Site Usability nitpick -- Hyperlinks should be underlined

[DaveC426913]

It really seems to be grinding your gears though!

We all have our battles. Usability is mine. PF is an oasis of quality content. I want to keep it that way.

The question is more: why is it grinding anybody else's gears?

This thread didn't have to be 30 posts long. Post #2 could have been "Huh. You're totally right. Function before form. We should enact W3C internet standards that facilitate accessibility for all, not just a few." - and garnished with two dozen Likes.

Watch your blood pressure and remember that the internet is basically anarchy.

No, the internet is mostly anarchy. That doesn't mean we all just give up.

 

[Dale]

Post #2 could have been "Huh. You're totally right. Function before form. We should enact W3C internet standards that facilitate accessibility for all, not just a few."

But do we have any actual evidence that underlining links promotes accessibility? I support the goal, but I am not convinced that this assists the goal.

I personally have a fairly strong astigmatism, so I find excessive underlining makes text more difficult to read, not less. I would not assume that more underlining would actually help others who are more severely visually impaired. It doesn't help me with my minor impairment.

 

[sbrothy]

We all have our battles. Usability is mine. PF is an oasis of quality content. I want to keep it that way.

The question is more: why is it grinding anybody else's gears?

This thread didn't have to be 30 posts long. Post #2 could have been "Huh. You're totally right. Function before form. We should enact W3C internet standards that facilitate accessibility for all, not just a few." - and garnished with two dozen Likes.


No, the internet is mostly anarchy. That doesn't mean we all just give up.

You’re right. But, as is often the case, I have a twinkle in my eye when posting (A habit I should really combat more vigorously.).

This seemed to me to be a battle which couldn’t be won - or at least one that would end in a Pyrrhic victory - however that would look?

But all you want seems to be to change how physicsforums shows links. Doesn’t sound impossible on the face of it, but how deep the behavior sticks I have no idea.

It seems it works for @TensorCalculus in “dark mode” which is what I meant with the look being customizable in the user end. Of course that probably entails other problems and I myself like to switch back and forth precisely to avoid straining my eyes.

But enough nitpicking from me. Just wanted to let you know that I understand that this is a real problem for you and not to be ridiculed.

 

[DaveC426913]

But all you want seems to be to change how physicsforums shows links. Doesn’t sound impossible on the face of it, but how deep the behavior sticks I have no idea.

By default, all links will show with an underline. When they don't it's because a designer has added a CSS style:

a { text-decoration: none; }

 

[DaveC426913]

But do we have any actual evidence that underlining links promotes accessibility? I support the goal, but I am not convinced that this assists the goal.

Do you mean has the World Wide Web consortium - the body that sets the standards for the trillion dollar critical global infrastructure technology that is the world wide web - actually done any government-mandated and funded research and acquired data on billions of user experiences, to define their standards - as opposed to merely guessing?

Yes.

 

[DaveC426913]

Official standard: WCAG (Web Content Accessibilty Guidelines) published by the World Wide Web Consortium (W3C), widely recognized internationally, including adoption as an ISO standard (ISO/IEC 40500:2012):

• Link-specific guidelines: WCAG 2.0 introduced criteria such as 2.4.4 Link Purpose (In Context) and 2.4.9 Link Purpose (Link Only) which require that the purpose of each link is clear to users, particularly through accessible text or context.

Yale Usability & Web Accessibility's guidance on links covers important UX considerations like:

• Always use a non-color indicator (typically an underline) to distinguish links from regular text; color alone is not enough for accessibility.

(Full disclosure: I did use ChatGPT to do a search for the relevant references, because it's fast, but I curated the results for brevity and clarity).

 

[sbrothy]

By default, all links will show with an underline. When they don't it's because a designer has added a CSS style:

a { text-decoration: none; }

OMG. Don’t get me started on CSS. Talk about battles you can’t win!

 

[DaveC426913]

OMG. Don’t get me started on CSS. Talk about battles you can’t win!

That's literally all it took to implement - and literally all it takes to remediate it.

 

[Dale]

Do you mean has the World Wide Web consortium - the body that sets the standards for the trillion dollar critical global infrastructure technology that is the world wide web - actually done any government-mandated and funded research and acquired data on billions of user experiences, to define their standards - as opposed to merely guessing?

