Accessibility vs Design
Andy Jacobs
Ok, having a crack at my own site with CSS to see if I can come up with
a design that I like and one that will pass the main tests.
So far, I've come up with this:
http://www.redcatmedia.net/new/
Yes, it uses a fixed width. I'm guessing the only way round this would
be to make the image fade to the right and then use the same colour as
the background for the div.
Can anyone think of another way?
I've used CSS for the drop down with a hack for IE. It passes the test
and just degrades to an unordered list with no style sheets. I've tried
in in Opera, FF, Safari and IE on the Mac, and IE on a PC and it looks
fine. Anyone with any other platforms who can have a look would be
useful.
--
Andy Jacobs
www.redcatmedia.net
Intelligent Websites For Intelligent Business People
Stevie D
> http://www.redcatmedia.net/new/
>
> Yes, it uses a fixed width. I'm guessing the only way round this would
> be to make the image fade to the right and then use the same colour as
> the background for the div.
If the banner image is in the background, as it appears to be, it
doesn't matter if the width of the element is different from the width
of the image. You could set max-width:760px to help decent browsers -
and if the user has a narrower screen, the banner image will be
cropped. You can then set either background-repeat:repeat-x; or
background-repeat:no-repeat; and set an appropriate dark background
colour, for browsers that don't support max-width and where users have
a wider window than that.
A few other comments:
• The text is _TINY_. I have my text set to a size that I like. I
don't want to be told that I'm going to read it at _70%_ of that size,
in an small and ugly font moreover. Yes, I know I can change it, but I
shouldn't have to.
• Some sort of rollover effect on the menu would help, make it easier
for people to see where they are clicking and to reinforce that there
is a link there.
• Website is one word. Web site is incorrect. (Conversely web page is
two words, webpage is incorrect. Don't ask...)
• I'm not sure that what you've got as <h2> can really be called a
heading.
--
Stevie D
\\\\\ ///// Bringing dating agencies to the
\\\\\\\__X__/////// common hedgehog since 2001 - "HedgeHugs"
___\\\\\\\'/ \'///////_____________________________________________
Andy Jacobs
Andy Jacobs <an...@redcatmedia.net> wrote:
> Yes, it uses a fixed width.
And now it doesn't.
Andy Jacobs
Stevie D <ste...@sjd117.freeserve.co.uk> wrote:
> Andy Jacobs wrote:
>
> A few other comments:
>
> € The text is _TINY_. I have my text set to a size that I like. I
> don't want to be told that I'm going to read it at _70%_ of that size,
> in an small and ugly font moreover. Yes, I know I can change it, but I
> shouldn't have to.
Hmm... so what do I need to do? I've got my browser set to 100% for
fonts and it looks slightly too big for my taste so I left it alone.
>
> € Some sort of rollover effect on the menu would help, make it easier
> for people to see where they are clicking and to reinforce that there
> is a link there.
Just noticed that. I think I have an error in my stylesheet somewhere
as I've specified a:hover to be ff0000. Something's obviously being
overwritten or ignored.
>
> € Website is one word. Web site is incorrect. (Conversely web page is
> two words, webpage is incorrect. Don't ask...)
Are you sure about that? I've changed it on the site but I'm just about
to have 10500 postcards (or post cards) printed and I'd rather make sure
it was right on there as well :o)
>
> € I'm not sure that what you've got as <h2> can really be called a
> heading.
I did wonder and I thought it would come up. How many words constitute
a headline? The reason for doing it was that it's actually in its own
div. This means that I could put the body in a central column and put
the headline in the left column, or something.
Cheers for the tips, especially the one about websites!
Stevie D
> Hmm... so what do I need to do? I've got my browser set to 100% for
> fonts and it looks slightly too big for my taste so I left it alone.
What your stylesheet tells browsers to do is to use a font that is 70%
of the default size. That means that for most people it will be too
small. You shouldn't set the size of body text to less than 90%
(0.9em). If it is too big on your screen, you need to adjust your
default font settings so they are comfortable.
>> • Website is one word. Web site is incorrect. (Conversely web page is
>> two words, webpage is incorrect. Don't ask...)
>
> Are you sure about that? I've changed it on the site but I'm just about
> to have 10500 postcards (or post cards) printed and I'd rather make sure
> it was right on there as well :o)
According to the local authority where I work, that is what we have
been instructed to use.
Google:
Results 1 - 10 of about 1,170,000,000 for website
Results 1 - 10 of about 912,000,000 for "web site"
Not as conclusive as I would have expected.
Restricting it to UK only gives:
Results 1 - 10 of about 83,300,000 for website
Results 1 - 10 of about 29,100,000 for "web site"
which suggests that "web site" is more common outside the UK.
Jim Ley
<ste...@sjd117.freeserve.co.uk> wrote:
>Andy Jacobs wrote:
>> Are you sure about that? I've changed it on the site but I'm just about
>> to have 10500 postcards (or post cards) printed and I'd rather make sure
>> it was right on there as well :o)
>
>According to the local authority where I work, that is what we have
>been instructed to use.
but that sounds like a local style guide (so the usage is consistent
in all you produce), there's certainly no basis for saying either is
wrong, as your stats on usage showed...
Jim.
jake
<an...@redcatmedia.net> writes
>Morning all
>
>Ok, having a crack at my own site with CSS to see if I can come up with
>a design that I like and one that will pass the main tests.
>
>So far, I've come up with this:
>
>http://www.redcatmedia.net/new/
>
>
A few thoughts:
(a) Hide your CSS from older browsers:
http://www.gododdin.demon.co.uk/ng/AQ04X.JPG (15k)
(b) Headings, etc.:
http://www.gododdin.demon.co.uk/ng/AQ05X.JPG (55k)
(c) Your <acronym> should be <abbr> ......
(d) Get someone to check your grammar and punctuation
(e) Deactivate any link on page that points to itself.
(f) Don't claim WAI-AAA until your page rates it.
regards.
--
Jake
jake
>
>>
>> € Website is one word. Web site is incorrect. (Conversely web page is
>> two words, webpage is incorrect. Don't ask...)
>
>Are you sure about that? I've changed it on the site but I'm just about
>to have 10500 postcards (or post cards) printed and I'd rather make sure
>it was right on there as well :o)
>
[snip]
'Web site' is correct; 'website' is trendy, and acceptable to anyone who
doesn't know any better ;-)
Suggestion: go to the Microsoft, Opera, Netscape, W3C, etc. sites (i.e.
the major players) and see what they do.
regards.
--
Jake
Jim Ley
wrote:
>[snip]
>
>>
>>>
>>> € Website is one word. Web site is incorrect. (Conversely web page is
>>> two words, webpage is incorrect. Don't ask...)
>>
>>Are you sure about that? I've changed it on the site but I'm just about
>>to have 10500 postcards (or post cards) printed and I'd rather make sure
>>it was right on there as well :o)
>>
>[snip]
>
>'Web site' is correct; 'website' is trendy, and acceptable to anyone who
>doesn't know any better ;-)
Trendy? the Economist Style Guide has it as website, and the 150 odd
year paper is hardly focusing on the trendy side of words...
Both are acceptable, both are in common usage, just pick a style
you're happy with.
>Suggestion: go to the Microsoft, Opera, Netscape, W3C, etc. sites (i.e.
>the major players) and see what they do.
Such technical sites are probably only relevant to technical
audiences, not everyone is such an audience
Jim.
jake
<j...@jibbering.com> writes>>>> € Website is one word. Web site is incorrect. (Conversely web page is
>>>> two words, webpage is incorrect. Don't ask...)
>>>
>>>Are you sure about that? I've changed it on the site but I'm just about
>>>to have 10500 postcards (or post cards) printed and I'd rather make sure
>>>it was right on there as well :o)
>>>
>>[snip]
>>
>>'Web site' is correct; 'website' is trendy, and acceptable to anyone who
>>doesn't know any better ;-)
>
>Trendy? the Economist Style Guide has it as website, and the 150 odd
>year paper is hardly focusing on the trendy side of words...
>
>Both are acceptable, both are in common usage, just pick a style
>you're happy with.
>
>>Suggestion: go to the Microsoft, Opera, Netscape, W3C, etc. sites (i.e.
>>the major players) and see what they do.
>
>Such technical sites are probably only relevant to technical
>audiences, not everyone is such an audience
>
>Jim.
As you say, both are acceptable.
regards.
--
Jake
Andy Jacobs
jake <ja...@gododdin.demon.co.uk> wrote:
> In message <andy-4FD000.0...@news.btinternet.com>, Andy Jacobs
> <an...@redcatmedia.net> writes
> >Morning all
> >
> >Ok, having a crack at my own site with CSS to see if I can come up with
> >a design that I like and one that will pass the main tests.
> >
> >So far, I've come up with this:
> >
> >http://www.redcatmedia.net/new/
> >
> [snip]
>
> >
> A few thoughts:
>
> (a) Hide your CSS from older browsers:
> http://www.gododdin.demon.co.uk/ng/AQ04X.JPG (15k)
If that means making a compromise on what current browsers will see then
I'm not prepared to do it. Anyone using NN4.7 isn't in my target
audience. If it can be done without compromising, how do I go about it?
>
> (c) Your <acronym> should be <abbr> ......
Done
>
> (f) Don't claim WAI-AAA until your page rates it.
Bobby and Site Valet both say there are no problems. If they pass all
automatic tests and I'm happy that the warnings aren't relevant then
that's AAA.
Stevie D
> but that sounds like a local style guide (so the usage is consistent
> in all you produce), there's certainly no basis for saying either is
> wrong, as your stats on usage showed...
From www.answers.com :
: USAGE NOTE The transition from World Wide Web site to Web site to
: website seems to have progressed as rapidly as the technology itself.
: The development of website as a single uncapitalized word mirrors the
: development of other technological expressions which have tended to
: evolve into unhyphenated forms as they become more familiar. Thus
: email has recently been gaining ground over the forms E-mail and
: e-mail, especially in texts that are more technologically oriented.
: Similarly, there has been an increasing preference for closed forms
: like homepage, online, and printout.
Chris Morris
> In article <$dJpLUDZ...@gododdin.demon.co.uk>,
> jake <ja...@gododdin.demon.co.uk> wrote:
> > In message <andy-4FD000.0...@news.btinternet.com>, Andy Jacobs
> > <an...@redcatmedia.net> writes
> > >
> >
> > (f) Don't claim WAI-AAA until your page rates it.
>
> Bobby and Site Valet both say there are no problems.
http://www.cs.tut.fi/~jkorpela/www/acctools.html might be worth a
read. You trust automatic checks more than you should.
Ultimately, it's not Bobby and Site Valet that are going to be reading
your pages.
> If they pass all automatic tests and I'm happy that the warnings
> aren't relevant then that's AAA.
I'm mildly surprised that
<h2>...</h2>
...
<h1>...</h1>
...
<h2>...</h2>
didn't give a warning from either, then. (Guideline 3.5)
Equally, it's only the fact that neither (to my knowledge) also checks
external stylesheets that prevents:
#top {
...
height: 80px;
...
}
from triggering a problem. (Guideline 3.4)
http://www.dur.ac.uk/c.i.morris/rcm1.png
It's not possible to get to the links under Services without using a
mouse (Guideline 6.4). Depending on what the navigation is like when
you're on the Services page, this may not actually be a problem.
You specify color: without background-color: and vice-versa, which in
certain conditions will give problems with Guideline 2.2
http://www.dur.ac.uk/c.i.morris/rcm2.png
...is now a <span class="redbold">legal requirement</span> under...
would be better using <em> or <strong> rather than <span>. Guideline
3.3, according to
http://www.w3.org/TR/WCAG10-HTML-TECHS/#text-emphasis
You use GIF rather than PNG in places (Guideline 11.1 *unless* you
have a good reason for using GIF)
So that's a few Priority 2 (-AA) guidelines that the automatic
checkers missed (though the W3C's CSS checker should have given
warnings about the 2.2 problem)
I also think it would be reasonable to argue that the presence of the
word 'icon' in the alt text for the WAI-AAA icon makes it less than
perfect as a text equivalent, and so to an extent fails the Priority 1
Guideline 1.1, but it's not a particularly serious failure in this case.
--
Chris
Andy Jacobs
Chris Morris <c.i.m...@durham.ac.uk> wrote:
> I also think it would be reasonable to argue that the presence of the
> word 'icon' in the alt text for the WAI-AAA icon makes it less than
> perfect as a text equivalent, and so to an extent fails the Priority 1
> Guideline 1.1, but it's not a particularly serious failure in this case.
