img alt attribute & photo galleries
Headless
thumbnails that link to html pages that contain the full size images.
Both the thumbnails and the full size images are content and I suppose
should therefore use the alt attribute for the images.
But I'm wondering what good this would do. Both the thumbnail index
pages and the full size image pages are irrelevant to the visually
impaired, and the sighted who initially browse with images off will have
to enable graphics mode to see them. I can't see a string of
alt="X smiling and wearing a green dress" being helpful.
Headless
Jukka K. Korpela
> Both the thumbnails and the full size images are content and I suppose
> should therefore use the alt attribute for the images.
At the end - i.e., _after_ analyzing how the page might work - the alt
attribute must be specified for each and every image. But you presumably
mean an alt attribute with a nonempty value.
> But I'm wondering what good this would do.
Scenario 1: I'm using a text-only browser and I just want to download and
save some images, which I will view later (or now, if I'm using a text-only
browser that can spawn an image display program). I surely appreciate
readable link texts.
> Both the thumbnail index
> pages and the full size image pages are irrelevant to the visually
> impaired,
Are they? Most visually impaired people are not blind. Scenario 2: I can
see, but not very well, and I will need extra work to view images. I would
surely like to know what each image is about.
> and the sighted who initially browse with images off will have
> to enable graphics mode to see them.
No, (scenario 3:) I can view the images _individually_, perhaps by clicking
on the area reserved for the image and selecting "View image".
> I can't see a string of
> alt="X smiling and wearing a green dress" being helpful.
It depends. If it's just a collection of home photographs, then it probably
doesn't matter. But mostly images benefit greatly from image captions. And
you could then make the image and the caption together a link, and _then_
you can say alt="" since the info is in the caption, e.g.
<div><a href="bigfoot.jpg"><img alt="" src="smallthumb.jpg">
Bigfoot as photographed in Calisota</a></div>
Scenario 4 (mostly futuristic, but not too far-fetched): I'm completely
blind but I can use equipment that lets me experience an image with my
fingertips, so that effectively different degrees of darkness in the image
are mapped to variations of the form of a surface, or to simulations of
such variation. I'd surely like to know about an image in text before I
invoke such presentation (which will probably be rather slow, in initial
implementations at least).
--
Yucca, http://www.cs.tut.fi/~jkorpela/
Pages about Web authoring: http://www.cs.tut.fi/~jkorpela/www.html
Headless
>Scenario 2: I can
>see, but not very well, and I will need extra work to view images. I would
>surely like to know what each image is about.
>
>(scenario 3:) I can view the images _individually_, perhaps by clicking
>on the area reserved for the image and selecting "View image".
I had thought of scenario 3 but this presumes that it's possible to come
up with meaningful descriptions. For more than 95% of the images in my
photo gallery this would be difficult if not impossible. I can come up
with a description but it would be like:
X in a creme top
X wearing jeans and a red top
X sitting wearing jeans and a red top
X in front of a window wearing a creme top
It's not a fashion site, so the descriptions of the clothing is afaics
meaningless, that leaves little or no meaning. Unlike your bigfoot
example I have no information on where or when the images were shot.
I'm beginning to think that if it's impossible to come up with a
meaningful description then images should actually be considered
decorative. That seems to suggest that using alt="" is appropriate.
Headless
Lauri Raittila
> "Jukka K. Korpela" <jkor...@cs.tut.fi> wrote:
> I had thought of scenario 3 but this presumes that it's possible to come
> up with meaningful descriptions. For more than 95% of the images in my
> photo gallery this would be difficult if not impossible. I can come up
> with a description but it would be like:
>
> X in a creme top
> X wearing jeans and a red top
> X sitting wearing jeans and a red top
> X in front of a window wearing a creme top
>
> It's not a fashion site, so the descriptions of the clothing is afaics
> meaningless, that leaves little or no meaning. Unlike your bigfoot
> example I have no information on where or when the images were shot.
How about just "photo of X". It is usually interesting who is in the
picture. Of course, if you meant with X that name is also unknown, it
don't help much. But it would be better to make that caption.
I'am doing some changes in my student guild's website, and I see that it
would greatly if even the persons name would be available, someway.
Especially when you are looking for picture of someone.
> I'm beginning to think that if it's impossible to come up with a
> meaningful description then images should actually be considered
> decorative.
It's not decorative just because there is no good textual alternative.
And usually there is, you just don't know it. Like "X in the party on Y's
house 21.2.1981". If you don't know the alternative, it doesn't mean it's
decoration.