Yes.

What goes into standards is based on the judgement and agreement of the members of the standards committee. Often they agree because there is published research on the topic, but other times they agree to choose a standard even in the absence of actual research.

So simply being a published standard for a big and regulated industry does not imply that a specific standard is based on scientific evidence. Scientific evidence for a standard is found in published scientific papers.

Official standard: WCAG (Web Content Accessibilty Guidelines) published by the World Wide Web Consortium (W3C), widely recognized internationally, including adoption as an ISO standard (ISO/IEC 40500:2012):

• Link-specific guidelines: WCAG 2.0 introduced criteria such as 2.4.4 Link Purpose (In Context) and 2.4.9 Link Purpose (Link Only) which require that the purpose of each link is clear to users, particularly through accessible text or context.

Yale Usability & Web Accessibility's guidance on links covers important UX considerations like:

• Always use a non-color indicator (typically an underline) to distinguish links from regular text; color alone is not enough for accessibility.

Interesting. So it seems that neither Google nor Wikipedia (nor PF) follow the WCAG. Or possibly the "mouse-over" underlining behavior is considered to meet the WCAG standard.

Do we know which is the case? If Google et al. do not follow the WCAG then is there a competing standard that they do follow?

 

[DaveC426913]

So simply being a published standard for a big and regulated industry does not imply that a specific standard is based on scientific evidence. Scientific evidence for a standard is found in published scientific papers.

Which is why I said:
"[they have] ...actually done ... government-mandated and funded research and acquired data on billions of user experiences..."


Do we need to produce the published papers on world wide web usability and wait for them to be vetted by members here (who are not field experts) before we make a one-line CSS change that helps PF be more accessibility-compliant?

I'm not being facetious. That seems a pretty high bar.


Put another way:

"...do we have any actual evidence that underlining links promotes accessibility? I support the goal, but I am not convinced that this assists the goal..."

How much evidence would convince you that colour-blind and vision-deficient users can't rely on colour alone to see links embedded in text?

As Yale says: "color alone is not enough for accessibility."

 

[DaveC426913]

"Blue hypertext is a good design decision: no perceptual disadvantage in reading and successful highlighting of relevant information" [underlining is addressed as well]
https://pmc.ncbi.nlm.nih.gov/articles/PMC5036113

Underline on hover/focus is still a failure
Having an underline but only on hover is still a failure of criterion 1.4.1. If the non-color cue only happens when the mouse hovers over the link or when the link receives focus, it is still a failure.” 3
[ https : //medium.com/vassar-design/links-with-underlines-as-a-best-practice-fe1ba0d6ba71 ]
(I don't understand why PF insists on rendering this link as if it is a [ media ] object. It's just a web page.)

"Textual links should be colored and underlined to achieve the best perceived affordance of clickability, though there are a few exceptions to these guidelines."
https://www.logarithmic.net/pfh-files/links.html

There are still "merely" recommendations from authorities. Do I need to provide the original papers?

 

[Dale]

How much evidence would convince you that colour-blind and vision-deficient users can't rely on colour alone to see links embedded in text?

Most color-blind users are not achromatic and would be effectively served by different color schemes. For example, the various viridis color maps that I use in R for this purpose.

Being a mildly vision-deficient (astigmatism) user myself, I personally don't believe that underlining helps. It certainly makes things more difficult to read for me personally. So, the evidence that I would require would actually be a peer-reviewed study or that included a full gamut of vision-deficiencies, or similar.

I don't think that it is as obviously beneficial as you assert. It would be worse for me personally, so I would want clear evidence that it is more beneficial over the whole population.

I'm not being facetious. That seems a pretty high bar.

Nor am I. For me it is not beneficial and actually worse. So yes, I do ask for a higher bar than it being in some standard that appears to not be followed by the most major sites.

 

[sbrothy]

Most people I know think color blindness means that those that suffer from it see in black & white only.

 

[Dale]

"Blue hypertext is a good design decision: no perceptual disadvantage in reading and successful highlighting of relevant information" [underlining is addressed as well]
https://pmc.ncbi.nlm.nih.gov/articles/PMC5036113

It seems from this research that there is very little benefit to underlining in addition to blue color.