That's quite funny then as that's the code the site gives you if you
want to display the logo!
Just shows how wrong things can be when open to interpretation ;o)
Alan J. Flavell
> In article <87mzqnb...@dinopsis.dur.ac.uk>,
> Chris Morris <c.i.m...@durham.ac.uk> wrote:
>
> > I also think it would be reasonable to argue that the presence of
> > the word 'icon' in the alt text for the WAI-AAA icon makes it less
> > than perfect as a text equivalent, and so to an extent fails the
> > Priority 1 Guideline 1.1, but it's not a particularly serious
> > failure in this case.
I agree with Chris on this, as you'd be able to deduce...
> That's quite funny then as that's the code the site gives you
You're talking about the W3C, http://www.w3.org/WAI/ , and as
recommended by http://www.w3.org/WAI/WCAG1-Conformance ?
Yes, the alt text which they use here is suboptimal, in my opinion.
The function of the icon is to represent a claim of conformance to a
particular WAI level. Thus, logically, the alternative text should be
a claim of conformance to that WAI level. Not a description of an
image.
If a description of an image *was* wanted, then the proper place for
it would be the title attribute of the "img" tag. However, in
practical terms that wouldn't be such a good idea, as it would mask
the well-chosen title attribute which they've provided on the
enclosing "a href=" to identify the target of the link.
And on a minor technical point, it's inadvisable to include a line
break within the value of an ALT attribute, in the way that their
samples have done. Some browsers make a mess of that.
> if you want to display the logo!
I was on the relevant W3C mailing list at the time that the 1.0
guidelines were developed. We had several rounds of discussion which
came down on the side of meaningful *alternative* text along the lines
that Chris and I are promoting here, but there were a couple of
hard-bitten types who kept smuggling "description of the image" back
in again after each round of correction had been done. I guess folks
finally got tired of arguing. (They dumped me off their mailing list
when our mailer rejected a virus which their mailing list had
forwarded, ho hum).
> Just shows how wrong things can be when open to interpretation ;o)
As I've said before, the guidelines are a compromise reached (at a
certain point in time) after discussion and argument between a large
number of folks each with their own agenda. Some cared specifically
about blind access, some about dyslexia, some about attention deficit,
and so on and so on.
You need to take your own responsibility for what you're doing, within
the scope of the guidelines, not slavishly follow some legalistic
interpretation of them.
Andy Jacobs
"Alan J. Flavell" <fla...@ph.gla.ac.uk> wrote:
> On Sun, 22 May 2005, Andy Jacobs wrote:
>
> > In article <87mzqnb...@dinopsis.dur.ac.uk>,
> > Chris Morris <c.i.m...@durham.ac.uk> wrote:
> >
> > > I also think it would be reasonable to argue that the presence of
> > > the word 'icon' in the alt text for the WAI-AAA icon makes it less
> > > than perfect as a text equivalent, and so to an extent fails the
> > > Priority 1 Guideline 1.1, but it's not a particularly serious
> > > failure in this case.
>
> I agree with Chris on this, as you'd be able to deduce...
>
> > That's quite funny then as that's the code the site gives you
>
> You're talking about the W3C, http://www.w3.org/WAI/ , and as
> recommended by http://www.w3.org/WAI/WCAG1-Conformance ?
>
> Yes, the alt text which they use here is suboptimal, in my opinion.
>
> The function of the icon is to represent a claim of conformance to a
> particular WAI level. Thus, logically, the alternative text should be
> a claim of conformance to that WAI level. Not a description of an
> image.
Yes. The penny has just dropped in ALT text! It's the *message* that the
image conveys: on a weather page - sun.jpg alt="It's sunny today" and
not alt="Image of sun".
Geoff Berrow
<andy-709B50.1...@news.btinternet.com> from Andy Jacobs
contained the following:
>Yes. The penny has just dropped in ALT text! It's the *message* that the
>image conveys: on a weather page - sun.jpg alt="It's sunny today" and
>not alt="Image of sun".
Hasn't dropped with me I'm afraid. When I use an image I might want to
convey something metaphorically. Describing the image keeps the
metaphor intact.
Let's suppose I write an article on wisdom. I illustrate the article
with a picture of an owl. Is it not somewhat patronising to explain the
meaning of the image?
--
Geoff Berrow 0110001001101100010000000110
001101101011011001000110111101100111001011
100110001101101111001011100111010101101011
Andy Jacobs
Geoff Berrow <blth...@ckdog.co.uk> wrote:
> I noticed that Message-ID:
> <andy-709B50.1...@news.btinternet.com> from Andy Jacobs
> contained the following:
>
> >Yes. The penny has just dropped in ALT text! It's the *message* that the
> >image conveys: on a weather page - sun.jpg alt="It's sunny today" and
> >not alt="Image of sun".
>
> Hasn't dropped with me I'm afraid. When I use an image I might want to
> convey something metaphorically. Describing the image keeps the
> metaphor intact.
>
> Let's suppose I write an article on wisdom. I illustrate the article
> with a picture of an owl. Is it not somewhat patronising to explain the
> meaning of the image?
Maybe to you or me it would be. But if you'd been blind from birth then
the metaphor for wisdom would mean nothing to you, as would the image of
the sun in which case "It's sunny today" is going to tell you much more
than a description of the image would. I think it's different when
you're talking about metaphors. Can't you use the extended description
is these cases? I still think metaphors won't fit any model we can come
up with. It's also going to be a cultural thing. Translating the text
into a different language is easy enough but if they don't use the same
metaphors then you're still buggered.
Alan J. Flavell
> When I use an image I might want to convey something metaphorically.
> Describing the image keeps the metaphor intact.
Feel free to be explicit. Show us what alt text you propose, and how
it will read in its context.
> Let's suppose I write an article on wisdom. I illustrate the
> article with a picture of an owl. Is it not somewhat patronising to
> explain the meaning of the image?
The reader in question isn't seeing the image. It's your job to
convey your meaning, if any, in words. The phrase "Image of an owl"
suddenly appearing in the text doesn't make any kind of sense in a
text-only rendering, does it? So let's see your version, in context.
Stevie D
> Andy Jacobs wrote:
>
>> Yes. The penny has just dropped in ALT text! It's the *message* that
>> the image conveys: on a weather page - sun.jpg alt="It's sunny today"
>> and not alt="Image of sun".
Even simpler than that would just be
alt="Sun"
In the same way that when you have a navigation button formed from an
image, all the alt text needs to do is replicate what the text in the
graphic says.
If you want to spell it out further, then you should use
title="It's sunny today"
> Hasn't dropped with me I'm afraid. When I use an image I might want
> to convey something metaphorically. Describing the image keeps the
> metaphor intact.
It's about levels. What one person will sensibly alt as "bullet", or
even better "*", another person will alt as "small blue globe bullet".
The alt should tell the reader what the picture is of, briefly.
> Let's suppose I write an article on wisdom. I illustrate the article
> with a picture of an owl. Is it not somewhat patronising to explain the
> meaning of the image?
Then alt="Owl" would be perfectly sufficient. If you feel that more
might be useful, use a longdesc to expand further, eg "Owl wearing a
mortarboard, glasses and a waistcoat, sitting at a desk, to represent
wisdom and intelligence".
Geoff Berrow
<Pine.LNX.4.62.05...@ppepc56.ph.gla.ac.uk> from Alan J.
Flavell contained the following:
>> Let's suppose I write an article on wisdom. I illustrate the
>> article with a picture of an owl. Is it not somewhat patronising to
>> explain the meaning of the image?
>
>The reader in question isn't seeing the image. It's your job to
>convey your meaning, if any, in words. The phrase "Image of an owl"
>suddenly appearing in the text doesn't make any kind of sense in a
>text-only rendering, does it? So let's see your version, in context.
Surely it's only my job to convey any meaning to the same extent as the
image would? In which case, a description of the image could often be
the most appropriate.
Chris Morris
> <andy-709B50.1...@news.btinternet.com> from Andy Jacobs
> >Yes. The penny has just dropped in ALT text! It's the *message* that the
> >image conveys: on a weather page - sun.jpg alt="It's sunny today" and
> >not alt="Image of sun".
>
> Hasn't dropped with me I'm afraid. When I use an image I might want to
> convey something metaphorically. Describing the image keeps the
> metaphor intact.
>
> Let's suppose I write an article on wisdom. I illustrate the article
> with a picture of an owl. Is it not somewhat patronising to explain the
> meaning of the image?
No, not really.
alt="The owl has been a symbol of wisdom for centuries."
and put the image of the owl at an appropriate point in the document
(It could make a nice opening line, depending on how you wrote the
article)
The only mildly annoying point is that there's then arguably
information in the alt text which isn't in the image, unless it's a
picture of an owl that's obviously centuries old.
Alternatively, alt="", if it's just intended to convey general
metaphorical wisdom, as the article presumably still conveys that
message without the picture.
Any chance you could write an article on wisdom and put a picture of
an owl in it? ;) This really isn't something that there can
meaningfully be a 'one-size-fits-all' answer for...
--
Chris
Chris Morris
> Geoff Berrow wrote:
> > Andy Jacobs wrote:
> >
> >> Yes. The penny has just dropped in ALT text! It's the *message* that
> >> the image conveys: on a weather page - sun.jpg alt="It's sunny today"
> >> and not alt="Image of sun".
>
> Even simpler than that would just be
> alt="Sun"
> In the same way that when you have a navigation button formed from an
> image, all the alt text needs to do is replicate what the text in the
> graphic says.
True, but this isn't quite the same thing.
> If you want to spell it out further, then you should use
> title="It's sunny today"
It all depends on context.
Context 1:
Today's weather: <img ...>
Context 2:
<img ...> Go and have a picnic!
Context 3: (A weather map, if you didn't guess...)
<img ...> <img ...> <img ...> <img ...> <img ...>
In context 1, alt="Sun" (well, alt="Sunny") would be better.
In context 2, alt="It's sunny today." would be better.
In context 3, alt="North-west England: Sun"
All from the same image.
> > Hasn't dropped with me I'm afraid. When I use an image I might want
> > to convey something metaphorically. Describing the image keeps the
> > metaphor intact.
>
> It's about levels. What one person will sensibly alt as "bullet", or
> even better "*", another person will alt as "small blue globe bullet".
I'd have difficulty describing alt="bullet" as sensible.
(IMO, putting the image as the list-style-image: on a <ul> would be
the best way of going about it, as that almost ensures it will have
the same non-visual representation as any other bullet)
> The alt should tell the reader what the picture is of, briefly.
That, I would disagree with.
> > Let's suppose I write an article on wisdom. I illustrate the article
> > with a picture of an owl. Is it not somewhat patronising to explain the
> > meaning of the image?
>
> Then alt="Owl" would be perfectly sufficient.
Would it? Does the following make sense?
...wisdom cannot be measured in terms of popular consensus.
<img alt="Owl" ...>
To acknowledge the existence of wisdom assumes...
Which linearises as:
"...wisdom cannot be measured in terms of popular consensus. Owl To
acknowledge the existence of wisdom assumes..."
which makes no sense.
So it might be good, or it might not, depending on context.
--
Chris
Alan J. Flavell
> I noticed that Message-ID:
> <Pine.LNX.4.62.05...@ppepc56.ph.gla.ac.uk> from Alan J.
> Flavell contained the following:
>
> >The reader in question isn't seeing the image. It's your job to
> >convey your meaning, if any, in words. The phrase "Image of an owl"
> >suddenly appearing in the text doesn't make any kind of sense in a
> >text-only rendering, does it? So let's see your version, in context.
>
> Surely it's only my job to convey any meaning to the same extent as the
> image would?
Just so.
> In which case, a description of the image could often be
> the most appropriate.
I've asked you to provide an example, in context. The choice of alt
text depends on what you purposed in using the image: only you can
decide what that purpose is. But when you've chosen your alt text,
and put it into its context, we'd all be in a position to evaluate
whether it makes any kind of sense as a piece of text in context.
I've got this worrying suspicion that you'd happily copy the example
of the US Embassy Belgrade:
Small red bullet Response to Terrorism
D.M. Procida
> > Let's suppose I write an article on wisdom. I illustrate the article
> > with a picture of an owl. Is it not somewhat patronising to explain the
> > meaning of the image?
