W3C

- DRAFT -

AGWG teleconference

22 Sep 2020

Attendees

Present
(sorry, AWK, Brooks, Chuck__, Detlev, Detlev_(sorry, Francis_Storr, GN, Glenda, Grady_Thompson, JakeAbma, Jennie, JohnRochford_, Katie_Haritos-Shea, Laura, Levon, MelanieP, MichaelC, Nicaise, Rachael, Raf, Sukriti, Wilco, alastairc, bruce_bailey, juliette_mcshane, kirkwood, mbgower, meeting, meeting_overran), overran), sarahhorton, stevelee
Regrets
JustineP, BruceB, DavidF, CharlesH, Wilco
Chair
alastairc
Scribe
Laura, kirkwood

Contents


<laura> Scribe: Laura

<kirkwood> yes

<AWK> +AWK

Accessible Authentication https://www.w3.org/2002/09/wbs/35422/wcag22-accesssible-auth-updates/results

ac: Accessible Authentication Results. 7 questions

Change "must" to "is" in the SC text #1280

ac: simple change. But it is normative.

<alastairc> https://github.com/w3c/wcag/pull/1408/files

ac: 13 agree with the change

suggest resolve to accept.

<Chuck__> +1 to accept

RESOLUTION: accept pull request 1408

Clarify exceptions in definition #1310

scribe: second topic Clarify exceptions in definition.

<alastairc> https://github.com/w3c/wcag/pull/1411/files

scribe: It was pointed out in github issue 1310 that exceptions need to be part of the definition, as notes are not normative. Pull request 1411 shows the change in the SC text, which is to move the note for "cognitive function test" into the main definition.
... all agree accept awk
... It is still memorization, even if it is easy because of repetition.

awk: The right way to handle this is not to say that remembering your own email address isn't memorization, it is to exempt these specific items that are memorized from the requirements of the SC.

<alastairc> Current proposed text: "memorization, such as remembering a username, password, set of characters, images, or patterns. Common personal identifiers such as name, e-mail, phone number, and handle on social media are not considered memorization as they are consistent across websites;"

<GN015> I n the first resolution above, should it say request 1280 instead of 1408?

sl: agree it is an issue. Not sure what to do about it.

ac: bit of a theme with this SC.

<kirkwood> long term memory is not often referred to as “memorization” think this is confusing

<kirkwood> +1 to Alastair

ac: line we have been drawing 1.) things that about you and 2. ) things that are about the web site.

<mbgower> "...are considered an acceptable level of memorization, as they are familar to the user and consistent across websites;"

<alastairc> q/

jk: making it confusing regarding memorization

<Glenda> could we say “not relying on short-term or mid-term memory”?

ac: can we keep it in the definition and make it clearer?

<mbgower> "...are not considered short- or mid-term memory, as they are familiar to the user and consistent across websites;"

awk: let me take a look here.
... I guess we could.

<alastairc> Current bullet: "memorization, such as remembering a username, password, set of characters, images, or patterns. Common personal identifiers such as name, e-mail, phone number, and handle on social media are not considered memorization as they are consistent across websites;"

<Glenda> we do have a handy dandy list of user identified info for 1.3.5 https://www.w3.org/TR/WCAG21/#input-purposes

<Zakim> mbgower, you wanted to say I've posted a suggested phrase. We can go into detail in the Understanding

awk: need to be specific as to what is allowed.

mg: need to explain in the understanding doc.

<mbgower> "...are not considered short- or mid-term memory, as they are familiar to the user and consistent across websites;"

ac: we have a couple of suggestions.

glenda: can we used the list we already included?

ac: most of them could work. but not all of them.

jr: memory types don’t make a difference.

<alastairc> akc aw

<Jennie> +1 to John R's suggestion

<alastairc> "Common personal identifiers such as name, e-mail, phone number, and handle on social media are not considered cognitive function tests as they are consistent across websites;"

awk: easiest fix “they are not considered cognitive functional tests”

<Chuck__> ack "such

<Chuck__> ack as"

mg: concerned with "such as”. not restrictive enough.

<kirkwood> social

<Glenda> +1 to what mbgower just said. let’s list them specifically. so it isn’t fuzzy or open to different interpretation.

