Per @Volker_E I also want to suggest that we change the blue link color in Vector to Accent50 (#36c) so it matches that in Minerva.

@Huji I made this task more general and created two sub tasks for the proposed changes

#3366cc is a poorly chosen colour for links. While you’re upping contrast level to almost 3:1 with normal text, like WCAG 2.0 tells us, you’re decreasing it almost in two times for the background (while not against AA level, making it against AAA level). What that really tells us is not that we need to have less visible link colour, but we should’ve not used #222 as main text colour because of purely stylistic choices (and fix contrast of the links in regards to pure black text instead of dark grey).

You can use this for debugging it:
https://contrast-checker.glitch.me

@stjn can you clarify what you mean here:

While you’re upping contrast level to almost 3:1 with normal text, like WCAG 2.0 tells us, you’re decreasing it almost in two times for the background (while not against AA level, making it against AAA level)

I'm not sure I understand

#3366cc is almost two times lighter against background compared to #0645ad and doesn’t pass WCAG 2.0 AAA level for text.

I see. To clarify:
#0645AD has a contrast ratio of 8.53:1 against #FFFFFF (passes WCAG AA, AAA)
#3366cc has a contrast ratio of 5.37:1 against #FFFFFF (passes WCAG AA, fails AAA)

I think it makes sense to balance these measurements with the measurements in the task description, regarding how distinguishable these colors are against black text. When considering only how they perform against a white background, I see your point about #0645AD. However when considering both measurements it seems like #3366cc is a better choice.

Yeah, I get that, that’s why I’m saying that reverting back to black and seeking accessible colours next to pure black is better than having pale link colours that still barely come around WCAG 2.0 AA level requirement (3:1, see https://webaim.org/blog/wcag-2-0-and-link-colors/) because of using dark grey.

I was asked whether T219707, which I created, should be merged here. My reply was that the two tickets are "Related, but not the same".

I am unclear as to why they have been merged nonetheless.

Solving T213778 will not solve T219707.

@Pigsonthewing can you clarify what you mean here?

In T213778#5096826, @Pigsonthewing wrote:

Solving T213778 will not solve T219707.

In T213778#5097479, @alexhollender wrote:

@Pigsonthewing can you clarify what you mean here?

In T213778#5096826, @Pigsonthewing wrote:

Solving T213778 will not solve T219707.

T213778 proposes "updating our colors used for links and visited links to ones that are more distinguishable (i.e. further apart) from the color we use for black text". It says nothing about making colours used for links and visited links more distinguishable from each other, which is the point of T219707.

@Pigsonthewing thank you for the clarification. Not sure why that task got closed. I think you could either edit the description of this task by adding a section about the contrast between links and visited links, or re-open that task.

I've updated the description to cover T219707.

@Volker_E @Pigsonthewing the results in the second table in the Description section above are confusing when you compare them with the before/after demonstration image in the Proposal section. Visually it appears that the proposed colors are more distinguishable from each other, but the Delta E and contrast values are lower. What are your thoughts?