The effects seem small and the analysis is not corrected for multiple comparisons. It looks like the only place it has an effect is in "go past" time and total viewing time. But with the go-past time underlining and blue made it worse than just blue.

And none of the participants were vision-deficient in any way that was not corrected in the study.

I have not read this yet, but it may be more useful, particularly if it actually cites the research:
https://www.nngroup.com/reports/usability-guidelines-accessible-web-design/
This was a further link from the second link you posted above.

 

[AlexB23]

Yes, by all means throw a wrench in the gears.

Haha, I did that to demonstrate why it is good to use underlines for links, as blue text could be just colorized.

 

[DaveC426913]

Most color-blind users are not achromatic and would be effectively served by different color schemes. For example, the various viridis color maps that I use in R for this purpose.

Being a mildly vision-deficient (astigmatism) user myself, I personally don't believe that underlining helps. It certainly makes things more difficult to read for me personally. So, the evidence that I would require would actually be a peer-reviewed study or that included a full gamut of vision-deficiencies, or similar.

I don't think that it is as obviously beneficial as you assert. It would be worse for me personally, so I would want clear evidence that it is more beneficial over the whole population.


Nor am I. For me it is not beneficial and actually worse. So yes, I do ask for a higher bar than it being in some standard that appears to not be followed by the most major sites.

OK. I give up.

It seems as if to enact even the slightest change and accommodate a wider audience, apparently it will require PF to ignore the entire accessibility industry, its experts, their standards and established best practices, and instead launch its own studies to analyze the raw data to become its own expert, and come to its own conclusions, which - so far - consists of, like, three users who have said "I dunno ... I looked at this page for twelve seconds and for me it's not benefical ..."

OK, now I'm being facetious.


Anyway, I have nothing further to contribute. You guys know best.

 

[AlexB23]

OK. I give up.

It seems as if to enact even the slightest change and accommodate a wider audience, apparently it will require PF to ignore the entire accessibility industry, its experts, their standards and established best practices, and instead launch its own studies to analyze the raw data to become its own expert, and come to its own conclusions, which - so far - consists of, like, three users who have said "I dunno ... I looked at this page for twelve seconds and for me it's not benefical ..."

OK, now I'm being facetious.


Anyway, I have nothing further to contribute. You guys know best.

I side with you, Dave.

See, this is a link about HTML, and this is blue text.

 

[vela]

Add to that google search results.

Kagi search results, on the other hand, do have underlined links. So do the sites for the New York Times, the LA Times, the Washington Post, and the Wall Street Journal. That doesn't seem too surprising because newspapers probably prioritize making their sites accessible. I just checked a number of sites I visit regularly and noticed the majority of them have underlined links.

Told you lack of underline has become a de facto standard.

I don't think following allegedly de facto standards on the web should be the goal.

The visual indicator for a link shouldn't be just color for the reasons mentioned earlier. It doesn't necessarily have to be underlining, but there should be some visual indication other than color, which doesn't require a mouseover to discover. That is, of course, if PF cares about accessibility.

 

[Greg Bernhardt]

I think some years ago, heavy link websites started to move away from the underline because it added so much visual noise to the page.

 

[DaveC426913]

I think some years ago, heavy link websites started to move away from the underline because it added so much visual noise to the page.

Yes. Pandering to the default ideal user at the expense of accessibility. Not a good ethos for PF.


We are an aging population. For the first time in census records, seniors are now the largest demographic.

 

[Dale]

Pandering to the default ideal user at the expense of accessibility. Not a good ethos for PF.

I am still not convinced of this. And although it clearly exasperates you, your exasperation is not a credible substitute for actual evidence. When someone is unconvinced about relativity, I can point them to actual scientific papers with real experiments and repeatable and confirmed experimental data. I don't have to rely on authority nor name-calling.

Also, wouldn't the best solution be at the browser level rather than the site level? Why should either user's experience come at the expense of the other's?