>
> Maybe to you or me it would be. But if you'd been blind from birth then
> the metaphor for wisdom would mean nothing to you,
Blind people understand metaphor too, or is that not what you meant? In
any case, the description "Wise owl picture" seems to meet all the needs
of that particular case.
Daniele
--
Apple Juice Ltd
Chapter Arts Centre
Market Road www.apple-juice.co.uk
Cardiff CF5 1QE 029 2019 0140
Alan J. Flavell
> any case, the description "Wise owl picture" seems to meet all the
> needs
I note that none of the proponents of these nonsensical alt texts have
responded to the invitation to present them in a realistic textual
context.
Geoff Berrow
<Pine.LNX.4.62.05...@ppepc56.ph.gla.ac.uk> from Alan J.
Flavell contained the following:
>I've asked you to provide an example, in context. The choice of alt
>text depends on what you purposed in using the image: only you can
>decide what that purpose is. But when you've chosen your alt text,
>and put it into its context, we'd all be in a position to evaluate
>whether it makes any kind of sense as a piece of text in context.
Earlier you wrote 'hard-bitten types who kept smuggling "description of
the image" back in' which implied to me that you did not think this
would ever be appropriate. I attempted to show that there were
occasions where description of the image would be perfectly appropriate.
If my purpose is to be metaphoric, or to provoke some reaction then a
description of the image may well be appropriate. As you say, it
depends on the context.
Another example. If I run an anti abortion site and show a picture of a
dismembered aborted fetus my purpose is to shock. I can't think of
anything, other than a graphic description of the image, that would have
the same effect.
I'd like to hear what blind people think. Personally, I think I would
find it patronising to be treated as if I could not deduce the meaning
of an image from a good description.
Geoff Berrow
Chris Morris contained the following:
>Any chance you could write an article on wisdom and put a picture of
>an owl in it? ;)
I'm not young, but I don't feel old enough to be an authority on wisdom
yet. ;-)
Alan J. Flavell
> I noticed that Message-ID:
> <Pine.LNX.4.62.05...@ppepc56.ph.gla.ac.uk> from Alan J.
> Flavell contained the following:
>
> >I've asked you to provide an example, in context. The choice of alt
> >text depends on what you purposed in using the image: only you can
> >decide what that purpose is. But when you've chosen your alt text,
> >and put it into its context, we'd all be in a position to evaluate
> >whether it makes any kind of sense as a piece of text in context.
>
> Earlier you wrote 'hard-bitten types who kept smuggling "description of
> the image" back in' which implied to me that you did not think this
> would ever be appropriate.
It's wrong as a principle, because it gives the wrong answer more
often than it gives the right one. I didn't say it was *always*
wrong, but if you follow the principle that the alt text supplies
alternative text "to serve as content when the element cannot be
rendered normally" (in the words of the more perceptive part of the
HTML4 specification) then you shouldn't go far wrong.
> I attempted to show
Not really, because you didn't provide any realistic context for
your proposal. To my way of thinking, as soon as you did so, you'd
realise that the text made no kind of contextual sense, like this
currently from BBC news:
BBC NEWS: VIDEO AND AUDIO
Camilla, Duchess of Cornwall Camilla carries out her first solo
official engagement
BBC Six O'Clock News
or these:
Boy using a laptop Children drive home internet use
Nintendo has not yet decided on colours for Revolution Long wait for
Nintendo Revolution
The sail training ship Tenacious in Milford dry dock, taken by David
Barrett Your pictures
> that there were occasions where description of the image would be
> perfectly appropriate.
Occasionally, yes, when you compose the "text to serve as content" you
might decide that the appropriate content is a brief description of
the image. But you'll still need to fit that into its textual context
so that it makes sense to a text-only reader.
> I'd like to hear what blind people think.
One of them told us that a web page was an inherently visual product,
that could only be presented to him by giving a detailed description
of *every* visual feature of the page. Another expressed exactly the
opposite view, that visual features were merely incidental to the
content, and he wanted to get the content, not be distracted with
what, to him, were irrelevant details of the visuals. You can satisfy
both of them, if you want to, by providing longdesc links; you don't
need to fool with the alternative text: its purpose is already
documented - "to serve as content" - and I go along with that. Think
of the alternative text and the image as alternative representations
of your content.
ttfn
Geoff Berrow
<Pine.LNX.4.62.05...@ppepc56.ph.gla.ac.uk> from Alan J.
Flavell contained the following:
>> any case, the description "Wise owl picture" seems to meet all the
>> needs
>
>I note that none of the proponents of these nonsensical alt texts have
>responded to the invitation to present them in a realistic textual
>context.
And why should we? I thought we were attempting to come to a working
guide on what the alt text should contain? Now you are saying 'well it
depends on the context'.
Sounds like a cop out to me.
Pyromancer
> I noticed that Message-ID: <87acmm5...@dinopsis.dur.ac.uk> from
> Chris Morris contained the following:
>
> >Any chance you could write an article on wisdom and put a picture of
> >an owl in it? ;)
>
> I'm not young, but I don't feel old enough to be an authority on
wisdom
> yet. ;-)
But that sentence is a statement of wisdom in itself, so you could do:
<html (doctype)>
<head>
<title>Wisdom</title>
</head>
<body>
<img src='owl-01.jpg' title='The Bird Himself' alt='???'>
<p>I'm not young, but I don't feel old enough
to be an authority on wisdom yet.
<img src='smile.gif' alt=' fx: smile '></p>
</body>
</html>
Now, what sensible text should replace the ???
And come to that, is ' fx: smile ' a suitable replacement for graphical
version of ":-)" ?
D.M. Procida
"Wise owl picture" might not be the ideal text, but I fail to see how it
is nonsensical.
Alan J. Flavell
> "Wise owl picture" might not be the ideal text, but I fail to see
> how it is nonsensical.
When you finally present it in a realistic text-only context, you'll
surely recognise how absurdly it reads.
:::Jerry::::
"D.M. Procida" <real-not-anti...@apple-juice.co.uk> wrote in
message
news:1gx15n0.3ssarvc17u4mN%real-not-anti...@apple-juice.co.uk...
> Alan J. Flavell <fla...@ph.gla.ac.uk> wrote:
>
> > On Mon, 23 May 2005, D.M. Procida wrote:
> >
> > > any case, the description "Wise owl picture" seems to meet all
the
> > > needs
> >
> > I note that none of the proponents of these nonsensical alt texts
have
> > responded to the invitation to present them in a realistic textual
> > context.
>
> "Wise owl picture" might not be the ideal text, but I fail to see
how it
> is nonsensical.
WTF is a 'Wise Owl' ?
In other words, if someone doesn't know what a Wise Owl is or what it
represents then the alt text is nonsensical. To those who have the
image available it is obvious that you are using a picture to
illustrate a point, surely any alt text should relate to the point
being made and not what the picture is - for example an image of the
BBC strike today with the alt text "People outside the BBC" means
nothing (as there are always people outside the BBC), even though
factually correct, "There were many union pickets outside BBC
buildings today" gives the information that seeing the image would.
Sorry if I've gate-crashed an argument and just re made a previous
point....
:::Jerry::::
"Pyromancer" <pyrom...@inkubus-sukkubus.co.uk> wrote in message
news:1116872653.7...@z14g2000cwz.googlegroups.com...
> Geoff Berrow wrote:
>
> > I noticed that Message-ID: <87acmm5...@dinopsis.dur.ac.uk>
from
> > Chris Morris contained the following:
> >
> > >Any chance you could write an article on wisdom and put a picture
of
> > >an owl in it? ;)
> >
> > I'm not young, but I don't feel old enough to be an authority on
> wisdom
> > yet. ;-)
>
> But that sentence is a statement of wisdom in itself, so you could
do:
>
> <html (doctype)>
> <head>
> <title>Wisdom</title>
> </head>
>
> <body>
> <img src='owl-01.jpg' title='The Bird Himself' alt='???'>
> <p>I'm not young, but I don't feel old enough
> to be an authority on wisdom yet.
> <img src='smile.gif' alt=' fx: smile '></p>
>
> </body>
> </html>
>
> Now, what sensible text should replace the ???
Surely the 'title' description is wrong to start with, that *could* be
title='The wise Owl', that would then allow for something like alt='A
pillar of wisdom?'
>
> And come to that, is ' fx: smile ' a suitable replacement for
graphical
> version of ":-)" ?
>
Hmm, does (or will) aural styles allow for emotions, will we be able
to tell the text to speech browser to 'speak' with a slight laugh
?.... :~)
Andy Jacobs
"Alan J. Flavell" <fla...@ph.gla.ac.uk> wrote:
> On Mon, 23 May 2005, Geoff Berrow wrote:
>
> > I noticed that Message-ID:
> > <Pine.LNX.4.62.05...@ppepc56.ph.gla.ac.uk> from Alan J.
> > Flavell contained the following:
> >
> > >I've asked you to provide an example, in context. The choice of alt
> > >text depends on what you purposed in using the image: only you can
> > >decide what that purpose is. But when you've chosen your alt text,
> > >and put it into its context, we'd all be in a position to evaluate
> > >whether it makes any kind of sense as a piece of text in context.
> >
> > Earlier you wrote 'hard-bitten types who kept smuggling "description of
> > the image" back in' which implied to me that you did not think this
> > would ever be appropriate.
<snip>
> Think
> of the alternative text and the image as alternative representations
> of your content.
But that assumes that the text depends on the image for context.
1. A page with a heading of "The Weather in Portsmouth Today" and a
picture of the sun would be useless in a text only environment without
alt text of "It's going to be sunny today".
2. The same page with and image of blue skies over the South Downs
could still be sensible with alt text of "It's going to be sunny today"
BUT... If you have a text description of the weather in Portsmouth on
the page with an image of blue skies over the South Downs then the alt
text on the image becomes superfluous, doesn't it?
I've just been doing some digging on the W3 site and this is an example
they give:
<IMG src="97sales.gif" alt="Sales for 1997"
longdesc="sales97.html">
So, does the alt text tell me that the company is doing well or not?
Have sales gone up or down?
Oh, and for the person who suggested fx: smile for a :-), the W3 example
gives:
<ABBR title="smiley in ASCII art">:-)</ABBR>
:::Jerry::::
"Andy Jacobs" <an...@redcatmedia.net> wrote in message
news:andy-210934.2...@news.btinternet.com...
<snip>
>
> BUT... If you have a text description of the weather in Portsmouth
on
> the page with an image of blue skies over the South Downs then the
alt
> text on the image becomes superfluous, doesn't it?
Surely the above is an argument about *if* there should be any alt
text and not just alt="", if the image is on the page solely for
'design' / presentation there is little need for an alternate textual
description - which is what you describe above, the textual content of
the page gives all the information?
>
> I've just been doing some digging on the W3 site and this is an
example
> they give:
>
> <IMG src="97sales.gif" alt="Sales for 1997"
> longdesc="sales97.html">
>
> So, does the alt text tell me that the company is doing well or not?
> Have sales gone up or down?
No, but if you want to know you can use the longdesc link.
Alan J. Flavell
> > Think
> > of the alternative text and the image as alternative representations
> > of your content.
>
> But that assumes that the text depends on the image for context.
You don't have any images in this situation, you only have text. It's
your job to adapt to that situation.
If the image doesn't convey any significant meaning, then the
appropriate alt text is "", or possibly white space - all the
respectable guidelines make that clear. Otherwise, it's your job to
convey your meaning in words.
> 1. A page with a heading of "The Weather in Portsmouth Today" and a
> picture of the sun would be useless in a text only environment
> without alt text of "It's going to be sunny today".
So that's a good alt text. We're getting there.
[...]
> BUT... If you have a text description of the weather in Portsmouth on
> the page with an image of blue skies over the South Downs then the alt
> text on the image becomes superfluous, doesn't it?
Just so. As I said, there's no correct alt text for a specific image:
it depends on what job the image is meant to be doing, and the only
person who knows that is the one who put it there. But we can all
evaluate whether their resulting choice of text makes sense in its
textual context, which is why we need to see these various proposals
exhibited in a realistic context, then we can all comment on how well
we think they read.
> I've just been doing some digging on the W3 site and this is an example
> they give:
>
> <IMG src="97sales.gif" alt="Sales for 1997"
> longdesc="sales97.html">
Go to http://ppewww.ph.gla.ac.uk/~flavell/alt/alt-text.html#howlers
and scroll down to "Plot of fish population" to deduce my opinion of
that.
