- From: Gregg Vanderheiden <gregg@vanderheiden.us>
- Date: Thu, 3 Aug 2023 15:14:36 -0700
- To: Andrew Somers <Andy@GeneralTitles.com>
- Cc: Jonathan Avila <jon.avila@levelaccess.com>, "w3c-waI-gl@w3. org" <w3c-wai-gl@w3.org>
- Message-Id: <4466A46F-6211-4926-B99C-9B26D8A9393F@vanderheiden.us>
Hi Andrew Can you provide the citations for the studies you are referring to? I would be good to read them as we work on WCAG 3. The colorspaces of computers have all changed since WCAG 2,0 days so we need to rework everything anyway. But It would be good to see the studies you cite. Also — were they based on the WCAG provisions? They only state the contrast ratio. Or were they based on tools? and did the sites have less contrast for people with color blindness and low vision — or just Lower contrast for normal vision? Ah... and were they based on how to calculate contrast? -One thing I found from this is that there are different camps on how to measure all this - and Aries Arditi - a vision scientist with the lighthouse for the blind who did research with low vision users finally said he was not going to get involved with any more of these discussions… since the seem to border on 'religious' discussions at times. - but the Oberdorfer Award in Low Vision, was presented by the Association for Research in Vision andOphthalmology (ARVO) and the ARVO Foundation for Eye Research in 2013 for the work done on WCAG and contrast. I think going forward we would do better to not throw stones and simply try to figure out what is the best methods with modern colorspaces — look for best evidence - and then make our new standards in W3 Two things I think are for sure The display technologies have changes — and so we DO need to update the contrast measures We need to keep in mind that we are talking about the best measures for people with low vision, people with different color vision differences — and people with BOTH. That will make a lot of the standard research on contrast a bit less useful. And the type of research we need much rarer. I ALSO suspect that We should be looking at studies that involve 'reading text' vs studies about letter identification or anything image base etc Color contrast and Contrast are different. Two colors can contrast with each other (e.g. a red and a green) but be identical to someone. I think we should drop "color contrast" and just talk about contrast Best Thanks All this will help us as we move forward gregg ------------------------------ Gregg Vanderheiden gregg@vanderheiden.us > On Aug 3, 2023, at 1:32 PM, Andrew Somers <Andy@GeneralTitles.com> wrote: > > Hi Jon, > > I think I'm being misunderstood here. There are functional, tested, peer reviewed, useful contrast methods available today. WCAG 2.x contrast is not one of them. It is not helpful, and a survey of sites from before WCAG 2 showed they used a demonstrably better contrast than after, with the adoption of HTML5, where color was abstracted to CSS, these sites such as Wired which formerly used good contrast, were/are REDUCING contrast. > > This is not good, and an unintended consequence. > > My position on this was first stated in April 2019; subsequent research and investigation has shown that: > > 1) There currently exists substantial misunderstanding regarding contrast for design. Many of these misunderstandings trace back to the WCAG 2.x documentation. > > 2) Surveys of major sites has shown a marked decrease in contrast in the years after WCAG 2.0 became the recommendation, this appears to also be related to the switch to HTML 5. It would be bad for accessibility if that were to happen to apps as well. > > 3) There is no evidentiary support that WCAG 2.x contrast does anything unique nor special to aide accessibility. > > 4) Promoting WCAG 2.x contrast to the point it is being added to legislation has the chilling effect of blocking actual accessibility. Certain aggressive individuals loudly claiming that “WCAG 2 is law” as if that was somehow more important than actual accessibility, and their related actions, have resulted in obstruction of actually accessible contrast methods and design systems. > > > > THEREFORE > > If this organization does not want to incorporate actually accessible methods, that’s one thing. But it is quite another for this organization to continue to promote methods are that > are widely known and demonstrated as not usefully