographic content. Etc.

Text is inherently scalable. Very little other content is. The page design is
typically constrained by its unscalable content. It makes sense to design for
a certain size, so that all the unscalable content fits into it. And 800 x 600
is the obvious choice at the moment. But in 5 years time ...?

--
Barry Pearson

http://www.Barry.Pearson.name/photography/

http://www.BirdsAndAnimals.info/

http://www.ChildSupportAnalysis.co.uk/

[I V]

Re: How to detect table width or height?

On Mon, 27 Oct 2003 22:39:02 +0000, Steve Pugh wrote:
[color=blue]
> jim@jibbering.c om (Jim Ley) wrote:[color=green]
>>On Mon, 27 Oct 2003 19:40:06 +0000, Steve Pugh <steve@pugh.net > wrote:
>>[color=darkred]
>>>If the scanning of the eye [...] and the distance from monitor to eyeball.[/color]
>>
>>I understoof the pixel was defined with reference to the distance from
>>the eyeball, please correct me if I'm wrong.[/color]
>
> That's how it's defined in the CSS spec. But no browser that actually
> uses that definition has ever been sighted, much less caught and
> brought back for study.[/color]

Well, yes and no. The CSS spec says that, "If the pixel density of the
output device is very different from that of a typical computer display,
the user agent should rescale pixel values," and then gives an explanation
of a 'reference pixel', which is presumably intended to aid anyone
implementing such a rescaling. It _doesn't_ say, at least as I interpret
it, that a CSS UA has to interpret pixel lengths precisely using this
reference pixel. So, a browser intended for desktop use which uses screen
pixels as CSS pixels, regardless of the DPI of the screen, is, I think,
doing what the CSS spec says it should.

--
"- Penny, I worry that you are loosing heart... You are not the sweet little
girl I once knew. Where's your sense of wonder?
- Currently flowing into a sanitary napkin... Guess where my childlike
innocence and idle dreams are currently wedged. Come on, I dare you."

http://www.huh.34sp.com/

[kchayka]

Re: How to detect table width or height?

Jim Ley wrote:[color=blue]
>
> I simply do not agree with such a blanket statement, reading across
> more than 800 pixels or so I find impossible (regardless of text size
> it's the actual scanning of the eye which is my problem)[/color]

The problem is the number of characters per line, not the number of
pixels. I need a larger than average text size and an 800px width
restriction often _causes_ readability problems.

Something you don't mention is that many of these sites that used a
fixed width layout also use multiple columns, plus they tend to set
microfonts. After I override their silly font sizes so I can read it,
there ends up being only a couple words per line in a column. IMO, this
is a far worse readability problem than lines that are a bit long.

With a fluid layout, at least I can adjust my window width to help
readability when I need to. Can't do that with a fixed design,
especially a table-based layout, which many of these sites are.
[color=blue]
> a maximum width is very important to the usability of
> your content IMO.[/color]

OK fine, but if that content is text, set a max width in em or ex units,
not a fixed with in pixels. Use the right tool for the job, and all
that jazz.

--
To email a reply, remove (dash)un(dash). Mail sent to the un
address is considered spam and automatically deleted.

[kchayka]

Re: How to detect table width or height?

Dennis wrote:
[color=blue]
> On Tue, 21 Oct 2003 14:08:33 -0500, kchayka <kcha-un-yka@sihope.com>
> wrote:
>[color=green]
>>With HTML, I have control over text size. I can make text whatever size
>>it needs to be for me to read it. Flash can't do that.[/color]
>
> You can put your own font buttons on your flash movie and change the
> font size that way. Admitedly, it would be slicker if the user's
> existing browser settings would do that automatically.[/color]

Ah, I see now that you are going for "slick". Usability and
accessibility normally take a back seat in these cases. I have never
seen a Flash site that had any kind of font user control, probably
because if the author really cared about this, I doubt they would have
used Flash in the first place.
[color=blue]
> But just out of curiosity, what do you do about graphics?[/color]

I could care less about eye-candy. Photos aren't normally a problem
unless the dimensions are tiny. Graphics that masquerade as text are
the only real issue. Turning off image loading helps only if the author
uses appropriate alt text. Many do not. In these cases, the browser's
back button works quite nicely. ;)

--
To email a reply, remove (dash)un(dash). Mail sent to the un
address is considered spam and automatically deleted.

[kchayka]

Re: How to detect table width or height?

Jim Ley wrote:[color=blue]
>[color=green][color=darkred]
>>>I simply do not agree with such a blanket statement, reading across
>>>more than 800 pixels or so I find impossible (regardless of text size
>>>it's the actual scanning of the eye which is my problem),[/color][/color]
>
> I believe it's common enough for authors to choose it as a
> default suggestion.[/color]

If it were simply a suggestion, it wouldn't be so bad, but the huge
majority of these 800px wide sites use table layouts, often nested
tables. Few browsers let the user disable table support and attempts to
get around table width issues via user stylesheets generally don't work
well.