Stevie D
> In context 1, alt="Sun" (well, alt="Sunny") would be better.
> In context 2, alt="It's sunny today." would be better.
> In context 3, alt="North-west England: Sun"
>
> All from the same image.
True. I was thinking of the use of a sun icon in one particular
context, where time and location are spelt out.
> ...wisdom cannot be measured in terms of popular consensus.
> <img alt="Owl" ...>
> To acknowledge the existence of wisdom assumes...
>
> Which linearises as:
> "...wisdom cannot be measured in terms of popular consensus. Owl To
> acknowledge the existence of wisdom assumes..."
> which makes no sense.
>
> So it might be good, or it might not, depending on context.
That is the problem with *any* alt text where the image has been used
for decoration rather than as text replacement.
It's one reason why, in situations like that, I often enclose the alt
text in [...] square brackets - this makes it clear in a text browser
that it is not in the main flow, and I would hope that a screen reader
gives a similar indication.
It is also why I generally eschew decorative images, or set null alt
text, or make sure that as many images as possible are background
images defined in the CSS.
D.M. Procida
> > "Wise owl picture" might not be the ideal text, but I fail to see
> how it
> > is nonsensical.
>
> WTF is a 'Wise Owl' ?
I didn't say "Wise owl", I said "Wise owl picture". A sighted person
seeing the picture of a wise-looking owl in the context of text about
wisdom will be able to make and enjoy the appropriate associations. A
person who is able to make and enjoy the same associations, but who is
blind, can use the text "Wise owl picture" in the same way.
If the text were merely "Wise owl" then it would leave open the purpose
of the picture, which in this case is nothing very essential. By
including the word "picture" in the alt text then it's clearer that it
isn't serving some other, mre important purpose.
If the text were "In western cultures, the owl is a symbol of wisdom"
(or something similar) that would be distracting, irrelevant and
patronising.
The picture of the owl is not carrying a message or information - it's
simply a small associative device. It's not like an image of a sun to
indicate sunny weather, or a photograph of a public demonstration. So
the alt text also needs to reassure anyone who reads it that they don't
need to be looking for that message or information.
Andy Jacobs
<1gx1fsn.yew6lk13mnlytN%real-not-anti...@apple-juice.co.uk>,
real-not-anti...@apple-juice.co.uk (D.M. Procida) wrote:
> The picture of the owl is not carrying a message or information - it's
> simply a small associative device. It's not like an image of a sun to
> indicate sunny weather, or a photograph of a public demonstration. So
> the alt text also needs to reassure anyone who reads it that they don't
> need to be looking for that message or information.
In that case, if what I'm starting to pick up is correct, the alt text
should be "" or " " (stil not sure if they mean different things?
As I understand it, in order for alt text to be appropriate, the image
must convey information that isn't available in the text (i.e. the "It's
Sunny" example). If the image is doing anything else then alt text is
not appropriate but some other device might be. A photography site, for
example. The alt text isn't appropriate to describe an image of a
sunrise but an extended description would be. The example on the W3
site with the "Sales figures for 2005" might have been better using the
title text and extended description rather than the alt and extended
description. It might have been better to have"
alt="Sales grew over 2005"
title="Graph of sales figures for 2005"
...and then had the extended description as well.
Maybe we're getting too hung up on alt text per se and ignoring the fact
that their might be more appropriate methods of describing an image.
I don't think the owl needs alt text. JM2PW.
:::Jerry::::
> :::Jerry:::: <m...@privacy.net> wrote:
>
> > > "Wise owl picture" might not be the ideal text, but I fail to
see
> > how it
> > > is nonsensical.
> >
> > WTF is a 'Wise Owl' ?
>
> I didn't say "Wise owl", I said "Wise owl picture".
What's a picture the blind man asks.... It makes no difference if it's
a picture, it's the information that it sends (hence the old saying,
'A picture saves a thousand words').
A sighted person
> seeing the picture of a wise-looking owl in the context of text
about
> wisdom will be able to make and enjoy the appropriate associations.
A
> person who is able to make and enjoy the same associations, but who
is
> blind, can use the text "Wise owl picture" in the same way.
No they can't if they don't know what a 'Wise Owl (picture)' means.
>
> If the text were merely "Wise owl" then it would leave open the
purpose
> of the picture, which in this case is nothing very essential. By
> including the word "picture" in the alt text then it's clearer that
it
> isn't serving some other, mre important purpose.
No, it just says that there is a image of a wise owl, alt="Golden
Eagle picture" could be a picture of just that or the Eagle from the
USA Presidential shield - to someone who doesn't have the image
available, they will never know. IYSWIM
>
> If the text were "In western cultures, the owl is a symbol of
wisdom"
> (or something similar) that would be distracting, irrelevant and
> patronising.
...and as useless a alt description as just saying 'Wise Owl (picture}
IMO, why are you using the image of the 'Wise Owl', that is what you
need to impart in the information given in any alt text - otherwise
the image is decoration and alt="" is correct.
>
> The picture of the owl is not carrying a message or information -
it's
> simply a small associative device. It's not like an image of a sun
to
> indicate sunny weather, or a photograph of a public demonstration.
So
> the alt text also needs to reassure anyone who reads it that they
don't
> need to be looking for that message or information.
>
So, you are using it as decoration, alt="" is then correct IMO.
D.M. Procida
> > > WTF is a 'Wise Owl' ?
> >
> > I didn't say "Wise owl", I said "Wise owl picture".
>
> What's a picture the blind man asks....
Blind people know what pictures are. We're talking about the blind, not
the alingual.
> It makes no difference if it's a picture, it's the information that it
> sends (hence the old saying, 'A picture saves a thousand words').
Not all pictures contain information or messages - what do you do with
pictures that do something else ?
> A sighted person
> > seeing the picture of a wise-looking owl in the context of text
> about
> > wisdom will be able to make and enjoy the appropriate associations.
> A
> > person who is able to make and enjoy the same associations, but who
> is
> > blind, can use the text "Wise owl picture" in the same way.
>
> No they can't if they don't know what a 'Wise Owl (picture)' means.
Why do you keep changing the alt text I suggested? I wrote "Wise owl
picture", not "Wise owl (picture)" or "Wise owl".
> No, it just says that there is a image of a wise owl, alt="Golden
> Eagle picture" could be a picture of just that or the Eagle from the
> USA Presidential shield - to someone who doesn't have the image
> available, they will never know. IYSWIM
Maybe "Picture of a wise owl" would be clearer (i.e. it's a picture of
*a* wise owl, not some special particular wise owl).
> ...and as useless a alt description as just saying 'Wise Owl (picture}
> IMO, why are you using the image of the 'Wise Owl', that is what you
> need to impart in the information given in any alt text - otherwise
> the image is decoration and alt="" is correct.
An image might provide a non-specific illustration of the thing being
discussed, a visual association (like the owl), or make a visual joke.
None of these images are imparting information, but neither are they
merely decorative. They are all doing something, and if what they are
doing is important enough it needs to be reflected in the alt text.
Arguably in the case of the owl it's not very important, but it could
be.
Chris Morris
> <Pine.LNX.4.62.05...@ppepc56.ph.gla.ac.uk> from Alan J.
> Flavell contained the following:
> >responded to the invitation to present them in a realistic textual
> >context.
>
> And why should we? I thought we were attempting to come to a working
> guide on what the alt text should contain? Now you are saying 'well it
> depends on the context'.
>
> Sounds like a cop out to me.
Hardly. There is no algorithm (no matter how advanced an AI or human
intelligence is used) that will, given an image - without any
information about how it will be used - provide the One True
Alternative Text. This is because there isn't one.
I used the 'sun' weather picture earlier in the thread as an example
of something that could reasonably have multiple alternative texts -
even on the same page.
<img src="sun" alt="North-west: Sunny">
<img src="sun" alt="North-east: Sunny">
...
<img src="sun" alt="Sun index:"> <img src="shadedbar3.5" alt="3.5">
How does the correct alternative text not depend on the context and
usage of the image?
--
Chris
Chris Morris
> "Andy Jacobs" <an...@redcatmedia.net> wrote in message
> news:andy-210934.2...@news.btinternet.com...
> > example they give:
> >
> > <IMG src="97sales.gif" alt="Sales for 1997"
> > longdesc="sales97.html">
> >
> > So, does the alt text tell me that the company is doing well or not?
> > Have sales gone up or down?
>
> No, but if you want to know you can use the longdesc link.
data table for the graph, or similar, in this context. Useful
information but not the "instant impression" that looking at the graph
would provide.
There's plenty of space in the alt text to say "Sales for 1997 rose
20% over the year", after all.
If you're going to put all the information on a separate page, at
least do <a href="sales97.html"><img src="97sales.gif" alt="Sales for
1997"></a> rather than relying on the very poorly supported longdesc
to do the same job.
--
Chris
Chris Morris
> Geoff Berrow wrote:
> > I noticed that Message-ID: <87acmm5...@dinopsis.dur.ac.uk> from
> > Chris Morris contained the following:
> >
> > >Any chance you could write an article on wisdom and put a picture of
> > >an owl in it? ;)
> >
> > I'm not young, but I don't feel old enough to be an authority on
> > wisdom yet. ;-)
>
> But that sentence is a statement of wisdom in itself, so you could do:
> <img src='owl-01.jpg' title='The Bird Himself' alt='???'>
> <p>I'm not young, but I don't feel old enough
> to be an authority on wisdom yet.
> <img src='smile.gif' alt=' fx: smile '></p>
...
>
> Now, what sensible text should replace the ???
It depends on the exact content of the picture and why it's there.
alt=""
alt="Unlike the author, owls are renowned for their wisdom"
alt="Like the author, owls are renowned for their wisdom"
alt="More wisdom than the average owl!"
alt="Owlcorp Wisdom Supplements Ltd."
--
Chris
Alan J. Flavell
> Blind people know what pictures are. We're talking about the blind,
> not the alingual.
The kind of images that we're discussing here are a *means* of
conveying information content to the user[1]. When images are
unavailable, it's your job to convey that *information content* by the
available means i.e text.
Trying to convey the information obliquely, by describing the image
that they can't see, instead of directly offering the information
which the image was meant to represent, seems nearly always to produce
sub-optimal results in context (I note that still none of the
proponents have produced a realistic example in context).
> Not all pictures contain information or messages - what do you do
> with pictures that do something else ?
I'd follow the WAI guidelines and set the alt text to an empty string
or white space. Or consider inserting the image via CSS, and ignoring
it as part of the information content.
> An image might provide a non-specific illustration of the thing
> being discussed, a visual association (like the owl), or make a
> visual joke.
All of which could qualify as information content.
> None of these images are imparting information,
I don't agree. All of that can qualify as information content.
> and if what they are doing is important enough it needs to be
> reflected in the alt text.
As a principle I don't disagree, if you can pull it off convincingly
in a text-only situation. I haven't yet seen a proposal which does
that, IMO.
[1] pages whose whole purpose is to exhibit images, such as photo
galleries etc. are a different matter, evidently.
D.M. Procida
> Trying to convey the information obliquely, by describing the image
> that they can't see, instead of directly offering the information
> which the image was meant to represent, seems nearly always to produce
> sub-optimal results in context (I note that still none of the
> proponents have produced a realistic example in context).
I thought the example of an owl is a perfectly good one, particularly
because the image in question performs a function that *isn't* about the
provision of information.
> > An image might provide a non-specific illustration of the thing
> > being discussed, a visual association (like the owl), or make a
> > visual joke.
>
> All of which could qualify as information content.
Just possibly the first of these, but the latter two, no. A visual joke
or association does not provide information content, but alludes to and
is dependent upon a context, and requires the reader to be able to make
the necessary connections in order to work. The picture of the wise owl
carries an association, not information. Even a more complex allusion
(say, an image of Goya's _The Sleep of Reason_ in an article about Kant)
is not information-bearing - in both cases, any content of knowledge has
to be with the reader *already*.
Or do you want to insist categorically that any image must either be
merely decorative *or* information-bearing, and that no kind of image
could exist?
Andy Jacobs
1gx2m09.joabp0lf3fbaN%real-not-anti...@apple-juice.co.uk, "D.M.
Procida" <real-not-anti...@apple-juice.co.uk> wrote:
> Or do you want to insist categorically that any image must either be
> merely decorative *or* information-bearing, and that no kind of image
> could exist?