functional, that > do not reference the relevant peer reviewed scientific consensus, that > do not do what is claimed, and > can create conditions that are worse for readability, and > as a result blocks the advancement or use of actually accessible methods. > > This is a visible problem that impacts the rest of the SCs. > > Rather than force 6.1.5.4 and 6.1.5.8 on developers (which is what would eventually be the case), it would be better to state: > > 6.1.5.4: shall use an accessible method for determining contrast for the context of use. > > Or similar language. This has the benefit of permitting open development of methods and guidelines relating to visual presentation, and allows the AGWG to focus on issues relating to ARIA, semantic markup, and so forth. > > > Getting it wrong is bad, but getting it wrong and then obstructing advances in the field is unacceptable. > > > Thank you for reading, > > Andy > > > Andrew Somers > Senior Color Science Researcher > PerceptEx Perception Research Project <https://www.myndex.com/perceptex/> > P.O.Box 1867 🜛 Hollywood, CA 90078 > Vm/Tx: 213-448-4746 🜛 Andy@Myndex.com <mailto:Andy@Myndex.com> > <https://www.myndex.com/perceptex/> > >> On Aug 3, 2023, at 9:38 AM, Jon Avila <jon.avila@levelaccess.com <mailto:jon.avila@levelaccess.com>> wrote: >> >> Andrew, what I’m understanding you are saying is that it would be better to have no contrast recommendations for the next several years for Window software, documents, and mobile apps. because the correct method is not ideal. So, no contrast minimum at all is preferrable and the impact of not having a contrast requirement would be a reasonable trade off to limit use by people with low vision in order to address the concerns of designers with the current approach. >> >> I can’t express in words what this stance if adopted by the organization would imply to a group of marginalized people. >> >> Jonathan >> >> From: Andrew Somers <andy@generaltitles.com <mailto:andy@generaltitles.com>> >> Sent: Thursday, August 3, 2023 12:28 PM >> To: Abou-Zahra, Shadi <sabouzah@amazon.at <mailto:sabouzah@amazon.at>> >> Cc: WCAG list (w3c-wai-gl@w3.org <mailto:w3c-wai-gl@w3.org>) <w3c-wai-gl@w3.org <mailto:w3c-wai-gl@w3.org>>; Silver TF <public-silver@w3.org <mailto:public-silver@w3.org>>; Chuck Adams <charles.adams@oracle.com <mailto:charles.adams@oracle.com>> >> Subject: Re: CFC - Publish WCAG2ICT First Public Working Draft >> >> Hi Shadi, >> >> To add and be absolutely clear, I can set the licensing as needed for either APCA, DPS contrast, or SACAM. >> >> A clear example supporting my formal objection to the continued promotion of WCAG 2.x contrast: >> >> This basic example is in part why the design community has no interest in WCAG 2.x, and also why the accessibility community should be up in arms as well. According to WCAG 2.x contrast math, both the left (dark text) and the right (white text) examples against orange #de7a01 are an identical 3.04 to 1 contrast. >> >> Do you find them to be the same contrast and same readability? Look especially at the colorblind examples, particularly protanopia. >> >> <image001.png> >> >> The claims that WCAG 2.x is somehow better for color vision deficiency are patently false. Following WCAG 2.x can be harmful to readability, regardless of impairments. >> >> WCAG 2.x contrast was never subject to the substantial testing we have conducted for APCA, has no peer review, was objected to by IBM and others back in 2007, and the fact that it continues today is nothing but a black mark on all of the other SCs of WCAG 2.x. That it is now being elevated to law in the EU in untenable and an atrocious overreach. The unfortunate inclusion of SC 1.4.3 in WCAG 2.x has had the unintended consequence of causing a lowering of contrast of major sites (based on a survey of major sites before WCAG 2.0, and later contrast reductions). >> >> The continued promotion of WCAG 2.x contrast as if it was somehow valid must stop, now. It is time to end this charade and focus on being actually accessible. >> >> >> >> Thank you for reading. >> >> Andy >> >> >> Andrew Somers >> Senior Color Science Researcher >> PerceptEx Perception Research Project <https://www.myndex.com/perceptex/> >> P.O.Box 1867 🜛 Hollywood, CA 90078 >> Vm/Tx: 213-448-4746 🜛 Andy@Myndex.com <mailto:Andy@Myndex.com> >> <https://www.myndex.com/perceptex/> >> >> >> On Aug 3, 2023, at 2:56 AM, Andrew