But this kind of thing usually should be caption anyway.
<img alt="Photo of">
X in the party on Y's house 21.2.1981
> That seems to suggest that using alt="" is appropriate.
Sometimes. But if you don't have other content than images, it is not.
--
Lauri Raittila <http://www.iki.fi/lr> <http://www.iki.fi/zwak/fonts>
Saapi lähettää meiliä, jos aihe ei liity ryhmään, tai on yksityinen
tjsp., mutta älä lähetä samaa viestiä meilitse ja ryhmään.
Andy Dingley
wrote:
> I can't see a string of
>alt="X smiling and wearing a green dress" being helpful.
Wrong attitude.
You should never (as a general rule of design, for anything, not just
the web) try to limit usage of your product, just because you can't
think of a need for it.
If it's do-able (i.e. technically feasible, and not an excessive cost)
then you should do it. Only if there's an identified downside to
adding these alts or titles, should you start thinking about taking
them off.
As an obvious example here, what about indexing spiders ?
Andreas Prilop
> Both the thumbnails and the full size images are content and I suppose
> should therefore use the alt attribute for the images.
> But I'm wondering what good this would do. Both the thumbnail index
> pages and the full size image pages are irrelevant to the visually
> impaired,
Is Google "visually impaired"? You should know that
<http://images.google.com/> indexes ALT texts.
BTW: <http://www.google.com/> does not index ALT texts.
--
Top posting.
What's the most irritating thing on Usenet?
Headless
>How about just "photo of X".
All images are of X, this would result in imo non meaningful alt tags.
Example of the type of images:
http://www.suziperry.utvinternet.ie/photos_1_ch.htm
Headless
Headless
>Is Google "visually impaired"? You should know that
> <http://images.google.com/> indexes ALT texts.
I hope you're not suggesting that I should use non descriptive alt text
to get indexed by a picture search engine?
Headless
Lauri Raittila
> Lauri Raittila <la...@raittila.cjb.net> wrote:
>
> >How about just "photo of X".
>
> All images are of X, this would result in imo non meaningful alt tags.
Why (repetative, but why is that nonmeaningful)? If someone actually
wants to look them with text based browser, and you say alt="", he may be
unable to download the images. If they are all from X, then "photo" with
possible running number would be better than nothing.
I would maybe need to do download these, when sending image to as email
attachments from public library, where I have no access to anything but
the browser. -> easiest way to get the image will be getting it to remote
computer using web based SSH with lynx/links, and then send it as email.
In this scenario, it doesn't even do harm if all alt texts are same.
> Example of the type of images:
> http://www.suziperry.utvinternet.ie/photos_1_ch.htm
See, try looking it whitout images, because you have alt="", you have
nothing sensible. So how do you download the image(s)?
Otherwise very nice page. Exept in:
http://www.suziperry.utvinternet.ie/photos.htm
Many people would think that clicking the image gets bigger version of
it. - Make the text also link? (I supposed it is your page)
http://www.guymacon.com/ "02@spamcop.net Guy Macon
Andreas Prilop wrote:
>
>Headless <invalid...@dna.ie> wrote:
>
>> Both the thumbnails and the full size images are content and I suppose
>> should therefore use the alt attribute for the images.
>> But I'm wondering what good this would do. Both the thumbnail index
>> pages and the full size image pages are irrelevant to the visually
>> impaired,
>
>Is Google "visually impaired"? You should know that
> <http://images.google.com/> indexes ALT texts.
>
>BTW: <http://www.google.com/> does not index ALT texts.
http://www.google.com/ does, however, read the ALT text and uses
it to help it decide what pages to display for a particular
search term.
Andreas Prilop
> http://www.google.com/ does, however, read the ALT text and uses
> it to help it decide what pages to display for a particular
> search term.
Could you please explain in more detail? Some examples, perhaps?
Andy Mabbett
<invalid...@dna.ie> writes
>Example of the type of images:
"MS Internet Explorer is not capable of displaying this site properly.
Why not try a better browser like Opera or Mozilla"
Anyone care to translate that..?
--
Andy Mabbett
Women are only able to visit Kuwait's stock exchange on a mezzanine floor
called the "Ladies Trading Floor". In Iraq, women can participate fully in the
surprisingly booming Baghdad stock exchange. Guardian 4 Feb 2003.