I think that there are different mechanisms to use and each is appropriate
for the type of image being used. There would appear to be 3 different
things: alt, title and extended description, each of these is appropriate
for different types of images.
I think that the owl would benefit from a title, but not alt text.
Andy
Alan J. Flavell
> Or do you want to insist categorically that any image must either be
> merely decorative *or* information-bearing, and that no kind of image
> could exist?
I don't "insist categorically" on anything, but the WWW will still do
what it does, no matter what you or I might want it to. In this sense
it separates content from presentation. You can either work with that
behaviour (which is what I'm recommending), or fight against it.
You still haven't supplied a realistic example in context. Just an
isolated alt text proposal, which in my opinion will read in a
ridiculous fashion when plonked-down in the middle of some realistic
text.
Frankly, this doesn't work:
Apple Juice Ltd - Macintosh consultancy, service and repairs
horizontal bar
Your faithful Mac works tirelessly for you. It's time for some
refreshment.
logo
The HTML4 spec itself (pre-dating the WAI guidelines) correctly
advises:
Do not specify irrelevant alternate text when including images
intended to format a page, for instance, alt="red ball" would be
inappropriate for an image that adds a red ball for decorating a
heading or paragraph. In such cases, the alternate text should be the
empty string ("").
I'd say read the spec, and follow its advice. You don't need me for
that.
Focus on that perceptive HTML4 specification: "text to serve as
content when the element cannot be rendered normally". Work out what
you purposed in putting the image there, and formulate some text which
is *appropriate* for that purpose - it might be alt="". Then you're
doing what the HTML spec and the WAI guidelines recommend, you don't
need my advice or opinion on the matter.
As I make plain on my page, my aim is to present the text-only user
with a text-only presentation of the content which makes sense in its
own terms. It seems to me (and to quite a number of long-standing www
participants whose opinions I've grown to respect) to work well in the
vast majority of cases - except where the images are the principal
purpose of the page.
References to the colour, font, or appearance of the page in some
other presentation than the text-only situation which these users are
in, only tends to distract from the content, rather than enhancing it.
Those who *want* such details could be recommended to get a client
agent which implements the longdesc, and that's where you could put
such details, if you wish.
There's also the title= attribute, but as I said earlier on this
thread, there's a practical limitation. Of course the title= of the "a
href=" should relate to the target of the link, whereas the title= of
the img should relate to the image itself - that's the theory. But in
practical terms, if you have an image as the sole content of an "a
href=", and you provide title attributes on both the img and the "a
href=", client agents will usually action only one of them, and ignore
the other.
D.M. Procida
> On Tue, 24 May 2005, D.M. Procida wrote:
>
> > Or do you want to insist categorically that any image must either be
> > merely decorative *or* information-bearing, and that no kind of image
> > could exist?
>
> I don't "insist categorically" on anything, but the WWW will still do
> what it does, no matter what you or I might want it to. In this sense
> it separates content from presentation. You can either work with that
> behaviour (which is what I'm recommending), or fight against it.
Well, that simply doesn't answer the question. Given categories of image
which are not information-bearing, nor mere decoration, but all the same
bear some kind of content, how are you going to deal with them in alt
text?
> Frankly, this doesn't work:
>
> Apple Juice Ltd - Macintosh consultancy, service and repairs
>
> horizontal bar
I wouldn't get too excited about that crappy old thing, which is indeed
a complete mess.
Chris Morris
> Alan J. Flavell <fla...@ph.gla.ac.uk> wrote:
> > I don't "insist categorically" on anything, but the WWW will still do
> > what it does, no matter what you or I might want it to. In this sense
> > it separates content from presentation. You can either work with that
> > behaviour (which is what I'm recommending), or fight against it.
>
> Well, that simply doesn't answer the question. Given categories of image
> which are not information-bearing, nor mere decoration, but all the same
> bear some kind of content, how are you going to deal with them in alt
> text?
It depends.
I teach an accessibility course occasionally. Part of that is picking
good alt text. I use three easy images (a purely decorative one, an
image of text, and a logo acting as a link to a homepage) and then
throw in a difficult one - a photograph of a tutorial.
I throw that in firstly to demonstrate that the apparently obvious
"photograph of tutorial" isn't particularly helpful, and secondly to
demonstrate that - if you don't believe it's solely decorative, as
some people taking the course have reasonably claimed - picking good
alternative text for it requires a bit of thought.
I would deal with it in alt text by thinking about why the image was
on the page, how the alternative text would fit in with the context of
the page, what text I would use if I didn't have the image available,
etc.
If you give me an example of an image of the sort you mean, *in its
real environment*, or at least in a plausible environment, then I can
come up with what I'd use for the alternative text, and tell you why.
If you give me a broad category of images I can only give you a broad
category of guidelines and a recommendation to "think about it".
--
Chris
Mark Goodge
keyboard and typed:
>Alan J. Flavell <fla...@ph.gla.ac.uk> wrote:
>
>> On Tue, 24 May 2005, D.M. Procida wrote:
>>
>> > Or do you want to insist categorically that any image must either be
>> > merely decorative *or* information-bearing, and that no kind of image
>> > could exist?
>>
>> I don't "insist categorically" on anything, but the WWW will still do
>> what it does, no matter what you or I might want it to. In this sense
>> it separates content from presentation. You can either work with that
>> behaviour (which is what I'm recommending), or fight against it.
>
>Well, that simply doesn't answer the question. Given categories of image
>which are not information-bearing, nor mere decoration, but all the same
>bear some kind of content, how are you going to deal with them in alt
>text?
I'm not sure what you mean here. If something has content, then it has
information, surely. Can you give me an example of an image which has
content but no information?
Mark
--
http://www.FridayFun.net - jokes, lyrics and ringtones!
"If you never know truth then you never know love"
Stevie D
> I throw that in firstly to demonstrate that the apparently obvious
> "photograph of tutorial" isn't particularly helpful, and secondly to
> demonstrate that - if you don't believe it's solely decorative, as
> some people taking the course have reasonably claimed - picking good
> alternative text for it requires a bit of thought.
Another point to recognise, that is often lost, is that some people
browse without images in general, but can choose to load them when
they wish to - either turning images off in a normal GUI browser, or
using a text-browser and calling a helper application. For images that
are, er, <fx: trying to think of the right word> supplementary to the
text [1], it is useful for these people to know (a) that there is a
picture there, and (b) roughly what it might be.
[1] Probably not the best word. Images that are not _purely_
decorative, but are not essential to the content, use or navigation of
the page.
D.M. Procida
> If you give me an example of an image of the sort you mean, *in its
> real environment*, or at least in a plausible environment, then I can
> come up with what I'd use for the alternative text, and tell you why.
>
> If you give me a broad category of images I can only give you a broad
> category of guidelines and a recommendation to "think about it".
Well, how about the Goya/Kant suggestion I made earlier, and if you want
more context: the text is a sarcastic critique of the junk-culture
fringes of post-structuralist thought and argues that we have a direct
obligation to rationality. The picture (of _The Sleep of Reason_) is
appears inline with a paragraph in which the phrase "the sleep of
reason" is used of a certain prevailing tendency towards post-scientific
medicine.
Obviously in this case the picture bears no *information*, but serves as
a *reminder* to someone who is already aware of it and what it means, so
it's certainly not mere decoration, either.
D.M. Procida
> >Well, that simply doesn't answer the question. Given categories of image
> >which are not information-bearing, nor mere decoration, but all the same
> >bear some kind of content, how are you going to deal with them in alt
> >text?
>
> I'm not sure what you mean here. If something has content, then it has
> information, surely. Can you give me an example of an image which has
> content but no information?
"It's raining" - a statement, bearing information
"Stop!" - an imperative, bearing no information, but having content
"Where's my lunch?" - a question, bearing no information, but having
content
A picture of an angry crowd in Karachi in a news report - information
A stock picture of a Pakistani policeman in a news report about police
corruption in Karachi - no information, but it has content
Unless you're using 'information' in a looser (and less satisfactory)
way than me, then I think you would agree that some of those above are
examples where there is content - meaningfulness, for want of a better
word - but not the bearing of information.
Liz
":::Jerry::::" <m...@privacy.net> wrote:
> alt="GoldenEagle picture" could be a picture of just that or the Eagle
> from the USA Presidential shield - to someone who doesn't have the
> image available, they will never know. IYSWIM
But that would just be plain *wrong*, since the one on the USA
presidential shield is a Bald Eagle.
BTW -
I agree with the subject line.
I'm not sure if you can really meld design with accessibility, and
coming from a time when bandwidth was everything (and still working
with a slow, dial-up connection) I am quite shocked by the huge
overhead in bandwidth caused by accessibility requirements (e.g.
title, longdesc) not to mention all the CSS hacks.
And still, no-one has answered the question I've been asking for a
couple of years: what when the images *are* the content?
(I emailed back and forth with Dave Shea on this one pointing out that
a tables-for-layout page was accessible to more browsers than a CSS
one. (I also don't have any fully (or even reasonably) CSS compliant
browsers available to me on this set-up at the moment, info for those
who haven't come across me before). He pointed out that I could easily
get all the text content of csszengarden. My opinion still is that the
text content *isn't* the essential content of his site - it's actually
just a hook to demonstrate the wonders of CSS design: the visual
design *is* in fact the point of the site, ergo the visual design *is*
the content of csszengarden. This *can't* be made accessible to the
blind any more than my site which essentially is pictures with a bit
of text thrown in to give people something to do while the images are
downloading - on most of my pages, the text is not the 'content', it's
the 'decoration'. Oh his page, the 'text' could equally well have been
gobbledegook.
(Mind you, I find it difficult to read some of the csszengarden pages
on my hi-res screen on the pc with my normal settings due to font size
and low contrast - and DS himself admits that some of the pages won't
let you resize even up to 150%, (though most resize to 150%, he clams)
because of the fixed width layouts. Admittedly, the small font sizes
and low contrast look much better: another reason why I don't think
you can have *really* good, contemporary design with full
accessibility, unless you provide accessible pages separately (if the
text content *is* your 'content'.
(Oooooh, if only I'd allowed myself the luxury of fixed-width tables -
sigh!
And I can't understand why deeply-nested divs are less morally
reprehensible than nested tables, don't get me started.)
BTW, my 2p worth on the 'wise owl scenario' is that it *should* be
alt="", but I'd probably put "wise owl picture" just because my usual
browser puts the alt text in the image 'box', so you can read it
before the pic downloads. In fact, before I knew that all browsers
didn't do this (!) I used to put little extra anecdotes in the alt
text, again to give people something to do while waiting for the pix
to download.
Slainte
Liz
--
Liz
"Alan J. Flavell" <fla...@ph.gla.ac.uk> wrote:
> As I've said before, the guidelines are a compromise reached (at a
> certain point in time) after discussion and argument between a large
> number of folks each with their own agenda. Some cared specifically
> about blind access, some about dyslexia, some about attention deficit,
> and so on and so on.
Which is a really interesting point, because there are very few sites
around anywhere, mine included, which are accessible by the
surprisingly large number of people with either a low reading age, or,
far more commonly, a low comprehension age. The latter is something
which even a lot of teachers don't seem to know: some kids can read
things and you'd think they had a reading age of 10 or 12, when aged
16, but they can't actually understand a lot of the words they've been
mechanically taught. In the school I work in (mainstream) I'd say at
least 10% of the leavers come into this category, and it's unlikely to
get better as they get older: in general they've had a lot of
one-to-one and small group help when in school, without any rise in
comprehension level. This *isn't* reading level: there are lots of
'systems' whereby you can work out the 'reading level' of a piece of
text, e.g. by average number of letters in words or other formulae,
but this doesn't say anything *at all* about understandibility.
So we talk about blind people, those who need a screen-reader (e.g.
dyslexics), people with motor problems (access keys) etc, but don't
think of the larger number of people with very low comprehension: but
let's face it - how many of us want to read 'Where's Spot?'
equivalents all day long? (In school, we find it very difficult to
'pitch' general (paper) communications with parents because we have to
try to be 'inclusive' and 'accessible' without being patronising and
insulting to those with average or above intelligence.)
I've looked at the sites of several 'must be accessible to the
detriment of everything else, it that's how it has to be' posters on
this group, as well as the 'highly recommended' sites about
accessibility whose sites would be totally inaccessible to those with
low comprehension age. Now, I'd personally totally support your
argument that they're not your 'target audience', but since on this
group I've been assured in the past that a photographer or artist
can't say that 'the blind' aren't in their target group, surely that
argument must apply to 'low comprehension' visitors also.
