# Accessibility: Semantic labeling

**URL:** https://meta.discourse.org/t/accessibility-semantic-labeling/77889
**Category:** Feature
**Tags:** accessibility
**Created:** [January 12, 2018, 9:24pm UTC](https://meta.discourse.org/t/accessibility-semantic-labeling/77889 "2018-01-12T21:24:31Z")
**Posts on this page:** 4
**Page:** 1

<div class="post-metadata">

### Author: ![kevinrobinson](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/kevinrobinson/32/120056_2.png) [@kevinrobinson](https://meta.discourse.org/u/kevinrobinson)
#### Post date: [January 12, 2018, 9:24pm UTC](https://meta.discourse.org/t/accessibility-semantic-labeling/77889/1 "2018-01-12T21:24:31Z")

</div>

This is split out from the master accessibility thread: [Accessibility audit and shepherd for making improvements](https://meta.discourse.org/t/accessibility-audit-and-shepherd-for-making-improvements/66620/16).

Accessibility question: Are form fields and links properly labeled?

> Not always.  
> Missing label noted in screen shot of checker tool below.  
> See Techniques for proper labeling here [H91: Using HTML form controls and links | Techniques for WCAG 2.0](https://www.w3.org/TR/WCAG20-TECHS/H91)

I tried gathering more data using [Free Web Accessibility Checker – WCAG Audit | achecker.ca](https://achecker.ca/checker/index.php), trying to find tools to automate some of these checks. I checked several pages as representative of the user interfaces that make up the bulk of user interactions (home, category, topic pages). There was one class of “known problems” (using `<b>` tags). There’s some previous conversation [here](https://meta.discourse.org/t/accessibility-audit-and-shepherd-for-making-improvements/66620/3) and [here](https://meta.discourse.org/t/accessibility-audit-and-shepherd-for-making-improvements/66620/4) about whether the `<b>` / `<strong>` distinction is meaningful (from the perspective of the accessibility audit the answer is a definite “yes,” even if this might feel pedantic or like it’s adding cost of bytes over the wire for the longer tag).

I didn’t yet investigate the “potential problems,” but could include that in the next phase of improvements.

---

<div class="post-metadata">

### Author: ![kevinrobinson](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/kevinrobinson/32/120056_2.png) [@kevinrobinson](https://meta.discourse.org/u/kevinrobinson)
#### Post date: [January 12, 2018, 9:25pm UTC](https://meta.discourse.org/t/accessibility-semantic-labeling/77889/2 "2018-01-12T21:25:43Z")

</div>

So the next steps here would be: what would it take to change `<b>` tags to `<strong>` on the accessibility checking tool’s recommendation? Or if folks have questions or concerns about trusting that tool’s recommendations, we could look for other guidance here (ideally that comes with automated tooling).

---

<div class="post-metadata">

### Author: ![codinghorror](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/codinghorror/32/110067_2.png) [@codinghorror](https://meta.discourse.org/u/codinghorror)
#### Post date: [January 12, 2018, 9:48pm UTC](https://meta.discourse.org/t/accessibility-semantic-labeling/77889/3 "2018-01-12T21:48:47Z")

</div>

One problem here is that the checker needs to be loading the full page in the browser, as rendered by JavaScript – if it is doing a `curl` command it is not getting the same page the actual browser is getting, as the whole page is built using JavaScript commands.

---

<div class="post-metadata">

### Author: ![kevinrobinson](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/kevinrobinson/32/120056_2.png) [@kevinrobinson](https://meta.discourse.org/u/kevinrobinson)
#### Post date: [January 12, 2018, 9:51pm UTC](https://meta.discourse.org/t/accessibility-semantic-labeling/77889/4 "2018-01-12T21:51:38Z")

</div>

@codinghorror Ah cool, that seems like a good thing to watch out for in general. 👍 I think there are bunch of better tools to do this as well, and it’d be good to find out more there too!

For the particular issue here, I don’t see as much how that’s related though, that tool seems to have done the trick in spotting this bit. 🙂