Headless
>> http://www.suziperry.utvinternet.ie/photos_1_ch.htm
>
>See, try looking it whitout images, because you have alt="", you have
>nothing sensible. So how do you download the image(s)?
Let's assume I add alt="Picture of X" to all images. Visitor Y lands on
the site with Opera, Y has images disabled because that's faster, Y
arrives at one of the thumbnail pages, Y sees a lot of links all labeled
"Picture of X".
This isn't going to help Y to decide which images he wants to right
click and download.
Now let's examine another scenario: another picture thumbnail page with
these alt texts:
Johnny age 4 blowing out the candles on his birthday cake
Johnny's new bike he got for his 4th birthday
Johnny crying after he fell off his new bike
Johnny sound asleep after his 4th birthday party
These descriptions allow Y to determine in advance which pictures he
wants to right click and download, no real need to switch on graphics in
his browser.
Johnny's photos tell a story that can be "translated" into meaningful
alternative text.
Now back to my photo gallery: basically all head shots of the same
person, they should be considered content because the images show the
facial expression, body posture, clothing etc, these things convey
subtle information about that person. But the story that these images
tell is not something that can be "translated" into meaningful
alternative text imo.
Which brings me to believe that unlike the photos of Johnny's birthday
party, this type of photo gallery can only be experienced with graphics
enabled, if that's true then alt="" seems appropriate to me.
Headless
Headless
>"MS Internet Explorer is not capable of displaying this site properly.
>Why not try a better browser like Opera or Mozilla"
>
>Anyone care to translate that..?
An expression of my annoyance with IE's poor support for standards
followed by a plug of 2 alternatives.
Headless
Dave Patton
news:l13l6vsfgbkoo7dn8...@4ax.com:
> Now back to my photo gallery: basically all head shots of the same
> person, they should be considered content because the images show the
> facial expression, body posture, clothing etc, these things convey
> subtle information about that person. But the story that these images
> tell is not something that can be "translated" into meaningful
> alternative text imo.
>
> Which brings me to believe that unlike the photos of Johnny's birthday
> party, this type of photo gallery can only be experienced with graphics
> enabled, if that's true then alt="" seems appropriate to me.
I'd suggest that a series of text such as "Fred, headshot 1",
"Fred, headshot 2", "Fred, headshot 3", etc would work.
--
Dave Patton
Canadian Coordinator, the Degree Confluence Project
http://www.confluence.org dpa...@confluence.org
My website: http://members.shaw.ca/davepatton/
Lauri Raittila
> Lauri Raittila <la...@raittila.cjb.net> wrote:
>
> >> http://www.suziperry.utvinternet.ie/photos_1_ch.htm
> >
> >See, try looking it whitout images, because you have alt="", you have
> >nothing sensible. So how do you download the image(s)?
>
> Let's assume I add alt="Picture of X" to all images. Visitor Y lands on
> the site with Opera, Y has images disabled because that's faster, Y
> arrives at one of the thumbnail pages, Y sees a lot of links all labeled
> "Picture of X".
>
> This isn't going to help Y to decide which images he wants to right
> click and download.
No, but Y may wan't to download them all, or at least some. And he might
have graphical browser, from which she can look which image is which as I
explauined earlier.
> Now let's examine another scenario: another picture thumbnail page with
> these alt texts:
>
> Johnny age 4 blowing out the candles on his birthday cake
> Johnny's new bike he got for his 4th birthday
> Johnny crying after he fell off his new bike
> Johnny sound asleep after his 4th birthday party
>
> These descriptions allow Y to determine in advance which pictures he
> wants to right click and download, no real need to switch on graphics in
> his browser.
Yes, that would be much better, but the "photo of X" still is better than
nothing.
> Now back to my photo gallery: basically all head shots of the same
> person, they should be considered content because the images show the
> facial expression, body posture, clothing etc, these things convey
> subtle information about that person. But the story that these images
> tell is not something that can be "translated" into meaningful
> alternative text imo.
That's somewhat true. But the fact there is more than one picture, that
would get same lousy alt text, dosen't mean that this lousy alt text
would be meaningless. As it is image callery, they wan't to see
images/alt texts, even if they can't display images. They wouldn't go to
image gallery otherwise.
> Which brings me to believe that unlike the photos of Johnny's birthday
> party, this type of photo gallery can only be experienced with graphics
> enabled, if that's true then alt="" seems appropriate to me.
Alt="" is only good when image is decorative, or when possible
information is presented as text. IMO "photo" is better than "", even if
there is lots of images.