<alastairc> The common personal identifiers name, e-mail, phone number, and handle on social media are not considered cognitive function tests as they are consistent across websites;

ac: any issues with the last wording?

awk: do we have a definition for social media?

ac: no
... if the company allows the social login for other sites may work.
... we are trying to allow authors to allow 3 party authentication
... like google and facebook

jk: People are using password managers. We should have that in mind.

ac: considered password managers and social media logins separate

jennie: maybe the problem is the term “social media handle” is the problem?

<Levon> I treat password managers like an accommodation, so I don't have to rely on my poor memory.

ac: don’t know what we have “social media handle” now.

<mbgower> The common personal identifiers name, e-mail, and phone number are not considered cognitive function tests as they are consistent across websites;

<Zakim> mbgower, you wanted to say take it out :)

ac: maybe we should just remove it.

wilco: something like twitter is very common.

<mbgower> The common identifiers name, e-mail, and phone number are not considered cognitive function tests as they are personal to the user and consistent across websites;

ac: do you need to rely on it in this scenario?

wilco: no.

<kirkwood> +1

<Chuck__> +1

ac: any objections to: “The common identifiers name, e-mail, and phone number are not considered cognitive function tests as they are personal to the user and consistent across websites”?

<JohnRochford> +1

<Jennie> +1

Laura: +1

<sarahhorton> +1

RESOLUTION: Accept amended pull request 1411, close issue 1343 and incorporate feedback from 1344

Understanding flow #1313

ac: top of the understanding document gets bogged down in 'cognitive function test' and loses sight of the authentication scope.

<alastairc> https://github.com/w3c/wcag/pull/1412/files

<alastairc> New text: "The general case of remembering a password is a <a>cognitive function test</a>, which are known to be problematic for many people with cognitive disabilities. "

ac: I put together a change in the understanding doc, which adds a line near the top.

<stevelee> +1 to Rachael

ac: Rachael said "The general case of remembering a password..." seems needlessly complex. Can we just say "Remembering a password..."

<Jennie> +1 to Rachael

ac: MG said I just want to break it up to be grammatical: <p>The general case of remembering a password is a <a>cognitive function test</a>. Such tests are known to be problematic for many people with cognitive disabilities.

<alastairc> Remembering a password is a <a>cognitive function test</a>. Such tests are known to be problematic for many people with cognitive disabilities.

ac: any objections to the change?

<kirkwood> +1

Laura: +1

<JohnRochford> +1

RESOLUTION: Accept amended pull request 1412

Confusing example: properly marked up email and password #1363

ac: 1st example in the understanding document confuses email & username issues.

<alastairc> Proposed text: A web site uses a properly marked up username and password as the login authentication. Copy &amp; paste is not blocked on the inputs, so the user's browser or extension can identify the purpose of the inputs and automatically fill in the username and password

ac: As well as pointing out the note on the same topic (which essentially answers the question), there is a small change outlined in Pull request 1415 that shows the change in the understanding doc, which adds to the first example.
... any objections to the text?

RESOLUTION: Accept pull request 1415

Clarify whether copy/paste is a cognitive function test #1359

ac: smhigley pointed out that there are cases (such as web-based command line tools) where copy-paste is necessary.
... question around making it more explicit.

<kirkwood> we should keep in mind going to a public machine

ac: From previous discussions, copy-paste is necessary for things like password managers, and we do not want to discourage that.
... However, allowing for any copy-paste to 'pass' can lead to very intricate processes where you have to swap apps multiple times whilst maintaining an item in the clipboard within 30 seconds.

jr: (inaudible)

ac: not sure it is available to web content

<kirkwood> we can not allow copy paste.

sl: question is if copy/paste is available or if we are requiring people to copy/paste.

ac: author may not know.

jr: preventing copy/paste is terrible.

ac: we talk about that in the understanding doc.

jk: have copy/paste in critical path NYC kiosk inaccessible.

<Zakim> AWK, you wanted to point out my suggestion in issue 1418

awk: suggest the problem is a problem with the kiosk

<Zakim> mbgower, you wanted to say that a kiosk is likely an exception. WIll we have any exception?

mg: kiosk is a closed system. Not applicable.