Somers <Andy@GeneralTitles.com <mailto:Andy@GeneralTitles.com>> wrote: >> >> Hi Shadi, >> >> Well, on that point, WCAG 2.x contrast is limited to web content and even further limited to the specific technology of an sRGB display, not valid elsewhere (for that matter not particularly valid for sRGB). If ICT is supposed to be used for technology agnostic applications, then WCAG2 contrast is a nonstarter. >> >> APCA is adjustable to any realizable color space. The W3 flavor is already equipped for sRGB, Display P3, and Adobe98 RGB. >> >> If you are referring to the “less than permissive” license, that was due to some bad actor’s monkey business. >> >> APCA is a subset of SACAM, apca-w3 is intended for WCAG 3, APCA-RC (which I linked to) is independent, part of the Inclusive Reading Technologies initiative, and scoped for apps and other use cases. There are other flavors that are not public as yet. >> >> It is an adjustable model, and the model can be adjusted for other contexts including e-ink, paper (traditional print), physical signage… >> >> I’m not sure if this response answers all your questions, but your comments do bring up further reasons to remove 1.4.3 and 1.4.11 (6.1.5.4 and 6.1.5.8) from this draft of ICT. >> >> Thank you, >> >> Andy >> >> Sent from Andy’s iPhone >> >> >> On Aug 3, 2023, at 02:22, Abou-Zahra, Shadi <sabouzah@amazon.at <mailto:sabouzah@amazon.at>> wrote: >> >> >> Hi Andrew, >> >> Please correct me if I’m wrong but I think the current APCA-W3 is limited to web content only (and on self-illuminated displays only), and WCAG2ICT is specifically intended for other contexts – apart from the fact that we can’t have WCAG2ICT deviate from WCAG itself, your suggestion does not seem workable to me. >> >> Also, I believe the APCA-W3 license only covers use in WCAG and not in other documents (including other technical standards that adopt WCAG2ICT, such as EN301549) but I guess that is more of a logistical question on licensing terms and agreements. >> >> Best, the >> Shadi >> >> --- >> Shadi Abou-Zahra >> Amazon Devices and Services >> Principal Accessibility Standards and Policy Manager >> --- >> >> >> From: Andrew Somers <Andy@GeneralTitles.com <mailto:Andy@GeneralTitles.com>> >> Sent: Thursday, 3 August, 2023 10:34 AM >> To: WCAG list (w3c-wai-gl@w3.org <mailto:w3c-wai-gl@w3.org>) <w3c-wai-gl@w3.org <mailto:w3c-wai-gl@w3.org>>; public-silver@w3.org <mailto:public-silver@w3.org> >> Cc: Chuck Adams <charles.adams@oracle.com <mailto:charles.adams@oracle.com>> >> Subject: RE: [EXTERNAL] CFC - Publish WCAG2ICT First Public Working Draft >> >> CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you can confirm the sender and know the content is safe. >> >> -1 >> Objection. >> >> This is the first opportunity I’ve seen to weigh in on this. >> >> SUMMARY >> >> While it has been established that WCAG 2.x is bound to require a certain backwards compatibility, that does not mean we must promote SCs that are unsupported by modern science into yet another new document, to continue to wreak havoc and misunderstanding in the accessibility and designer/developer communities regarding this important topic. >> >> Thank you, >> >> Andy >> >> Additional Reading For a Deeper Dive: >> _____________________ >> BACKGROUND >> >> My objection is rooted in my long-standing objections¹ to certain WCAG 2.x SCs which can be harmful to readability, as well as being confusing, not aligning with human perception, and resulting in unintended consequences²⁻³. For WCAG those include 1.4.3 and 1.4.11, regarding the present document, these are 6.1.5.4and 6.1.5.8 >> >> 6.1.5.4 (1.4.3) does not predict nor align with the human perception of contrast on self illuminated displays, and in fact the values it generates can create conditions that are harmful to readability, particularly for those with color vision deficiencies such as protanopia⁴. 