BTW, my site isn't any better, but I'm not offering goods or services,
thank goodness. :-)
Slainte
Liz
--
Alan J. Flavell
> A picture of an angry crowd in Karachi in a news report - information
>
> A stock picture of a Pakistani policeman in a news report about
> police corruption in Karachi - no information, but it has content
Despite Chris's patient attempt to explain it, and my less patient
attempt to do the same, you seem to be proving the truth of of that
comment by "Callie" (pen-name for an earlier contributor to
c.i.w.a.h):
Many authors haven't figured out exactly what they are trying to
present; they don't know what it is about the image that's important
to the page's intended audience. The reason you can't figure out why
their alt [texts] aren't working is that they don't know why the
images are there.
For determining the appropriate alt text we don't really need to be
told what your images look like, we *do* need to be told what you
purposed by putting them there. And, short of us guessing on the
basis of hints that you're dropping, that's still missing, as far as I
can see.
> Unless you're using 'information' in a looser (and less satisfactory)
> way than me,
I prefer to use the term "content", but in a general sense
"information" isn't really wrong. Whatever you call it, it's whatever
you purposed by putting the image there. That's what the "alternative
text" is needed for. For the other purposes you've still (modulo any
practical issues) got the title= attribute and the longdesc= link, you
shouldn't need to pollute the alternative text with what, for
text-mode readers, are usually extraneous details, even if some of the
text-mode users might want to take a specific interest in them, as
we've previously discussed on this thread.
Lynx, for example, has a keycommand to put handles on all images, so
that they can be loaded, let's say into an external image viewer, you
don't have to do *anything* at the authoring side to enable that,
indeed you have no idea whether the user has such an application
available nor what it is nor whether it's of any use to them, that's
their own private business.
Trying to second-guess what one group of users might want, and make
provision for that from the authoring side, will often produce
suboptimal results, especially for users whose situation is
sufficiently different or whose wishes/requirements are sufficiently
different - maybe even contradictory.
That's why HTML+CSS has this idea of separation of content from
presentation (which the www is going to do anyway - they're just
providing a codified way of achieving what is going to happen whether
we want it or not). Use the HTML to mark-up your content honestly,
*not* to influence what you think the users might want to do with it.
Offer (optional) CSS style sheet(s) proposing presentation(s) for as
many common situation(s) as you see fit, but keep open the option that
some situations won't want any of them, so don't use CSS to provide
substantive content.
It all hangs together and makes sense, and it's finally (after early
stylesheet proposals from 199-mumble were trampled into the dirt by
Netscape, who at that time evidently wanted to turn HTML into a
pixel-exact page design language for the glossy-brochure big spenders)
got the upper hand again, and become realistic for practical use IMHO.
Jim Ley
>I'm not sure if you can really meld design with accessibility, and
>coming from a time when bandwidth was everything (and still working
>with a slow, dial-up connection) I am quite shocked by the huge
>overhead in bandwidth caused by accessibility requirements (e.g.
>title, longdesc) not to mention all the CSS hacks.
title/longdesc are unlikely to increase the number of frames of data
downloaded. CSS hacks are bad, will always be bad, and have nothing
to do with accessibility, I think you're confusing the 2 together.
>(I emailed back and forth with Dave Shea on this one pointing out that
>a tables-for-layout page was accessible to more browsers than a CSS
>one.
I have nothing against tables to make up for the mistakes of the CSS
WG, the grid layout was and still is the single commonest one on the
web, and it's far and away the most understandable to the majority.
However there's nothing particularly inaccessible with a single table
used to provide a grid, multiple nested tables are the problem with
accessibility, and these are unecessary in almost any design.
> He pointed out that I could easily
>get all the text content of csszengarden.
csszengarden is very inaccessible to me, I find it very hard to
understand almost any of the examples, of course it's art, not a
website, as you note below.
> This *can't* be made accessible to the
>blind any more than my site which essentially is pictures with a bit
>of text thrown in to give people something to do while the images are
>downloading.
Which is true, which is why the alt saying "A Gerenuk in the long
grass" or similar is enough. (I'm doing your site from memory, it may
have morphed into something completely different, but I'm stilll
imagining it's a photo site with lots of animals from safari's and
similar) You could do more and had title and longdesc, but it's not
necessary, the images are the content in your site, and as you say it
cannot be made accessible.
There's no problem with design and accessibility, it's just that too
many people still have a mechanical viewpoint of it, and it's
especially difficult to talk about it in the abstract, as there are no
absolutes - you certainly do not need to make every site accessibile
to all, that's simply impossible.
Jim.
Jim Ley
>So we talk about blind people, those who need a screen-reader (e.g.
>dyslexics), people with motor problems (access keys) etc, but don't
>think of the larger number of people with very low comprehension: but
>let's face it - how many of us want to read 'Where's Spot?'
>equivalents all day long?
I'm sure a great many people in the group have heard about Jonathon
Chetwynd and his excellent evangelism for this community,
unfortunately his website http://www.peepo.com is currently
unavailable, but I'm in no doubt that the web-accessibility community
know of many of the problems you're discussing. (Other members of
WCAG etc. are of course also knowledgeable about the problem.)
>Now, I'd personally totally support your
>argument that they're not your 'target audience', but since on this
>group I've been assured in the past that a photographer or artist
>can't say that 'the blind' aren't in their target group.
The problem I think is confusion about what "target audience" means
and what it gets you out of, a simple alt text is trivial and helps
lots of people, you can't really argue against doing that, however
providing longdesc, of course not, just like the person writing "how
to write accessibile websites" doesn't need to create a duplicate page
in a symbolic language.
Jim.
Jim Ley
<fla...@ph.gla.ac.uk> wrote:
>I prefer to use the term "content", but in a general sense
>"information" isn't really wrong. Whatever you call it, it's whatever
>you purposed by putting the image there.
In most of the examples (such as the policeman in the news story) it's
there to do nothing but re-enforce the impression given by the text
for the visually orientated, and to break up the text into less
imposing chunks, it's artistic. The picture rarely contains any
additional content, alt="" would often be appropriate, but so are
other things that could help to re-enforce the message.
>It all hangs together and makes sense, and it's finally (after early
>stylesheet proposals from 199-mumble were trampled into the dirt by
>Netscape, who at that time evidently wanted to turn HTML into a
>pixel-exact page design language for the glossy-brochure big spenders)
>got the upper hand again, and become realistic for practical use IMHO.
But the good news is that we've now got Mozilla, Safari and Opera, and
they're giving us <canvas> and javascript for the pixel exact page
design, so we don't need any of that crappy mark-up any more. Great
news!
Cheers,
you purposed by putting the image there.
In most of the examples (such as the policeman in the news story) it's
there to do nothing but re-enforce the impression given by the text
for the visually orientated, and to break up the text into less
imposing chunks, it's artistic. The picture rarely contains any
additional content, alt="" would often be appropriate, but so are
other things that could help to re-enforce the message.
Alan J. Flavell
> Which is a really interesting point, because there are very few
> sites around anywhere, mine included, which are accessible by the
> surprisingly large number of people with either a low reading age,
> or, far more commonly, a low comprehension age.
That's why the WAI guidelines urge authors to use the simplest
language compatible with the nature of the content.
It's recognised that even the best authors won't be able to explain
General Relativity, or Fermat's Last Theorem or whatever, to a reading
age of 10, and it's not asking authors to dumb-down the content, only
to express it as simply and straighforwardly as possible compatible
with its nature.
As for dyslexia, it's now quite feasible for the reader to have the
content spoken out to them, instead of or alongside the visual
presentation, so as long as your authoring style is compatible with a
speaking browser you've probably done what you can/should.
> think of the larger number of people with very low comprehension: but
> let's face it - how many of us want to read 'Where's Spot?'
> equivalents all day long? (In school, we find it very difficult to
> 'pitch' general (paper) communications with parents because we have to
> try to be 'inclusive' and 'accessible' without being patronising and
> insulting to those with average or above intelligence.)
Point taken, and I can doubtless be faulted on aiming at an
above-average readership, even where the topic could be of wider
interest.
In the UK context one might take a cue from the plain english folks
and their "crystal mark". http://www.plainenglish.co.uk/
Though I'd fault their web page on *some* features: the style sheet
starts off on the right foot with 100% body text, and does not put
Verdana into the mix, which is good - but then it goes and applies pt
font sizes to some elements, which is not good. It seems to have
mucked-up its HTML syntax. It fails a few objective checkpoints of
the WAI content guidelines: though they're not very severe, it
indicates an unfortunate lack of attention to detail in something
that's supposed to be encouraging us to pay attention to detail, so I
don't think it's unfair to mention it.
But their content does seem to be compatible with the principles which
they promote, so by and large it seems to be a worthy operation, and
it is quite widely known, so an author would be able to put a
reference into their portfolio and have it recognised by potential
customers or employers as the case may be.
IMHO and YMMV, anyway.
Alan J. Flavell
> In most of the examples (such as the policeman in the news story)
> it's there to do nothing but re-enforce the impression given by the
> text for the visually orientated, and to break up the text into less
> imposing chunks, it's artistic. The picture rarely contains any
> additional content, alt="" would often be appropriate, but so are
> other things that could help to re-enforce the message.
Yes, I know that and you know that, but I'm trying to persuade the
questioner to prove it for themself by working it out from the
general principles, rather than taking my word for it.
cheers
D.M. Procida
> > A picture of an angry crowd in Karachi in a news report - information
> >
> > A stock picture of a Pakistani policeman in a news report about
> > police corruption in Karachi - no information, but it has content
>
> Despite Chris's patient attempt to explain it, and my less patient
> attempt to do the same, you seem to be proving the truth of of that
> comment by "Callie" (pen-name for an earlier contributor to
> c.i.w.a.h):
> For determining the appropriate alt text we don't really need to be
> told what your images look like, we *do* need to be told what you
> purposed by putting them there. And, short of us guessing on the
> basis of hints that you're dropping, that's still missing, as far as I
> can see.
As far as I can see, you're simply failing, or refusing, to make a
distinction between "content" and "information", and insist on
collapsing them together. As it happens a stock image of a Pakistani
policeman is pretty thin on content - it's not merely decoration, but
it's not very much more than decoration - but the in the wise owl
example we have a picture which is rich in content (allusive,
associative) but not informative content.
> > Unless you're using 'information' in a looser (and less satisfactory)
> > way than me,
>
> I prefer to use the term "content", but in a general sense
> "information" isn't really wrong. Whatever you call it, it's whatever
> you purposed by putting the image there. That's what the "alternative
> text" is needed for. For the other purposes you've still (modulo any
> practical issues) got the title= attribute and the longdesc= link, you
> shouldn't need to pollute the alternative text with what, for
> text-mode readers, are usually extraneous details, even if some of the
> text-mode users might want to take a specific interest in them, as
> we've previously discussed on this thread.
Going back to the wise owl picture, the purpose of the image is to
elicit those associations, without rendering the owl itself an object of
attention. If there is to be alt text, it needs to do the same kind of
work - not only the positive work of eliciting the associations, but
also diffusing expectations of the object itself.
:::Jerry::::
"Liz" <inv...@v-liz.co.uk> wrote in message
news:80ca26704d...@liz13.uklinux.net...
> In message <4292e58a$0$46354$892e...@authen.white.readfreenews.net>
> ":::Jerry::::" <m...@privacy.net> wrote:
>
> > alt="GoldenEagle picture" could be a picture of just that or the
Eagle
> > from the USA Presidential shield - to someone who doesn't have
the
> > image available, they will never know. IYSWIM
>
> But that would just be plain *wrong*, since the one on the USA
> presidential shield is a Bald Eagle.
I did wonder about that after I sent the message, but I think the
point I was making is still valid ?
:::Jerry::::
<Liz's post left un snipped intentionally>
I'm not sure if I follow what you are saying here, ISTM that you are
suggesting that the site must be understandable to (say) a seven year
old even though it's aimed at adults.
I'm really going to put the cat with the pigeons now... Surely
comprehension is to do with intelligence [1] and is _not_ an *access*
(disability) problem unless the person also has a disability that
causes low comprehension. As you imply, many people are able to read
any word put in front of them but they have never bothered to find out
what "those long complicated, technical sounding words" mean ?