<kirkwood> don’t think so

mg: are there situations where this is not possible?

<AWK> "closed environment" rather than "closed system"

mg: should be ways of doing it.

jr: assume this will be in the understanding. doc.
... should work for everybody.

mg: I have had very high security clearance. used app on iphone, etc.
... some extreme cases but I don’t think we have to address them.

<Jennie> * But their supervisor could be blind, and may need to know what the training covers.

<alastairc> "transcription, such as typing in characters. Copy and paste, such as that automated by a password manager, is not considered transcription but requiring multiple steps or applications to complete a paste is a cognitive functional test;"

ac: PR talks about updating the definition of transcription.

<mbgower> +1 to tackling in understanding

ac: probably should simplify to Copy and paste is not transcription.
... is that okay with coga folks?

jennie: time where formatting may be different. Is that something for the understanding doc?
... yes.
... gives example with a hyphen

<JohnRochford> What a great memory Jennie has!

ac: when copying from an SMS you get the full message

<GN015> Sorry, I have to leave early today. have a good meeting!

ac: suggest not to accept PR and update understanding doc

<alastairc> Proposal: Do not accept the PR, propose a new understanding, allowing for copy/paste.

<Wilco> +1

<Jennie> +1

Laura: +1

<Levon> +1

RESOLUTION: Do not accept the PR, propose a new understanding update, allowing for copy/paste.

<alastairc> scribe:kirkwood

Include each step in an authentication process #1358

<alastairc> https://github.com/w3c/wcag/pull/1414/files

AC: fairly straight forward, think multistep. change is here
... just changin to ‘for each step’
... compare between choice options, assumed applied across steps

SteveL’ one or two factors and then all the rest

AC: each stage has to pass

AD: reworded what we had previously

SL: that’s good wasnt quite sure what the issues was

AC: anyoone abject to except pull request

Wilco; seems like sam requirment, what is the differnce?

AC: you could have a passing method for first but not second

WF: seems like it would pass

AC: if not negative impact it would be helpful
... blocker for you WF?

WF: no

AC: no objections

RESOLUTION: Accept pull request 1414

Does this include CAPTCHAs? #1256

AC: last of authentication questions
... what i suggested taking picure of common objects, remeber a cat in future
... mulifunction test would not include personal info
... need to expand on examples
... the example was identify which image contained image
... need to add something around cpaptures and examples
... do people agree with approach
... think we should update understanding to make clearer
... agree with Sarah comments
... good approach?

SL: PR doesn’t seem to match
... culture issues is there
... going to add sums, dysclaculua issue?

AC: that is a failure
... there are formatting issues which i fixed
... andrews point was addressed
... bruce agrees its tough
... see if people agrees with raitionale

MG: understand

<laura> s/. not restrictive /. It is not restrictive /

AC: its a greay line
... have used capture as authentication
... do agree that balancing act and need too update understanding
... not hearing objections

RESOLUTION: Accept amended pull request 1407 and we will update the understanding doc instead.

Target pointer spacing https://www.w3.org/2002/09/wbs/35422/wcag22-target-spacing-issues/

New (UPDATED) target spacing proposals #1312 & #1361

AC: next have target spacing, only first question, have several issue about spacing
... we got to situration trying to make target samller, sems like be better dealt with personlization

<alastairc> Proposal: "The size of the target for pointer inputs is at least 26 by 26 CSS pixels except when:"

AC: make smaller minimal siaze 26 x 26 with same exceptions as previously

SH: had some questions about affect of it, my understanding moving aswy from spacing and towards specifying target size, seems name should be updated
... to reduce activate control using space if we should circle back to that spacing for this SC

AC: yes generally we know bigger better, so many scenerios inroduced issues so think AA
... came because even using spacing it still impacts heavily

<alastairc> https://www.w3.org/2002/09/wbs/35422/wcag22-target-spacing-issues/results

DF: no seeing in survey, is it in there?