 

[DaveC426913]

I am still not convinced of this. And although it clearly exasperates you, your exasperation is not a credible substitute for actual evidence. When someone is unconvinced about relativity, I can point them to actual scientific papers with real experiments and repeatable and confirmed experimental data. I don't have to rely on authority nor name-calling.

Well, yes. I didn't expect site usability decisions to have to undergo the same rigor as the science it proffers.

Presumably, one would avail themselves of the good work of the experts, we're not in the business of reinventing the wheel are we?

Minor changes are made here all the time. Is every one of the them preceded by throwing away industry guidelines and diving directly into the raw data?

We didn't delve into the data to examine the efficacy of gender labeling in user avatars.

Also, wouldn't the best solution be at the browser level rather than the site level? Why should either user's experience come at the expense of the other's?

But again, you are second-guessing, not only the industry experts, but the government's usability guidelines, as defined by the ADA in faovur of your personal opinion.

People with less than perfect vision should no longer be marginalized, to be left to solve their own problems, while the rest of the population enjoy being pandered to by policy. Seniors are now the largest demographic.


USA Accessibility mandates for web content:
In the U.S., web accessibility is mandated by the Americans with Disabilities Act (ADA) (for state/local government and businesses open to the public) and Section 508 of the Rehabilitation Act (for federal agencies and vendors selling to the government). Both laws use the Web Content Accessibility Guidelines (WCAG) as the technical standard, requiring content to meet WCAG 2.1 Level A and AA conformance. State and local governments have compliance deadlines in 2026 and 2027 depending on their size, while federal agencies and their vendors must be compliant now.

https://www.ada.gov/resources/web-guidance/#why-website-accessibility-matters
"Inaccessible web content means that people with disabilities are denied equal access to information. An inaccessible website can exclude people just as much as steps at an entrance to a physical location. Ensuring web accessibility for people with disabilities is a priority for the Department of Justice. In recent years, a multitude of services have moved online and people rely on websites like never before for all aspects of daily living. For example, accessing voting information, finding up-to-date health and safety resources, and looking up mass transit schedules and fare information increasingly depend on having access to websites."
...
"Use of color alone to give information. People who are color-blind may not have access to information when that information is conveyed using only color cues because they cannot distinguish certain colors from others. Also, screen readers do not tell the user the color of text on a screen, so a person who is blind would not be able to know that color is meant to convey certain information (for example, using red text alone to show which fields are required on a form)."

 

[Dale]

People with less than perfect vision should no longer be marginalized

I am one such person and your suggested "improvement" marginalizes me.

But again, you are second-guessing, not only the industry experts, but the government's usability guidelines, as defined by the ADA in faovur of your personal opinion.

And you are misrepresenting them in favor of your personal opinion. Reading the actual statement in the WCAG, underlines are NOT required. According to the standard the requirement is only that a link should be identified by some indication other than just hue. So bold or brightness would also suffice, according to the standard. Personally, I would find bold the least problematic for astigmatism.

 

[weirdoguy]

People with less than perfect vision should no longer be marginalized

@Dale also has less than perfect vision, which he stated multiple times, and you seem to deliberately ignore it.

EDIT: Oh sorry, I didn't see @Dale that you yourself wrote about it.

 

[DaveC426913]

@Dale also has less than perfect vision, which he stated multiple times, and you seem to deliberately ignore it.

I'm not ignoring it. I don't think we should be making site-wide policy based on one user's experience, when there is a mountain of evidence purported by experts in the field to be had.

Yes, not every user will be able to have an ideal experience, but does that mean we ignore the needs of the many in favour of the preferences of the one?

 

[DaveC426913]

And you are misrepresenting them in favor of your personal opinion.

It's not my personal opinion! How many industry experts do you require?

Reading the actual statement in the WCAG, underlines are NOT required. According to the standard the requirement is only that a link should be identified by some indication other than just hue. So bold or brightness would also suffice, according to the standard. Personally, I would find bold the least problematic for astigmatism.

OK, let's do that, then*.

* unfortunately, in blocks of text, users do not tend to associate bolding with hyperlinks. The indicator that users associate with hyperlinks is underlining.

But again, Dale's personal preference versus the industry experts?

 

[vela]

I think some years ago, heavy link websites started to move away from the underline because it added so much visual noise to the page.