BTW, we all know that fixing font-size falls under the evil heading,
right? So what makes fixing page width OK?

--
To email a reply, remove (dash)un(dash). Mail sent to the un
address is considered spam and automatically deleted.

[Jim Ley]

Re: How to detect table width or height?

On Tue, 28 Oct 2003 07:56:46 -0600, kchayka <kcha-un-yka@sihope.com>
wrote:
[color=blue]
>Jim Ley wrote:[color=green]
>>
>> I believe it's common enough for authors to choose it as a
>> default suggestion.[/color]
>
>If it were simply a suggestion, it wouldn't be so bad, but the huge
>majority of these 800px wide sites use table layouts, often nested
>tables.[/color]

That's still a suggestion, we can do nothing but suggest on the web.
Sure user agents are poor in this area (like they're poor at other
things) but that does not, nor should it invalidate the idea of
suggesting widths.

If we don't suggest anything that some UA's are poor at - we're left
with nothing to use at all, we certainly can't use CSS which user
agents are almost entirely atrocious at.

Jim.
--
comp.lang.javas cript FAQ - http://jibbering.com/faq/

[kchayka]

Re: How to detect table width or height?

Dennis wrote:
[color=blue]
> On Mon, 20 Oct 2003 09:27:10 -0500, kchayka <kcha-un-yka@sihope.com>
> wrote:[color=green]
>>
>>So what are the relevant commands that tell you what the user's
>>preferred text size is?[/color]
>
> But are you familiar with the zoom feature in Flash?[/color]

Yes, and it has serious flaws. I have submitted bug reports to
Macromedia, but there have been no improvements, nor do I really expect
any. Seems it's not really important to them.
[color=blue]
> With it you can make EVERYTHING larger, not just text.[/color]

And why would I need to zoom everything? Are your eye-candy images
equally as important as the textual content? Or is maintaining your
pixel-perfect layout more important to you than my ability to read it?
Page zoom, like Opera does, has more drawbacks than benefits to me, and
the Flash player's zoom has even more drawbacks than Opera's. I'll take
just plain text zoom any day.
[color=blue]
> So images (which aren't affected by
> the user's preferred text size) also get magnified for the user's
> benefit. It seems like kind of a wash in my opinion.[/color]

It is not a wash by a long shot. If you ever _had_ to use it, you might
agree.

--
To email a reply, remove (dash)un(dash). Mail sent to the un
address is considered spam and automatically deleted.

[Jim Ley]

Re: How to detect table width or height?

On Tue, 28 Oct 2003 08:20:56 -0600, kchayka <kcha-un-yka@sihope.com>
wrote:
[color=blue]
>Dennis wrote:[color=green]
>> With it you can make EVERYTHING larger, not just text.[/color]
>
>And why would I need to zoom everything?[/color]

because Flash like other graphics formats uses position for
information, just changing the size of one part of the text could
dramatically change the meaning of the document. (e.g. in SVG where
you stupidly can override text size in CSS
http://jibbering.com/2002/8/text-mixup.svg )
[color=blue]
>Are your eye-candy images
>equally as important as the textual content?[/color]

It's rather ridiculous to suggest that people only use vector (or
other) graphics for eye candy, there are a lot of things you simply
cannot do in HTML, that you need to use graphics for.

Jim.
--
comp.lang.javas cript FAQ - http://jibbering.com/faq/

[kchayka]

Re: How to detect table width or height?

Jim Ley wrote:
[color=blue]
> On Tue, 28 Oct 2003 08:20:56 -0600, kchayka <kcha-un-yka@sihope.com>
> wrote:
>[color=green]
>>Dennis wrote:[color=darkred]
>>> With it you can make EVERYTHING larger, not just text.[/color]
>>
>>And why would I need to zoom everything?[/color]
>
> because Flash like other graphics formats uses position for
> information,[/color]

You assume the author is using Flash appropriately. This is usually
not the case. So far, the OP hasn't said anything that leads me to
believe he will be using it for a suitable reason, either. More like he
wants total control over layout, which Flash will definitely give him.
[color=blue][color=green]
>>Are your eye-candy images
>>equally as important as the textual content?[/color]
>
> It's rather ridiculous to suggest that people only use vector (or
> other) graphics for eye candy, there are a lot of things you simply
> cannot do in HTML, that you need to use graphics for.[/color]

I'll agree that there are some things that cannot be done (well) in
HTML, but the overwhelming majority of Flash sites do not fall into this
category. Most probably shouldn't be using Flash at all.

--
To email a reply, remove (dash)un(dash). Mail sent to the un
address is considered spam and automatically deleted.