AC: yes

DF: see it now

<laura> s/definition of transcription /definition of transcription /

DF: you put it quite well, discussed spacing option. to combine minimal target size and spacing and found didn’t go well together, hard to control, diifficult to express
... smaller size would go well with AAA
... 26 acceptable, but dropdowns, tqables often smaller than 26. just link with no p;adding. difficult

WF: might need to think about, not conviniced smaller target with spacing rather than larger target with no space. no evidence seen with targets being flush next wo other
... think current is more clear

AC: did 26 come form somewhter?

DF: don’t remember Micorosoft or empircle google docs, email , if it was a common size Jake meay now

<AWK> "Kathy from the TF noted: "The 24 CSS px comes from Microsoft’s research where they are recommending a target area of 24px with a minimum of 16% spacing. If we look at the minimum that is 3.8px so we rounded that to 4 CSS pixels.""

<alastairc> AWK - that was their prefered icon size, it wasn't a usability/accessibility recommendation

Jake: ther was research with 24 and 4 pixel spacing, there were combined minimal size with spacing
... why 24 with space or 28 without
... so according to all comments basic thought was to focus on AA version of target size

Jake; it was 24 then 24 may be Kathy or others gave input in discusssion

<Glenda> “Microsoft's Windows Phone UI Design and Interaction Guide suggests a touch target size of 34px with a minimum touch target size of 26px.” https://www.smashingmagazine.com/2012/02/finger-friendly-design-ideal-mobile-touchscreen-target-sizes/

Jke: t was 24 then 26 may be Kathy or others gave input in discusssion

<Sukriti> Alaistar, Kathy mentioned your research for 26

Jake: seemed 26 was best alternative

Ac: we’ll come back to specific number

WF: occurs to me all links would fail this
... seems extreme

Jake: i was thinking the same

]Jake: every interacttive element below might need spacing for example

Jake: couldn’t think of logical defensible one
... there is a need but where to draw line need help

AC: 44 Wilco sayd pixel spacing goog for some scenerios
... this version captures something different

DF: one option we have is unless element is a native element, if you have text just text, native link, not ure, as soon as you start exceptions list will grow longer, such as in line links, wrapping to second line becomes an option, is list in line, lots of issues
... what is in scope what are we trying to get at.
... in line text or dropdown list not sure where to draw line

BrooksN: target on bulls eye for example

BN: what i’m getting at is about finger size issue
... stylus versus finger, is this really getting at user need here
... can imagine use of wrong click but not too small

WF: if you put content in its not user agent, i cna only imagine user agent controls like check boxes and radio buttons
... seems we are getting sutck here, maybe we should scope out to say these things at the very least should be explored. limiting to buttons or menus

Jake: we are hitting some dead ends, started more than a year ago, preventing accidental activation for target smaller than 44 x 44 then went smaller to 26 x 26
... if need to define size, with perfect list we have two forces at play.
... if make too big can’t put enough info on page might have push back

<Brooks> I haven't heard of anyone skinny-thumbing a mobile touch target. I'm not sure the problem to solve for the mobile space is creating rule that makes it easier to activate a target. It seems to me that the intent of this SC is to prevent accidental activation of nearby targets. An SC that only speaks to minimum target size doesn't solve the original problem.

Jake: provide spacing to cover 44 and at least 24 by 24 stil ned spacing as mentioned in github
... we got rid of spacing and even tried AAA
... so difficult

AC: getting something in AA doesn’t seem likely

<alastairc> Brooks - The other aspect is that you need to know about is the heuristics that the devices use, if you miss a small target it will often guess which you meant. So spacing them out helps.

DF: differnce between pointer and touch. with touch harder to hit small targes, but wanted to be conscious of that, didn’t want to create a carve out niche
... thats why stuck with what work for both. probleme with spacing allow for target very small. then designer would make target very samll meking target too smalll
... need to figure out how to combine minimum target size with miminum spacing

WF: there may be some way out of hole but may need more work in tadk force

BN: i agree iwht Detlev with comibined approach of target and spacing

AC: when talking target size of spacint. how links in vericlaly spce list without going off screen,
... maybe good to go on to comments in survey, don’t fhink we’ll find research around the size
... what they use for minimum size, and other examples can show, don’t want ot create a increased workload across industry
... anything above 16 for example will capture quite a lot
... its always going to be somewhat arbitrary
... glenda put a link in around 26

Glenda: its from windows 2012

<sarahhorton> https://docs.microsoft.com/en-us/windows/uwp/design/style/app-icons-and-logos#more-about-app-icon-assets

GS: thats what i think its about size of human finger

AC: might be a way forward

<sarahhorton> https://developer.apple.com/design/human-interface-guidelines/macos/icons-and-images/custom-icons/

<Glenda> “A past study from the MIT Touch Lab found that the average person’s fingertips are 1.6– 2cm (0.6­­–0.8 in) wide. The impact area of the typical thumb is even larger — an average of 2.5cm (1 inch) wide! Designing touch targets to account for the physical dimensions of users is basic user-centered design.”