[1] or at least an unwillingness to learn.
As for accessibility, should all web sites write their content in the
style of 'The Sun newspaper', but then perhaps "those long
complicated, technical sounding words" just need to have a URL
pointing towards dictionary.com [2] or perhaps they need to be
enclosed between span tags and have an alt description of the words
meaning?!....
[1] a bit like foxnews.com places 'search' links into their content,
such as 'President Bush (search) today said...'.
:::Jerry::::
"D.M. Procida" <real-not-anti...@apple-juice.co.uk> wrote in
message
news:1gx33u9.1jatxje1ix5emhN%real-not-anti...@apple-juice.co.uk...
<snip>
>
> As far as I can see, you're simply failing, or refusing, to make a
> distinction between "content" and "information", and insist on
> collapsing them together. As it happens a stock image of a Pakistani
> policeman is pretty thin on content - it's not merely decoration,
but
> it's not very much more than decoration - but the in the wise owl
> example we have a picture which is rich in content (allusive,
> associative) but not informative content.
I would suggest that it is you who is refusing to understand that
content is information, unless it is design. If you had no image of a
Pakistani policeman or a Wise Owl what textual content would you
include on the page that re enforces the points made that the
inclusion of the picture does, if there isn't a textual alternative
then the picture is nothing but decoration (and thus alt="" is the
correct way to go), if there is a textual alternative you are on your
way to deciding what should be in any alt text IYSWIM.
Andy Jacobs
j...@jibbering.com (Jim Ley) wrote:
> The problem I think is confusion about what "target audience" means...
<snip>
This came up in the office today. We're just doing a site for a company
that makes racing harnesses for formula one and other high profile
racing cars. Do we have to design the site to cater for or take into
account a visually impaired market?
--
Andy Jacobs
www.redcatmedia.net
Intelligent Websites For Intelligent Business People
Andy Jacobs
<1gx2yhr.miv90j1q4tlkbN%real-not-anti...@apple-juice.co.uk>,
real-not-anti...@apple-juice.co.uk (D.M. Procida) wrote:
> Chris Morris <c.i.m...@durham.ac.uk> wrote:
>
> > If you give me an example of an image of the sort you mean, *in its
> > real environment*, or at least in a plausible environment, then I can
> > come up with what I'd use for the alternative text, and tell you why.
> >
> > If you give me a broad category of images I can only give you a broad
> > category of guidelines and a recommendation to "think about it".
>
> Well, how about the Goya/Kant suggestion I made earlier, and if you want
> more context: the text is a sarcastic critique of the junk-culture
> fringes of post-structuralist thought and argues that we have a direct
> obligation to rationality. The picture (of _The Sleep of Reason_) is
> appears inline with a paragraph in which the phrase "the sleep of
> reason" is used of a certain prevailing tendency towards post-scientific
> medicine.
Have I missed something or did you just make all that up?
<abbr title="Smiley face represented by ascii characters">:o)</abbr>
Andy Jacobs
Stevie D <ste...@sjd117.freeserve.co.uk> wrote:
> Chris Morris wrote:
>
> > I throw that in firstly to demonstrate that the apparently obvious
> > "photograph of tutorial" isn't particularly helpful, and secondly to
> > demonstrate that - if you don't believe it's solely decorative, as
> > some people taking the course have reasonably claimed - picking good
> > alternative text for it requires a bit of thought.
>
> Another point to recognise, that is often lost, is that some people
> browse without images in general, but can choose to load them when
> they wish to - either turning images off in a normal GUI browser, or
> using a text-browser and calling a helper application. For images that
> are, er, <fx: trying to think of the right word> supplementary to the
> text [1], it is useful for these people to know (a) that there is a
> picture there, and (b) roughly what it might be.
And how many people do you know who do that? I've been in contact with
internet users every day in an commercial environment for the last 5
years and I've never met anyone, or heard of anyone, who browses with
images turned off. Even in the days when I was connecting at 9600 I
still browsed with them on.
:::Jerry::::
"Andy Jacobs" <an...@redcatmedia.net> wrote in message
news:andy-3828AD.2...@news.btinternet.com...
> In article <42938092....@news.individual.net>,
> j...@jibbering.com (Jim Ley) wrote:
>
>
> > The problem I think is confusion about what "target audience"
means...
> <snip>
>
> This came up in the office today. We're just doing a site for a
company
> that makes racing harnesses for formula one and other high profile
> racing cars. Do we have to design the site to cater for or take
into
> account a visually impaired market?
>
Yes, although the drivers (and pit crew staff) shouldn't have any
vision problems, what about office staff etc.
Away from the hype, motor sport is just another business....
Alan J. Flavell
> And how many people do you know who do that? I've been in contact with
> internet users every day in an commercial environment for the last 5
> years and I've never met anyone, or heard of anyone, who browses with
> images turned off.
I know someone who counts page accesses by counting a unique image
that the page loads. He's now convinced from his statistics that 100%
of WWW clients load images. Without exception. Presumably you can
see the "non sequitur".
> Even in the days when I was connecting at 9600 I still browsed with
> them on.
It only takes one exception to prove him (and you) wrong. Of course
from the server side you can't count "people". Some of the clients
presenting what look like browser client agent strings are really
bots, and some of the real people presenting strings that you don't
recognise are really browsers. I think you'll find that the majority
of users of text-mode browsers (w3m, links browser, and lynx, at
least, and some proportion of emacs-w3) are not automatically loading
images, even though some of them are in a position to load selected
images if they so choose.
Might amount to less than 1% of the total web user population, but
hey, if they just happened to be the ones who were interested in your
pages, you'd lose quite a number of prospective users/ clients/
customers/ whatever, by not taking them into account. And if you do
it properly, you're not going to cause the slightest harm to your
majority population. What they call nowadays a win-win situation.
D.M. Procida
> > Well, how about the Goya/Kant suggestion I made earlier, and if you want
> > more context: the text is a sarcastic critique of the junk-culture
> > fringes of post-structuralist thought and argues that we have a direct
> > obligation to rationality. The picture (of _The Sleep of Reason_) is
> > appears inline with a paragraph in which the phrase "the sleep of
> > reason" is used of a certain prevailing tendency towards post-scientific
> > medicine.
>
> Have I missed something or did you just make all that up?
No, I just made it all up. But I'm sure someone has done something along
those lines.
Liz
":::Jerry::::" <m...@privacy.net> wrote:
>
> "Liz" <inv...@v-liz.co.uk> wrote in message
> news:692834704d...@liz13.uklinux.net...
[low comprehension level]
>> ... something which even a lot of teachers don't seem to know: some
>> kids can read things and you'd think they had a reading age of 10 or
>> 12, when aged 16, but they can't actually understand a lot of the
>> words they can mechanically read.
>> In the school I work in (mainstream) I'd say at least 10% of the
>> leavers come into this category, and it's unlikely
>> to get better as they get older: in general they've had a lot of
>> one-to-one and small group help when in school, without any rise in
>> comprehension level. This *isn't* reading level: there are lots of
>> 'systems' whereby you can work out the 'reading level' of a piece of
>> text, e.g. by average number of letters in words or other formulae,
>> but this doesn't say anything *at all* about understandibility.
> I'm not sure if I follow what you are saying here, ISTM that you are
> suggesting that the site must be understandable to (say) a seven year
> old even though it's aimed at adults.
Not me personally: I'm generally speaking a utilitarian [1], but not
with a capital U and selectively.
I'm suggesting that people who constantly advocate accessibility for
all [2] generally totally forget that a largeish number of adults have
a comprehension age of seven or under.
> I'm really going to put the cat with the pigeons now... Surely
> comprehension is to do with intelligence [1] and is _not_ an *access*
> (disability) problem unless the person also has a disability that
> causes low comprehension.
Doesn't 'limited intelligence' (caused by e.g. genetics, oxygen
starvation at birth, foetal alcohol syndrome, brain damage caused by
accident/illness etc. etc.) count as a disability?
> As for accessibility, should all web sites write their content in the
> style of 'The Sun newspaper', but then perhaps "those long
> complicated, technical sounding words" just need to have a URL
> pointing towards dictionary.com [2] or perhaps they need to be
> enclosed between span tags and have an alt description of the words
> meaning?!....
You can't have had much experience with people with low comprehension.
I write worksheets for these pupils all the time, and it really isn't
easy. Putting the meanings of words as footnotes or sidebars or
bracketed in italics just doesn't do it, and while they're looking at
that, they've completely lost the thread of the sentence.
<rant>
We constantly get 'notes' from Learning Support about specific
learning difficulties as they occur with pupils - you just wouldn't
believe how many different 'syndromes' there are: Beattie used to say
you've got to have an ology, nowadays, you're no-one if you haven't a
syndrome.
</rant>
I don't know the answer, and neither do any of my colleages: as I
said, most think that doing a simple 'reading age' test on the
material is enough (it certainly isn't!). The issue doesn't even seem
to be mentioned at our two connected teacher training establishments
even nowadays (nothing changes).
I personally don't think that all sites should have to cater to all
people: I think they should have a target audience and cater to them,
so photography and art sites needn't cater to the blind (but without
deliberately excluding them) or, more particularly, to people who
choose to browse without images. The (fka Glasgow) Herald and the Sun
cater for different audiences, and no-one suggests that the Herald
should dumb down.
[1] Greatest good for the greatest number.
But then there was that uk electrical goods site mentioned on websites
that suck a couple of months back where you can't get in unless you're
using MSIE (I don't know about faking: I know that you couldn't get in
with default Opera, and apparently you can't get in with Firefox.
[2] <facetious>
Maybe I should complain because the Ferrari website has content which
is inaccessible to me because I'm (relative to Ferrari owners)
financially disabled.
</facetious>
Slainte
Liz (Grumpy Old Woman)
--
Alan J. Flavell
> > Frankly, this doesn't work:
> >
> > Apple Juice Ltd - Macintosh consultancy, service and repairs
> >
> > horizontal bar
>
> I wouldn't get too excited about that crappy old thing, which is indeed
> a complete mess.
You're entirely free to discuss how you'd propose to fix it
(particularly as you're calling attention to it in every posting), and
it might make an interesting case study for some of us to comment on.
Regardless of what you might think of us, we *are* trying to make
constructive proposals (aren't we, Chris? - Jim too, at least, and
others).
Stevie D
<Snipp-ex>
> I'm not sure if I follow what you are saying here, ISTM that you are
> suggesting that the site must be understandable to (say) a seven year
> old even though it's aimed at adults.
There is a significant proportion of adults in the UK who are
"functionally illiterate" - they are unable to read and comprehend
straightforward text such as an typical article in a mainstream
broadsheet newspaper. Quite possibly a higher proportion than those
with sufficient visual or motor problems to cause them difficulty
accessing the web.
If a website is to be fully accessible, it has to be accessible to
people of low mental acuity as well as to people with low visual
acuity.
That doesn't mean that you have to dumb down the content so that a
five year old could read it, if that isn't appropriate. What it does
mean is that, unless your site is aimed at readers of a high
intellect, you say "use" instead of "utilise", you don't write words
like "acuity" - keep the language as simple as you reasonably can do
without it becoming patronising or losing important nuances and
aspects of the message.
> I'm really going to put the cat with the pigeons now... Surely
> comprehension is to do with intelligence [1] and is _not_ an *access*
> (disability) problem unless the person also has a disability that
> causes low comprehension.
Who are you to say that stupidity isn't a disability? It is just as
wrong to discriminate against dumb people as it is against blind
people.
Andy Jacobs
"Alan J. Flavell" <fla...@ph.gla.ac.uk> wrote:
Ok, but being a bit flippant:
1. My current site probably does ok without images and my new site
certainly will.
2. If I haven't come across them yet and I'm still in business then I
guess they aren't a great part of my audience - bearing point 1 in mind.
Andy Jacobs
":::Jerry::::" <m...@privacy.net> wrote:
But then I think the chances of anyone visually impaired being involved
in the buying process where 'sight' of the product is probably mandatory
to specifying it properly is so small that it just ain't worth worrying
about. Even the DDA says that you can't get it right 100% of the time!
D.M. Procida
> > I wouldn't get too excited about that crappy old thing, which is indeed
> > a complete mess.
>
> You're entirely free to discuss how you'd propose to fix it
> (particularly as you're calling attention to it in every posting), and
> it might make an interesting case study for some of us to comment on.
