Non-normative, passing mention of APCA

[svgeesus]

On https://www.w3.org/WAI/GL/WCAG3/2022/how-tos/visual-contrast-of-text/, on the "Get Started" tab, it currently says:

The complex, interrelated nature of font and graphic elements to provide visual readability is aided by the use of the APCA algorithm (Advanced Perceptual Contrast Algorithm) which allows authors to input color pairs to find a predicted contrast value. That contrast value then defines the minimum useable font weight and size.

The phrase "is aided" gives the impression of an optional item; perhaps one choice among many for predicting a contrast value.

If there are other options available, which WCAG considers equivalent, please list them.

If APCA is the sole option, please strengthen the wording to indicate that this is the correct and sole way to calculate a predicted contrast value.

[bruce-usab]

@Myndex – do you think the How tab could/should be the algorithm used by APCA? If not, please suggest prose for the How tab that is more generic, but still strong enough to push the reader beyond 2.x metrics.

[Myndex]

Hi Chris @svgeesus

The phrase "is aided" gives the impression of an optional item....If there are other options available, which WCAG considers equivalent, please list them.

Hi Chris, I see how that can be misinterpreted, and I don't recall the genesis of that statement, but I agree it should be strengthened. No, there are no other suitable methods for the stated purpose of readability.

In 2019, I evaluated dozens of contrast methods, published, casually discussed, and novel. None were well tuned for readability on self illuminated monitors, nor had polarity sensitivity in keeping with perception. This led to SAPC and looking at ways to simplify/combine/tune various other CAM models into the narrower purpose of contrast of stimuli of high spatial frequency (particularly text). APCA is the result of this exploration.

I'll work on a PR for this.

[Myndex]

do you think the How tab could/should be the algorithm used by APCA? If not, please suggest prose for the How tab that is more generic, but still strong enough to push the reader beyond 2.x metrics.

Hi Bruce @bruce-usab

I added an svg and MathML of the base algorithm to the PR w3c/silver#630 last week, though I added it to the methods section. That still is awaiting review (did I not ping you? I thought I clicked you for review...)

Do you think it's better in the How To? Should I amend the PR? The PR for the method section dives deeper into the specifics, and also contains specific techniques, such as the correct method for using an eye-dropper tool.

I'm standing by....Thank you!

[bruce-usab]

I thought I clicked you for review...

Okay, I tried. I don't think that really advances the publication much at all. It just results in:

You’re not to merge this pull request.

Ping to @michael-n-cooper to ask for him to hit me with a clue bat.

Do you think it's better in the How To?

I forgot about your PR @Myndex – so my suggestion for the How To is applicable to this Issue only, and the version of Get Started tab in OP.

I am inclined to think an algorithm is attached to an Outcome and then APCA is a Method to achieve the Outcome.

[Myndex]

I forgot about your PR – so my suggestion for the How To is applicable to this Issue only, and the version of Get Started tab in OP.

I am inclined to think an algorithm is attached to an Outcome and then APCA is a Method to achieve the Outcome.

Thank you @bruce-usab — what I was thinking was to update the "get started tab" with a link to the resources tab, and put pseudocode or basic JS in the resources tab.

I'm assuming this should be added to the existing pull request...

[svgeesus]

Thanks @Myndex that wording looks much better.

[Myndex]

Thanks @Myndex that wording looks much better.

No problem @svgeesus thank you for bringing it to my attention.

[svgeesus]

@bruce-usab so @Myndex has answered all my questions on this pull request to my satisfaction and thus I would like to see it merged.

Do you have any outstanding comments? I saw you wrote these edits all look fine to me.

@michael-n-cooper what is the policy here, are edits merged solely by document editors? Or can edits reviewed by subject specialists (which is the case here) also be merged?

[bruce-usab]

I do not have any outstanding comments. I appreciate the ping!

I also would like to see the edits merged.

[Myndex]

Hi @cwadamsoforacle This thread should migrate over with the other that have.