Is PF a link-heavy site?

In my brief look at a number of sites the other day, I noticed most sites don't underline every link, but they do for links in a block of text where the visual indicators help them stand out from the surrounding text.

@Dale, does the underlining on a site like Ars Technica cause you problems?

https://arstechnica.com/science/202...have-been-as-uniformly-massive-as-we-thought/

 

[Dale]

It's not my personal opinion! How many industry experts do you require?

Yes, it is. The experts writing the standards you are citing are not actually saying what you claim they are saying. The standard does not require underlining. It requires some non-color difference. This could be underlining, but it could also be bold, luminosity, or anything else.

Here is what the industry experts that you are misrepresenting actually say in their standard:

https://www.w3.org/TR/WCAG21/#use-of-color

Success Criterion 1.4.1​

Color is not used as the only visual means of conveying information, indicating an action, prompting a response, or distinguishing a visual element.

https://www.w3.org/WAI/WCAG21/Understanding/use-of-color

Understanding SC 1.4.1:
...
Note

If content is conveyed through the use of colors that differ not only in their hue, but that also have a significant difference in lightness, then this counts as an additional visual distinction, as long as the difference in relative luminance between the colors leads to a contrast ratio of 3:1 or greater. For example, a light green and a dark red differ both by color (hue) and by lightness, so they would pass if the contrast ratio is at least 3:1. Similarly, if content is distinguished by inverting an element's foreground and background colors, this would pass (again, assuming that the foreground and background colors have a sufficient contrast).

https://www.w3.org/WAI/WCAG21/Techniques/general/G183

Technique G183:
Using a contrast ratio of 3:1 with surrounding text and providing additional visual cues on hover for links or controls where color alone is used to identify them
...
To meet Success Criterion 1.4.1: Use of Color a relative luminance (lightness) difference of 3:1 or greater with the text around can be used. This technique goes beyond the success criterion and asks for visual highlights when the user hovers over each link, such as an underline, a change in font style such as bold or italics, or an increase in font size. Such a hover effect provides confirmation to pointer users that the text is a link, similar to how the link alters its appearance for keyboard users when it receives focus, in order to meet 2.4.7 Focus Visible.
...
Example 2
The hypertext links in a document are medium-light blue (#3366CC) and the regular text is black (#000000). Beyond the difference in color, the links have no other styles differences compared with the regular text. Because the blue text is light enough, it has a contrast of 3.9:1 with the surrounding text and can be identified as being different from the surrounding text by people with all types of color vision deficiency, including those individuals who cannot see color at all. In addition to the contrast difference, the links have :focus and :hover styles that reintroduce the underline when the links receive keyboard focus or when a mouse pointer hovers over them. Hover or focus style changes alone are not sufficient to meet the criterion.

https://www.w3.org/WAI/WCAG21/working-examples/link-contrast/

The following 26 web-safe colors pass at 3:1 vs black and 5:1 vs. white

From what I can tell Google, Wikipedia, and PF do in fact meet the actual standards. Just not your individual preference, which you are pushing as though it were something more. They appear to all be using the technique G183 explicitly given in the industry standard as a specifically allowed means of being accessible.

Underlines are not required for accessibility.

 

[Dale]

Is PF a link-heavy site?

In my brief look at a number of sites the other day, I noticed most sites don't underline every link, but they do for links in a block of text where the visual indicators help them stand out from the surrounding text.

@Dale, does the underlining on a site like Ars Technica cause you problems?

https://arstechnica.com/science/202...have-been-as-uniformly-massive-as-we-thought/

Their underlining is better than most because the underline is separated from the baseline of the text by a greater distance than most. So it makes less "interference" with the text. Even the "g" does not intersect the underline. I don't find it difficult to read like I often do with closer underlines.

 

[robphy]

I have just come from Wikipedia. They use the same style as PF: they only underline links on rollover.

Shame on them.

Here are some sites you might like a little better:

• https://physics.stackexchange.com/ (you have to find a post with hyperlinks in the text or comments)
• https://www.reddit.com/r/Physics/ (you have to find a post with hyperlinks in the text or comments)
• https://gregbernhardt.com/