Alt attribute for images on the HP empty / HTML validator gives up on HP

[birkenbihl]

The first thing that comes into mind wrt accessibility is the "alt" attribute for images.
Good: all images on the HP have an alt attribute - which keeps validators happy.
Bad: all alt attributes are empty. Which perverts their purpose.
Speaking about validators (or standard compliance): After feeding https://www.internetsociety.org/ to W3Cs validator the validator ends with "Error 6: Fatal Error: Cannot recover after last error. Any further errors will be ignored."
It took 2 min to find these severe issues on the HP. Hard to believe that somebody seriously checked the site.

[danyork]

Klaus, Many thanks for this report. It prompted our team to identify that there was a misconfiguration in one of the test tools being used. That's now been fixed and we're going back through to identify and fix issues.

[birkenbihl]

a) mis-configured test tools might explain the validator crash though ...
but
b) providing meaningful text substitutes for pictures (aka alt attribute) should be part of the authoring rather than something that is done during final clean-up or testing. Since you use wordpress (and probably include the pictures from the WP media library): every item in the media library should get a text substitute when is imported into the library. So you even shouldn't wait until you make use of it.
btw. Dan, would you mind to let me know which test tools you use?

[danyork]

Klaus, totally agree that adding alt tags should be part of the authoring process and we are doing that for all new images. We do have a need to go back and fix many of the imported images. Some are missing alt tags or have non-useful alt tags that are just the file names.

The tool our developers are using for testing is AXE - https://www.deque.com/products/axe/

FYI, we have gone through and made a number of updates. More to come.

[danyork]

Klaus, Which W3C validator are you using? (The one that crashed.) I went to https://www.w3.org/WAI/ER/tools/ and there is a long list of tools. I couldn't see one specifically from the W3C. I tried https://validator.w3.org ... but that is only about HTML compliance, not accessibility (and it completed fine).

Also, at this point, there are ALT tags on all images on the home page. A couple could be improved, but they are there. Can you please confirm you are seeing this? Thanks.

[birkenbihl]

Hm, there is a suggestion in the FF-web-developer-tools. (sortcut shift-alt-v) that ends up at https://validator.w3.org . The results for the page are here https://validator.w3.org/nu/?doc=https%3A%2F%2Fwww.internetsociety.org%2F. And no, I still see the image below.
Taking a closer look the validator obviously gets into trouble by a rather simple error. Which is: an "img" tag in the "noscript" in the html "head". It obviously lets the validator assume a missing a "body" tag before the "img" tag which in turn result in an error for the "/noscript" tag and the "/head" tag and the following "body" tag. And then - pity - the validator gives up.

But yes you are right: AFAIK there is no W3C/WAI validator at W3C - at least there weren't as I asked last time. I assume the rational for this are the limitations of automatic WAI conformance checking (which might change in the future).

And yes: no empty alt-attributes any more.

[birkenbihl]

just checked: https://www.facebook.com/business/help/952192354843755:
surprise: facebook is the author of the flagged code. It is one by one copied from there.
This said (and knowing why I block access to fb completely in my regular browser profile): I wonder if at least ISOC should not abstain from this kind of tracking games that are in favour of FB on the cost of visitor's privacy..

[danyork]

@birkenbihl - Ah, interesting. Yes, we are testing the use of the least-intrusive version of the Facebook Pixel to see if it helps us understand if people are in fact visiting our site based on ads we run on Facebook. We don't want to waste ISOC funds on ads if they are not effective. On the other hand, if the ads work well and get more people involved in ISOC work, they may be a great use of funds. We are definitely aware of the privacy issues, though, and are using the 'standard events' version that collects only the bare minimum of information (and theoretically no PII). It's a temporary trial to see whether it helps us or not. It's disappointing to see that this block of code doesn't meet accessibility standards. Thanks for digging into that!

[danyork]

@birkenbihl - I just wanted to let you know that we haven't forgotten this issue. We've been working through a number of performance-related issues with the site which we are hoping to finish up in the next week or so. Our developers will be moving to address some of these outstanding accessibility issues.

[danyork]

@birkenbihl - Another update on this issue. We've removed the Facebook Pixel and so a number of our validation issues went away. Our developers also spent some time fixing various other issues. We have a few more to fix, but we're getting close to where all that may be left will be some warnings:

https://validator.w3.org/nu/?doc=https%3A%2F%2Fwww.internetsociety.org%2F

Still working on it...