Headless
>Alt="" is only good when image is decorative, or when possible
>information is presented as text. IMO "photo" is better than "", even if
>there is lots of images.
I'm not convinced, I read back Jukka's recent posting on the subject and
it contains a clear pointer to what to do in situations like this:
>And finally, there are
>images that have content that cannot be described in words, such as a
>picture of Mona Lisa; then say that, and name the image in the ALT text so
>that it becomes obvious that the ALT text is not meant to be taken as such
>but as a reference to essentially visual information.
This seems the best way to do it imo, but I'll have to think about
translating that principle into an actual text.
Headless
Jukka K. Korpela
>> http://www.google.com/ does, however, read the ALT text and uses
>> it to help it decide what pages to display for a particular search
>> term.
>
> Could you please explain in more detail? Some examples, perhaps?
It's a bit difficult to find examples, since authors who write useful alt
texts tend to use similar words in the textual content as well, so the page
would be found anyway. Besides, Google may also use URLs - after all, quite
often URLs are named so that when taken as words, they reflect actual
content, and Google seems to find lone images on such grounds, i.e. the
results contain URLs of images that are not presented as belonging to any
HTML page (though it's difficult to say whether it actually found the
images via src attributes on HTML pages).
But it seems to me that Google's normal search does _not_ use alt texts
whereas Google's image search http://images.google.com/ does. I did a
simple test with the string
"ruusukuoriainen muistuttaa"
(with the quotes) as the search string, and http://www.google.com/ did not
find anything whereas the image search found an old page of mine where that
particular string occurs in an alt text only.
Another example: image search with "ruutukaappaus" ('screen shot' in
Finnish) finds http://www.ooodocs.org/dictinstall/fi/ where this word
occurs in an alt text only.
This isn't very regular though. I would expect that e.g. image search with
"gorod helsinki"
would find http://www.helsinki.fi/ which has alt="Gorod Helsinki", but it
doesn't find anything.
Andreas Prilop
> But it seems to me that Google's normal search does _not_ use alt texts
> whereas Google's image search http://images.google.com/ does.
That's exactly what I found.
Therefore I wonder what
| http://www.google.com/ does, however, read the ALT text and uses
| it to help it decide what pages to display for a particular
| search term.
<news:v6knhlr...@corp.supernews.com>
should mean.
Andreas Prilop
> This isn't very regular though. I would expect that e.g. image search with
> "gorod helsinki"
> would find http://www.helsinki.fi/ which has alt="Gorod Helsinki", but it
> doesn't find anything.
images.google.com hasn't indexed anything from the main page /index.html.
<http://images.google.com/images?q=kuva02+site:helsinki.fi>
--
ObJoke:
Ville d´Helsinki
Jim Ley
<la...@raittila.cjb.net> wrote:
>But this kind of thing usually should be caption anyway.
>
><img alt="Photo of">
>X in the party on Y's house 21.2.1981
I'm rather wary of this using alt as part of the caption, since it's
clearly only being used for the situation where the image cannot be
seen, and not for where it cannot be understood - whilst with the alt
above that's probably perfectly safe, I don't think it always is - and
when I interogate the image I cannot understand for the ALT
information half a sentance would tend to be confusing.
Jim.
--
comp.lang.javascript FAQ - http://jibbering.com/faq/
Jukka K. Korpela
> On Sat, 8 Mar 2003 13:34:13 +0200, Lauri Raittila
> <la...@raittila.cjb.net> wrote:
>
>>But this kind of thing usually should be caption anyway.
>>
>><img alt="Photo of">
>>X in the party on Y's house 21.2.1981
>
> I'm rather wary of this using alt as part of the caption, since it's
> clearly only being used for the situation where the image cannot be
> seen,
I think we are equally strongly against such alt attributes, though for
rather different, and largely opposite, reasons. First, I would like to
point out and emphasize that the alt attribute is by definition for use
when the image is not displayed. (This is not the same as "cannot be seen",
though this is not much of an issue here.)
> and not for where it cannot be understood - whilst with the alt
> above that's probably perfectly safe, I don't think it always is - and
> when I interogate the image I cannot understand for the ALT
> information half a sentance would tend to be confusing.
Maybe, but what's wrong then is the idea that the alt information should be
something to be interrogated for, especially if you think it would be sort
of a title for the image in isolation - which is, more or less, what the
title attribute is supposed to be.
