The answers to the �Looking for a good interface developer?� questions Followup (now with Javascript!) to thi [...]

Personally, I think information put in the alt shouldn’t necessarily be of use to all users, but only to those who don’t see the image.

That’s exactly my point because if the information put in alt attributes is for all users (I mean both sighted and visually impaired, whatever their browser), then IE’s tooltip will benefit only IE users.

(Maybe I am not writing clearly.)

Interesting question. Some people think it’s up to the interface developer to know about user focus and interactivity while others think that coding is the only aspect.

I tend to think an interface developers should know both, but depending on what company/assignment we’re talking about, there may also be people with advanced degrees in cognitive thinking, pure interaction designers with numerous years of experience and so on also present in the situation.

So’ I’d say that in some cases it’s a must-know, but not in all.

Aaron,

Thanks!

Jules,

Well, that’s the gist of it. Personally, I think information put in the alt shouldn’t necessarily be of use to all users, but only to those who don’t see the image.

Stijn, Nick,

Nick, you answered Stijn about the same thing I would’ve about that topic; and no, Stijn, it’s not off-topic, but instead good that you ask.

Don’t wory about not continuing the discussion; I think it’s good to talk about and ventilate different opinions. So, feel free to go on!

Gustaf,

Well, I felt obliged to answer them myself.
Also, I’m happy to hear that they were of use to you!

Anyway, I actually made good use of some of your questions today during a few interviews I held.

I wish the W3C would make this point more clear though, perhaps in an erratum.

Anyway, it’s time to be off to the pub

If you accept the logical validity of the corollary I present in my second paragraph

But I do not

You are right that the specs should be extrapolated at some points. However, I do not agree with you in this specific case, at least not with the way you interpret them.

…as the specs do not explicitly state it’s forbidden…

If you accept the logical validity of the corollary I present in my second paragraph, then the specs implicitly state that it’s forbidden. Specifications and standards should not be interpreted solely in terms of the specific words on the page, but in light of the logical implications thereof. Otherwise one can argue that almost anything one cares to think of is permitted.

It isn’t reasonable to conclude that, because no standard explicitly says “Don’t do that”, that it is therefore permissible to do so.

It isn’t reasonable to conclude that, because no standard explicitly says “Don’t do that”, that it is therefore permissible to do so. As far as I know there is no standard forbidding me from writing a rendering engine that displays a large animation of a bouncing rabbit whenever the user types the letter “Z” in a form, but the fact that I wasn’t explicitly told not to do it doesn’t mean it’s to be done.

(~ Nick Fitzsimons)

No, of course not, but that doesn’t mean you shouldn’t only do what the specs suggest and nothing else. I see serving alt as an alternative (rly), and title as an attribute that gives information that’s not immediately obvious from the image only – and the specs seem to agree with me (though they are somewhat vague about title to me).

But, when you do not understand what the image is supposed to mean (because you’re plain dumb, or for whatever reason, it happens to me sometimes) checking out the alt text might be useful. I find the tooltip the best way for this, and as the specs do not explicitly state it’s forbidden I don’t see much problems with it. Heck, even when they stated it was forbidden I wouldn’t seemuch problems with it, though that’s a whole different discussion Note that Mozilla’s explaination (linked to by Robert) does not say anything about specifications but gives the way web developers abuse the attribute as a reason not to show alt as a tooltip.

I do agree that that is a real problem; alt being used as title. However, I don’t find that a good reason not to display it as a tooltip. Mozilla doesn’t seem to have making every web developer code in a standards/semantics-aware way as goal, or they would’ve dropped support for a lot of legacy elements (embed for example) too, so I don’t see why they would make an exception for this.

It seems to me that the obvious corollary is that alternate text should not serve as content when the element can be rendered normally. Presenting the alternate text in a tooltip is making it serve as content, even if it is dynamic content.

It isn’t reasonable to conclude that, because no standard explicitly says “Don’t do that”, that it is therefore permissible to do so. As far as I know there is no standard forbidding me from writing a rendering engine that displays a large animation of a bouncing rabbit whenever the user types the letter “Z” in a form, but the fact that I wasn’t explicitly told not to do it doesn’t mean it’s to be done.

Given that the title attribute exists, and that HTML 4.01 specifically suggests its use for tooltips, it is obvious that any requirement for tooltip text is already met, and IE is wrong to misuse the alt attribute in this way.

Or does this count as hijacking the post and going way off-topic?

If so, then I was the one to start it.

I can’t find the reference to it but apparently the W3C does state that alt text should not appear as a tooltip, that “behaviour” should be reserved for the title attribute.

Or does this count as hijacking the post and going way off-topic?

The problem with placing useful information in the alt attribute (let’s assume that the information put there really is useful for all users), then users who aren’t using IE don’t see the information unless they disable the image.

BTW, I wrote a comment in the book about IE displaying alt text as a tooltip and after thinking about it, I came to the following conclusion. Alt text as a tooltip, for designers who focus only on IE as their target browser, is dangerous because it intices them to make use of the alt attribute to present information that is not a text alternative to the image such as “Click here for more hockey skates” or worse in an image map showing a province with hotspot icons representing mine locations, “Click on one of the mining symbols for more information about that particular mine” (I was guilty of that in years past).

Unobtrusive JavaScript and the like is a part of this, but also things like the knowledge you expect of users (whether they understand a capsule-shaped “Send”-text is something clickable without explicit hints, for example), and well, a lot more. For those interested I recommend Interaction Design by Preece et al.