<Glenda> https://www.nngroup.com/articles/touch-target-size/

AC: not really saying about good size target in Microsoft material

GS: a MIT touchlab reference average person finger size

AC: not even center of finger

SH: thinking about it phisical world we have specification. if building a building

<Glenda> cool references at the bottom of the NNGroup article: References

<Glenda> Parhi, P., Karlson, A. K., and Bederson, B. B. 2006. “Target size study for one- handed thumb use on small touchscreen devices.” In Proceedings of the 8th Conference on Human-Computer interaction with Mobile Devices and Services. MobileHCI ‘06. DOI= http://doi.acm.org/10.1145/1152215.1152260

<Glenda> Dandekar K., Raju B.I., Srinivasan M.A. (2003). 3-D finite-element models of human and monkey fingertips to investigate the mechanics of tactile sense. Journal of Biomechanical Engineering, 125, 682–691. DOI= 10.1115/1.1613673

SH: we producing operating systems can adopt determination

<Glenda> Look at this research: https://dl.acm.org/doi/10.1145/1152215.1152260

SC: looks like average size of finger is 40 pixels, and 24 touch size

<Chuck__> kirkwood: having difficult time with this, this is an arbitrary number. There are user agent functionality that is built into every device where...

<Chuck__> kirkwood: you are able to zoom in. As far as touch target size is concerned, I don't think that... we just need to pick a number. We can't compare...

<Glenda> “The results showed that while speed generally improved as targets grew, there were no significant differences in error rate between target sizes =9.6 mm in discrete tasks and targets =7.7 mm in serial tasks. Along with subjective ratings and the findings on hit response variability, we found that target size of 9.2 mm for discrete tasks and targets of 9.6 mm for serial tasks should be sufficiently large for one-handed thumb use on touchscreen-based

<Glenda> handhelds without degrading performance and preference.”

<Chuck__> kirkwood: a point size with digital and printed materials, it's arbitrary.

<alastairc> Kirkwood: Difficult time with this, it is an arbitary number. There's UA features built into screen devices where you zoom in, so the target size can get bigger. We just have to pick a number. We can't compare point size like we can with printed size, and not like physical architecture either.

<Chuck__> kirkwood: door sizes make sense in physical world, but doesn't make sense in digital world.

<alastairc> ... just pick a number

<Detlev> The size of the target for pointer inputs, or the non-interactive space surrounding the target, measures at least 26 by 26 CSS pixels except when

DF: trying to include target size suggestion with spacing in it

<alastairc> Suggested update: "The size of the target for pointer inputs, including non-interactive space surrounding the target not shared with other targets, measures at least 26 by 26 CSS pixels except when:"

DF: this would leave out number and issue of shared spacing and small target size

AC: suggest with shared spacing need to be addressed

DF: gave examples not sure

GS: while we are dealing with digital stuff this is the moment when physical

<Sukriti> http://touchlab.mit.edu/publications/2003_009.pdf

GS: thinks mobile HCI not free but has been scientific research on human beings and digital interfaces
... prefer a study rther then 24 sounds good

AC: think in where to go narrowing down to agree that 44 not going to work

<Glenda> @Sukriti, it is a different research article. I think both the one you referenced and THIS one should be reviewed..so we can use this research to explain whatever size is finally chosen. Parhi, P., Karlson, A. K., and Bederson, B. B. 2006. “Target size study for one- handed thumb use on small touchscreen devices.” In Proceedings of the 8th Conference on Human-Computer interaction with Mobile Devices and Services. MobileHCI ‘06. DOI=

