See also: IRC log
<scribe> scribe: shadi
JW: good discussion last
time
... want to capture it some and investigate key questions
... consistent finding in studies is that audio captchas are
quite useless for everyone
... expect on-going contest between people designing audio
captchas and people trying to crack them
... just like non-audio captchas
... interesting work at University of Washington
... wonder how vulnerable algorithms are to
machine-learning
... questions approaches
JS: want to be careful claiming
"unreliable"
... think identified certain percentage
... should compare to non-disabled users
... want to quantify vis-a-vis, not just sole number
... also think need to point out that audio modality is
different than visual
... it is temporal, not like visual where one can stare for
longer
... need to be able to initiate repeat
SH: think the literature will
strengthen our arguments
... could maybe summarize key literature
... i could take that action item
... concerned that there may be hidden gems in the
literature
... want to summarize key points of the papers
JW: thinking along similar
lines
... welcome that proposal
... any reactions to that?
JB: think there may be several
research paths we could go down
... can we prioritize particular directions
... so many tempting directions
... but maybe alternative ones
SH: yes, if there is a quicker way to pull up the different options
JB: maybe pick a question like
"what alternative options have been proposed and which seem
most feasible"
... this may toss half the literature
JW: agree with Judy
<scribe> ...new research may be available
UNKNOWN_SPEAKER: captcha is
anyonymous as opposed to password, biometrics, etc
... actual tasks are different
... from authentication situation
SH: alternative focus may give faster progress
JW: hearing proposal coming
forward
... want to focus on alternatives to captcha to establish human
interaction
... seems to be agreement
... Scott offered to take that action
SH: happy to do that
JW: can do bibliographic
searches
... anyone else want to contribute?
SAZ: wonder if others not on the call may be able to help?
JW: will send out a message
JW: status and progress on literature review for other topics?
[silence]
JW: suggest we take the same
approach as with captchas
... define questions that would help us advance the
topics
... with authentication seems user requirements
... like needing to memorize information
[silence]
JS: trying to think about the
best way to approach this
... feel bad we did not take minutes with authentication group
in Lisbon
... now that it pops up in CoGa and APA
... will not get solves in any one group
... but will need a number of groups
... don't think we have an issue with the specification
... or rather we were not able to identify anything in the
spec
... some issues in the devices
... but that is beyond W3C, really
... maybe only relevant to user agents
... that would take us to Web Platforms
... need to work together to address
... concerned about pushing where there is no implementation
commitment
JW: attended that group
... planning an API that allows different authentication
mechanisms
... then cryptographic assertion about the authentication
made
... servers to implement risk analysis
... like IP, device, etc
... raises questions about user requirements
... also wonder about the server-side analysis
... if there may be differentiation factors
... for people with disabilities
... also if devices that are trusted could include assistive
devices
[silence]
JW: could try to summarize the
references we have so far
... any actions that people want to take?
[silence]
JW: will bring this back at a future meeting
<MichaelC> trackbot, status?
<trackbot> #rqtf is associated with #apa.
<jasonjgw> ACTION: jasonjgw to summarize tentative authentication questions/issues on the wiki (as derived from recent W3C work in the area). [recorded in http://www.w3.org/2017/02/08-rqtf-minutes.html#action01]
<trackbot> Error finding 'jasonjgw'. You can review and register nicknames at <http://www.w3.org/WAI/APA/track/users>.
<MichaelC> https://www.w3.org/WAI/APA/track/users
<jasonjgw> ACTION: jjwhite to summarize tentative authentication questions/issues on the wiki (as derived from recent W3C work in the area). [recorded in http://www.w3.org/2017/02/08-rqtf-minutes.html#action02]
<trackbot> Created ACTION-2114 - Summarize tentative authentication questions/issues on the wiki (as derived from recent w3c work in the area). [on Jason White - due 2017-02-15].
<MichaelC> trackbot, associate action-2114 with product-52
<trackbot> action-2114 (Summarize tentative authentication questions/issues on the wiki (as derived from recent w3c work in the area).) associated with product-52.
<MichaelC> https://www.w3.org/WAI/APA/track/products/52
<jasonjgw> ACTION: shollier to categorize CAPTCHA references and to identify those which may lead to efficacious alternatives to CAPTCHA as means of proving human (rather than machine) interaction. [recorded in http://www.w3.org/2017/02/08-rqtf-minutes.html#action03]
<trackbot> Created ACTION-2115 - Categorize captcha references and to identify those which may lead to efficacious alternatives to captcha as means of proving human (rather than machine) interaction. [on Scott Hollier - due 2017-02-15].
SH: had a look around for
literature and didn't find much
... thought of emergency situations
... for example medical devices
... other idea on new technologies replacing old ones
... multiple interfaces may be useful
... also thought of physical fallbacks
... like door knob
... or physical light switch
<jasonjgw> Shadi welcomes Scott's comments and invites him to integrate his comments into the wiki page.
<jasonjgw> Shadi also plans to review Scott's comments with regard to the question (often raised in these discussions) of whether the challenges raised disproportionately affect people with disabilities.
SH: think emergency scenario could be more specific to people with disabilities
<jasonjgw> Shadi notes a connection between problems arising in emergency situations and related work in security and disaster management.
JS: appreciate your comments
Scott, very spot-on
... but have had push-back so need to be specific about things
specific to people with disabilities
... like advice not to use elevators in case of fire doesn't
help someone who cannot do stairs
... things like that need to be teased out
... also alternatives to default interfaces
... preference for people to login manufacturer website to get
services
... like vulnerability of hacking
... not sure what the specific aspect of people with
disabilities is
... but old discussion about alternatives
... that's a fundamental principle
<jasonjgw> q?
<jasonjgw> Shadi briefly notes accessibility issues emerging in connection with payments: encompassing UI issues as well as hardware-related concerns. Payment scenarios bring together several of the topics we're already considering. Shadi asks whether the Web payments work is relevant here.
JS: long term commitment on
this
... happening in APA
... with Web Payments
... specific APIs
... big target audience
<jasonjgw> Shadi acknowledges the value of its being on the APA agenda, while noting the further issues raised by public kiosks, payment devices, etc.
This is scribe.perl Revision: 1.148 of Date: 2016/10/11 12:55:14 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Found Scribe: shadi Inferring ScribeNick: shadi Default Present: jasonjgw, MichaelC, shadi, scott_h Present: jasonjgw MichaelC shadi scott_h scribe WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth Found Date: 08 Feb 2017 Guessing minutes URL: http://www.w3.org/2017/02/08-rqtf-minutes.html People with action items: jasonjgw jjwhite shollier[End of scribe.perl diagnostic output]