What's wrong with alt="Photo of" in my view is that is not an adequate
replacement for the image in contexts where the image is not seen. Even
when taken in conjunction with the caption text that follows, it does _not_
perform the function that the image is expected to perform when seen.
In some cases, alt="There is photographic evidence of the fact that X was
present in the party at Y's house 21.2.1981." could be adequate. In some
other cases, alt="X behaved really wildly in the party at Y's house
21.2.1981." would do. Unfortunately we have no simple way of telling that
the alt text gives just _part_ of the message of the photo.
In a normal photo gallery, I would use just attributes like
alt="[Photo of X in the party at Y's house 21.2.1981]"
in lack of anything more adequate, irrespectively of the potential presence
of a caption text with similar content. The brackets are the best way I
know to tell 'the message of this image cannot be expressed in words; a
description of the image follows: ...'.
(In practice, such attributes would best be generated from the same source
as captions. If captions are omitted, because the images are "obvious", it
would be a good idea, in principle at least, to generate title="X in the
party at Y's house 21.2.1981" as well.)
I guess such alt texts would, incidentally, correspond to your idea of
using alt attributes explanations of images. It's also helpful to someone
who uses, say, a text browser for loading some images - he might be able to
work on the alt texts alone.
Naturally, a photo gallery page should be clearly indicated as such, so
that e.g. people using speech browsers won't need to listen to the page for
several minutes before they realize that it has nothing but photos.
Jim Ley
<jkor...@cs.tut.fi> wrote:
>j...@jibbering.com (Jim Ley) wrote:
>
>> On Sat, 8 Mar 2003 13:34:13 +0200, Lauri Raittila
>> <la...@raittila.cjb.net> wrote:
>>
>>>But this kind of thing usually should be caption anyway.
>>>
>>><img alt="Photo of">
>>>X in the party on Y's house 21.2.1981
>>
>> I'm rather wary of this using alt as part of the caption, since it's
>> clearly only being used for the situation where the image cannot be
>> seen,
>
>I think we are equally strongly against such alt attributes, though for
>rather different, and largely opposite, reasons. First, I would like to
>point out and emphasize that the alt attribute is by definition for use
>when the image is not displayed.
We'll go over old ground if we argue this one, but I don't accept that
ALT is purely for when the image is not displayed - to me it's for
when the information in the image is not available to the user. This
is not clear by the HTML 4.01 definition of the attribute, however I
think we would agree that the WAI guidelines on how to author provide
more detail on how attributes should be used to their best ability,
and this is clearly what I get from the WAI documents.
>Maybe, but what's wrong then is the idea that the alt information should be
>something to be interrogated for
How else should it be made available?
>especially if you think it would be sort
>of a title for the image in isolation - which is, more or less, what the
>title attribute is supposed to be.
Yes, but a title is liking to only by accident give any detail on the
information contained in the image - that is what the ALT is for.
>What's wrong with alt="Photo of" in my view is that is not an adequate
>replacement for the image in contexts where the image is not seen.
Of course not, we can certainly agree on that, I was also going
further and suggesting it was misleading in the situation where the
author has decided the cost of providing high quality alt text
outweighs the benefits.
>I guess such alt texts would, incidentally, correspond to your idea of
>using alt attributes explanations of images.
Not incidentally, those ALT texts used correctly are the suitable
"explanations" - I'm not asking for anything but what the alt
attribute is for, just not with the narrow vision that it's only see
and not understand that matters to ALT.
Andy Mabbett
Korpela <jkor...@cs.tut.fi> writes
>In some cases, alt="There is photographic evidence of the fact that X
>was present in the party at Y's house 21.2.1981." could be adequate. In
>some other cases, alt="X behaved really wildly in the party at Y's
>house 21.2.1981." would do. Unfortunately we have no simple way of
>telling that the alt text gives just _part_ of the message of the
>photo.
Part of the problem being that the image will convey different meaning
to different viewers.
To the photographer/ site manger, it might say "X was at the party". To
he host of the party, it might say "I'm so glad X made the effort to
travel so far to celebrate my birthday"; to X, it might say "my bum does
look big o that!", to anther attendee, it might say "I remember, that
was when I first kissed X", and to someone who as not at the party; "I
wish I'd been there".
To another photographer; the meaning might be "what poor focussing", and
to a random surfer, who has never heard of these people, let alone the
party, it might say "that wallpaper doesn't go with those curtains". And
so on...
--
Andy Mabbett
Not in my name: <http://www.stopwar.org.uk/>