6.1.5.8 (1.4.11) also ignores the primary drivers of contrast perception, the spatial characteristics⁵, which arguably are most important for non-text. 1.4.11 reuses the faulty math and an arbitrary threshold from 1.4.3, citing an unpublished self-referential document for support. (The cited informal email actually implies the importance of spatial characteristics, but not otherwise recognized/developed in that SC.) >> >> >> NOT JUST ME >> >> Independent peer review demonstrates that with a set of thousands of random color pairs, 47% that “pass” WCAG2 should instead be rejected. Paradoxically, ~22% that are rejected should pass⁶, and in fact are demonstrably better for color vision deficiencies than the related colors that were incorrectly passed. >> >> And no peer review is required to see that 1.4.3 does not properly calculate nor predict useful colors for the increasingly popular dark mode—in this use case adhering to the WCAG2.x contrast results creates an unreadably poor contrast value for all users, not even considering impairments⁷. This becomes important for automation.⁸ >> >> >> >> >> >> >> AUTOMATION >> >> Of course, no designer or developer would intentionally use such poor colors (this being one big reason the design community has disregarded WCAG 2.x), but the need for and push for automated color adjustmentdemands a functional replacement for WCAG 2.x contrast, referred to as "not-fit-for-purpose” as one peer reviewer stated.⁹ >> >> Automation or automated color selection, such as that envisioned by the CSS color-contrast() function, absolutely requires a perceptually uniform prediction method. Without it, automated color adjustment becomes meaningless and unreliable. >> >> One example is that of spreadsheet automatic text color, flipping from black to white, depending on the background cell color in a spreadsheet. WCAG 2.x contrast puts that flip point at a significantly incorrect value, resulting in poor to unreadable combinations. And here is an example of an unintended consequence: an application that is doing the color flip accurately will degrade the user experience if the app is modified to instead use the invalid flip point resulting from WCAG 2.x >> >> >> HOW TO CORRECT >> >> There are other methods¹⁰ and guidelines written¹¹ which have been subject to public testing for years, and have ample third party and peer review, including a recent journal-published review.¹² >> >> At this point, there is no valid reason to create new documents that cite obsolete methods, nor is there a reason to demand backwards compatibility here. >> >> A “simple” replacement method is the Bronze Simple Mode <https://readtech.org/ARC/tests/bronze-simple-mode/?tn=intro>¹³, developed as a way to use perceptually uniform contrast prediction methods, with a guideline of simple thresholds similar to the WCAG 2.x SCs (no lookup table). This is a criterion that echos the simplicity of WCAG 1.4.3 but is, at the very least, using perceptually uniform math/methods. >> >> >> References: >> 1. https://github.com/w3c/wcag/issues/695 >> 2. https://tangledweb.xyz/please-stop-using-grey-text-3d3e71acfca8?sk=7caf4a84900aec25cfec31bf6fa996af >> 3. https://uxdesign.cc/better-reading-on-the-web-c943c4cfc91a?sk=f84c349a331d0290028cb76cca36615e >> 4. https://tangledweb.xyz/whats-red-black-also-not-read-573b9c0a97ed >> 5. https://www.smashingmagazine.com/2022/09/realities-myths-contrast-color/ >> 6. https://www.cedc.tools/article.html >> 7. https://github.com/Myndex/SAPC-APCA/discussions/30#discussioncomment-1888870 >> 8. https://github.com/w3c/csswg-drafts/issues/7310#issuecomment-1145775322 >> 9. https://github.com/w3c/low-vision-a11y-tf/discussions/131#discussioncomment-1553545 >> 10. https://apcacontrast.com <https://apcacontrast.com/> >> 11. https://readtech <https://readtech/>.org/ARC/ >> 12. https://git.myndex.com/#apca-peer-review--third-party-discussion >> 13. https://readtech.org/ARC/tests/bronze-simple-mode/?tn=intro >> >> >> >> Andrew Somers >> Senior Color Science Researcher >> PerceptEx Perception Research Project <https://www.myndex.com/perceptex/> >> redacted for public list >> >> >> >> On Aug 1, 2023, at 1:05 PM, Chuck Adams <charles.adams@oracle.com <mailto:charles.adams@oracle.com>> wrote: >> >> Call For Consensus — ends Monday August 7th at midday Boston time. >> >> We previously agreed upon accepting the First Public Working Draft of WCAG2ICT. No concerns were raised in our pre-cfc email. >> >> This call is to approve publishing the First Public Working Draft of WCAG2ICT. The draft document can be viewed here: https://w3c.github.io/wcag2ict/ >> >> Kind regards, >> Charles Adams >> >> >> >> >> Amazon Development Center Austria GmbH >> Brueckenkopfgasse 1 >> 8020 Graz >> Oesterreich >> Sitz in Graz >> Firmenbuchnummer: FN 439453 f >> Firmenbuchgericht: Landesgericht fuer Zivilrechtssachen Graz
Received on Thursday, 3 August 2023 22:15:19 UTC