<Glenda> http://doi.acm.org/10.1145/1152215.1152260

AC: seems to setup things getting push back, seems llik focusing in on a smaller size
... think covers survey questions

<alastairc> https://github.com/w3c/wcag/issues?q=is%3Aissue+is%3Aopen+label%3A%222.5.8+Pointer+Target+Spacing%22+-label%3A%22Survey+-+Added%22+-label%3A%22Duplicate%22+

AC: like to get proposal is with explainer texxt with where number came from out to those on github that raised issue
... would be useful to get organization feedback

<alastairc> https://www.w3.org/WAI/GL/wiki/WCAG_2.2_Issue_tracking_and_resolution

<sarahhorton> A note for the proposal, update the SC title to reflect the updated intent

AC: would appreciate read through one issue and propose response, we have somehting like 10 issues per success criteria
... need people to read and respond
... need to agree then work through
... any questions?

presnet+

Summary of Action Items

Summary of Resolutions

  1. accept pull request 1408
  2. Accept amended pull request 1411, close issue 1343 and incorporate feedback from 1344
  3. Accept amended pull request 1412
  4. Accept pull request 1415
  5. Do not accept the PR, propose a new understanding update, allowing for copy/paste.
  6. Accept pull request 1414
  7. Accept amended pull request 1407 and we will update the understanding doc instead.
[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version (CVS log)
$Date: 2020/09/22 16:59:05 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision of Date 
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Succeeded: s/sinacio/scenario/
Succeeded: s/auther/author/
Succeeded: s/Resutls/Results/
Succeeded: s/aggree /agree /
FAILED: s/. not restictive  /. It is not restrictive  /
Succeeded: s/comapny /company /
Succeeded: s/tryig /trying /
Succeeded: s/and. facebook/and facebook/
Succeeded: s/consideedr /considered /
Succeeded: s/socal /social /
Succeeded: s/someting /something /
Succeeded: s/probelm /problem /
Succeeded: s/secuity /security /
Succeeded: s/addess /address /
FAILED: s/defintion of transciption /definition of transcription /
Succeeded: s/transciption/transcription/
Succeeded: s/fairyl straignt/fairly straight/
Succeeded: s/restictive /restrictive /
Succeeded: s/36/24/
Succeeded: s/i is available /it is available /
Succeeded: s/peole /people /
Succeeded: s/defintion of transcription/definition  of transcription/
Succeeded: s/transciption/transcription/
Succeeded: s/whhen copying from an sms /when copying from an SMS /
Succeeded: s/not accept PR /not to accept PR /
Succeeded: s/understanding doc/understanding doc instead./
Succeeded: s/restictive /restrictive /
Succeeded: s/defintion /definition /
Succeeded: s/transciption/transcription/
Default Present: Chuck__, MichaelC, Rachael, alastairc, Laura, Francis_Storr, stevelee, Raf, MelanieP, AWK, Nicaise, kirkwood, GN, sarahhorton, Jennie, Glenda, Brooks, Grady_Thompson, juliette_mcshane, mbgower, Sukriti, Levon, Katie_Haritos-Shea, JohnRochford_, Wilco, Detlev, (sorry, meeting, overran), JakeAbma, bruce_bailey
Present: (sorry AWK Brooks Chuck__ Detlev Detlev_(sorry Francis_Storr GN Glenda Grady_Thompson JakeAbma Jennie JohnRochford_ Katie_Haritos-Shea Laura Levon MelanieP MichaelC Nicaise Rachael Raf Sukriti Wilco alastairc bruce_bailey juliette_mcshane kirkwood mbgower meeting meeting_overran) overran) sarahhorton stevelee
Regrets: JustineP BruceB DavidF CharlesH Wilco
Found Scribe: Laura
Inferring ScribeNick: laura
Found Scribe: kirkwood
Inferring ScribeNick: kirkwood
Scribes: Laura, kirkwood
ScribeNicks: laura, kirkwood

WARNING: No date found!  Assuming today.  (Hint: Specify
the W3C IRC log URL, and the date will be determined from that.)
Or specify the date like this:
<dbooth> Date: 12 Sep 2002

People with action items: 

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)


[End of scribe.perl diagnostic output]