I don't propose fixing it, actually. It can stay broken
> Regardless of what you might think of us, we *are* trying to make
> constructive proposals (aren't we, Chris? - Jim too, at least, and
> others).
D.M. Procida
> Not me personally: I'm generally speaking a utilitarian [1]
>
> [1] Greatest good for the greatest number.
Just the greatest good, actually. Not that it matters.
Daniele
D.M. Procida
> Alan J. Flavell <fla...@ph.gla.ac.uk> wrote:
>
> > > I wouldn't get too excited about that crappy old thing, which is indeed
> > > a complete mess.
> >
> > You're entirely free to discuss how you'd propose to fix it
> > (particularly as you're calling attention to it in every posting), and
> > it might make an interesting case study for some of us to comment on.
>
> I don't propose fixing it, actually. It can stay broken
Oops. Well, yes, it can stay broken, at least for the forseeable future,
and stay rubbish. It needs to be torn down completely and replaced.
> > Regardless of what you might think of us, we *are* trying to make
> > constructive proposals (aren't we, Chris? - Jim too, at least, and
> > others).
I've had pretty much nothing but constructive proposals from round here,
thanks very much. But why, what do you suppose I might think of you (or
anyone)?
Daniele
Liz
Stevie D <ste...@sjd117.freeserve.co.uk> wrote:
> What it does mean is that, unless your site is aimed at readers of a high
> intellect,
[snip]
And if it is, should we put:
"This site is aimed at readers of at least average IQ" as a disclaimer
at the top?
> you say "use" instead of "utilise", you don't write words
> like "acuity" - keep the language as simple as you reasonably can do
> without it becoming patronising or losing important nuances and
> aspects of the message.
Keeping language simple 'enough' for low comprehension readers almost
always loses important nuances, except where you use utilise for use
or 'at this point in time' instead of 'now'.
What if the 'essential content' of your site was poetry?
One of the things I notice lots of our kids don't 'get' most of the
time is figures of speech, and lots of even those nudging up towards
'normal' intelligence have difficulty with language or imagery which
isn't 'literal'. E.g. they'd wouldn't have a clue why there was an
image of an owl on a page about wisdom. But since they almost
certainly wouldn't know what 'wisdom' means that's a whole side issue.
Another thing a lot of people in general seem to find difficulty with
is analogy, which I naturally tend to use a lot, to my husband's
chagrin (I suspect this might be gender-biassed, but my experience is
hardly wide enough to be able to generalise.
Another thing which I seem to have naturally used forever is what I
discovered at uni to be the 'reductio ad absurdum' with which D also
shares a dislike with my pupils.
>> I'm really going to put the cat with the pigeons now... Surely
>> comprehension is to do with intelligence [1] and is _not_ an *access*
>> (disability) problem unless the person also has a disability that
>> causes low comprehension.
>
> Who are you to say that stupidity isn't a disability? It is just as
> wrong to discriminate against dumb people as it is against blind
> people.
Luckily, people who can't speak currently have few (no?) specific
problems in accessing the Web.
Slainte
Liz
--
Liz
j...@jibbering.com (Jim Ley) wrote:
> csszengarden is very inaccessible to me, I find it very hard to
> understand almost any of the examples,
Personally, I've never really been able to learn much from 'looking at
other people's source code'. That's one of the limitations of having a
'bookish brain'.
Like Andy, I'm reading the book just now.
It's very readable, and raisies some interesting general points, but
my CSS isn't moving on any - as I said, I just can't be 'hacked' with
the hacks!
Slainte
Liz
--
Alan J. Flavell
> > > Regardless of what you might think of us, we *are* trying to make
> > > constructive proposals (aren't we, Chris? - Jim too, at least, and
> > > others).
>
> I've had pretty much nothing but constructive proposals from round
> here, thanks very much. But why, what do you suppose I might think
> of you (or anyone)?
I was just trying to assure anyone interested, that even when my
patience threshold is exceeded, I'm still trying to be constructive.
btw, take a look at message-id
_8Nke.2292$oT1...@newsread1.news.pas.earthlink.net
http://groups.google.co.uk/group/comp.infosystems.www.authoring.html/msg/1a977de507be6e0e
if you like it that way.
We're not asking for anything to be done that would impair the results
for the mainstream user. It's about improving results for non-usual
browsing and other client scenarios. I can only stress again this
idea that it's a win-win situation.
Alan J. Flavell
> In message <42937da0....@news.individual.net>
> j...@jibbering.com (Jim Ley) wrote:
>
> > csszengarden is very inaccessible to me, I find it very hard to
> > understand almost any of the examples,
> Personally, I've never really been able to learn much from 'looking at
> other people's source code'. That's one of the limitations of having a
> 'bookish brain'.
csszengarden can be a fascinating source of ideas that are do-able
with CSS, as well as being a convincing demonstration of what can be
achieved by separating content/structure markup (in HTML) from
presentation proposals (in CSS). But almost none of the actual CSS
stylesheets offered are fit for use in production web sites, so they
have to be used as interesting proposals, to be adapted, if ever, for
production use - not as design templates.
cheers (IMHO and YMMV, natch)
Mark Goodge
keyboard and typed:
It's probably fairly uncommon now. It was a lot more common back in
teh days of slow modems - most of the people I knew who new how to
turn image loading off did so. The most common reason for turning
image loading off these days is when following a link sent round in an
email at work, just in case the resulting site isn't "office-safe" :-)
Mark
--
http://www.FridayFun.net - jokes, lyrics and ringtones!
"Well it's true today"
Mark Goodge
keyboard and typed:
>
>It only takes one exception to prove him (and you) wrong. Of course
>from the server side you can't count "people". Some of the clients
>presenting what look like browser client agent strings are really
>bots, and some of the real people presenting strings that you don't
>recognise are really browsers. I think you'll find that the majority
>of users of text-mode browsers (w3m, links browser, and lynx, at
>least, and some proportion of emacs-w3) are not automatically loading
>images, even though some of them are in a position to load selected
>images if they so choose.
In this context, it's also worth noting that the search engine spiders
are text-only browsers. Having the appropriate alt text (particularly
when an image is used as a link) will make a significant difference in
how well someone's site is indexed, and thus to the resulting page
rank.
When dealing with my clients, I've always used this as the main
argument - appealing to self-interest ("do you want your site near the
top of the search listings?") works better than arguments based on web
standards. Unfortunately.
Mark
--
http://www.FridayFun.net - jokes, lyrics and ringtones!
"We were meant to live for so much more"
Geoff Berrow
from Stevie D contained the following:
>That doesn't mean that you have to dumb down the content so that a
>five year old could read it, if that isn't appropriate. What it does
>mean is that, unless your site is aimed at readers of a high
>intellect, you say "use" instead of "utilise", you don't write words
>like "acuity" - keep the language as simple as you reasonably can do
>without it becoming patronising or losing important nuances and
>aspects of the message.
That's just a feature of good writing. You should read some sociology
books. They must practice being unintelligible.
--
Geoff Berrow 0110001001101100010000000110
001101101011011001000110111101100111001011
100110001101101111001011100111010101101011
D.M. Procida
> > > > Regardless of what you might think of us, we *are* trying to make
> > > > constructive proposals (aren't we, Chris? - Jim too, at least, and
> > > > others).
> >
> > I've had pretty much nothing but constructive proposals from round
> > here, thanks very much. But why, what do you suppose I might think
> > of you (or anyone)?
> We're not asking for anything to be done that would impair the results
> for the mainstream user. It's about improving results for non-usual
> browsing and other client scenarios. I can only stress again this
> idea that it's a win-win situation.
I don't need to be persuaded about the value of that end, or even about
the means of achieving it. I think the reasoning behind some particular
cases is not beyond dispute, that's all.
Chris Morris
Well, I do it. So you've heard of at least one now.
All the time in w3m and lynx (which I use on a daily basis), and if
I'm on a connection slower than ADSL, I turn images off in graphical
browsers. I'm obviously less patient than you.
--
Chris
:::Jerry::::
The people buying the product will be far more interested in the
published spec' than what it looks like, indeed they will probably be
telling your client how they want the equipment to look like [1],
IYSWIM.
[1] in terms of colour and branding etc.
:::Jerry::::
"Stevie D" <ste...@sjd117.freeserve.co.uk> wrote in message
news:jf47911jnn0r8cjjr...@4ax.com...
> :::Jerry:::: wrote:
>
> <Snipp-ex>
>
> > I'm not sure if I follow what you are saying here, ISTM that you
are
> > suggesting that the site must be understandable to (say) a seven
year
> > old even though it's aimed at adults.
>
>
> That doesn't mean that you have to dumb down the content so that a
> five year old could read it, if that isn't appropriate. What it does
> mean is that, unless your site is aimed at readers of a high
> intellect, you say "use" instead of "utilise", you don't write words
> like "acuity" - keep the language as simple as you reasonably can do
> without it becoming patronising or losing important nuances and
> aspects of the message.
Sorry, but to those who can read to the level of the Telegraph,
expecting them to read The Sun is highly patronising... I really do
think what you say above would be dumbing down content and the 'feel'
of a site.
>
> > I'm really going to put the cat with the pigeons now... Surely
> > comprehension is to do with intelligence [1] and is _not_ an
*access*
> > (disability) problem unless the person also has a disability that
> > causes low comprehension.
>
> Who are you to say that stupidity isn't a disability? It is just as
> wrong to discriminate against dumb people as it is against blind
> people.
>
Well, unless there is an undisclosed medical reason (such as dyslexia)
why are they 'stupid', and should they not be encouraged to improve
rather than just have Janet & John style publications served up?
Andy Jacobs
Morris" <c.i.m...@durham.ac.uk> wrote:
But then are you typical of the market that I'm selling into? Probably not.
I would say that most of the people here can be excluded from any statistics
as we're probably much more aware of technology than the average internet
user and will just represent spikes on a graph (if it could be plotted).
Andy
Andy Jacobs
4n5891he46jajsd9l...@news.markshouse.net, "Mark Goodge"
<use...@listmail.good-stuff.co.uk> wrote:
But then I spent most of my initial days on the net browsing the sort of
sites where images were integral to the experience... If you get my drift
;o)
Andy
Chris Morris
> In the UK context one might take a cue from the plain english folks
> and their "crystal mark". http://www.plainenglish.co.uk/
For many things, yes. I think they have a tendency in certain cases to
over-simplify. The WAI says "...clearest and simplest..." which
perhaps is a recognition that after a certain point the two can be
contradictory.
From their website, for example, there's:
http://www.plainenglish.co.uk/translations.html where they've transformed
"...Posters and leaflets issued by the Central Office of
Information, the Department of Health and Social Security and other
authoritative bodies are usually displayed in libraries, but items of
a disputatious or polemic kind, whilst not necessarily excluded, are
considered individually."
into
"...Before we can give you an answer we will need to see a copy of the
posters to make sure they won't offend anyone."
Now, I'm not disputing that the initial text is unnecessarily complex
and would benefit from heavy rewording. *However*, the suggested
changes go too far and _change_ the meaning of the text by removing
and not replacing "whilst not necessarily excluded".
So, simplicity *and* clarity are both important. I think - when a
compromise needs to be made - I'd always choose the latter. Someone
not being able to read some text is probably better than someone
*thinking* they can read it, and getting entirely the wrong idea.
--
Chris
(under no illusions that this post is either simple or clear, of course)
Chris Morris
> I'm not sure if you can really meld design with accessibility, and
I've never been convinced that there's any real problem there. While
the WCAGuidelines aren't exactly complete, very few of them impose
limitations on the design.
The main three that do are:
1) Have sufficient colour contrast
2) Use relative units for font size and layout size
3) Don't use images where text or markup could do the same thing
I'd add to that, as things that aren't in the WCAGuidelines, but I
still consider important.
4) Use a sensible base font size and don't go below 85% of browser
default at all.
5) Distinguish visited and unvisited links and ideally stick to
the blue/purple of common browsers.
I don't think there's any reason a competent visual designer can't get
a good design within those constraints. There is a tendency in modern
graphic design to break 1) repeatedly, and most print-background
designers find it difficult to get their heads around 2)
--
Chris
Alan J. Flavell
> But then are you typical of the market that I'm selling into?
It would cost you more to evaluate that than to simply follow accepted
web authoring standards. So just do it, and stop wasting your time
trying to find excuses.
> Probably not. I would say that most of the people here can be
> excluded from any statistics
You'd be excluding indexing bots too, you know.