Re: Accessible Authentication and issue responses

On 24/12/2017 09:34, Alastair Campbell wrote:

>> This aspect (of the copying from a piece of paper, or presumably then
>> from another handheld device) isn't clear from the normative language.
> 
> We are not trying to specify what is happening on the user-end, just what the site cannot rely on.

You *are* if you normatively say that a site cannot rely on "transcribe 
information", aka copying.

>> Is this "copying" related to CAPTCHAs? Because again, this is a separate
>> issue I'd argue, not an authentication one. It's a "challenge" rather
>> than "authentication".
> 
> If you have to transcribe a captcha to authenticate then yes, if it is separate, no.

That's not *authentication* (who I am). It's a challenge/verification (I 
am human). While "authentication" as a general term means "verifying a 
particular fact to be true", in web/computing it has the narrower 
definition of "verifying a user's identity".

A challenge can be *part* of an authentication form/dialog, but it's a 
separate issue (just like a login/sign-in form can have things like use 
of "color alone", etc). CAPTCHAs are a topic to themselves, and 
combining authentication/CAPTCHA so blurrily into one topic is what's 
also causing problems here.

>>> Sure, it could have a username/password and the content doesn't block pasting.
>>> The kiosk may not have anything to paste from, there is no user-benefit in that scenario, but the content passes.
>>
>> Hoping that this gets very explicitly mentioned as an example in
>> understanding then. Maybe it even warrants a note in the normative
>> language, to talk about user agent/environment limitations?
> 
> Is the content requirement not clear? Do we state that you can't (usually) use a screenreader in a kiosk environment?
> 
> The understanding doc does need updating, I can help with that early next year so long as everyone can live with the SC text for now.
> But for now I'm hitting the road for Christmas, see you on the other side 😊

I'd argue that the normative language of the SC is currently (as 
demonstrated in this thread) not clear enough. Hoping the 
soon-to-be-updated understanding helps, but as it's non-normative, I 
think there's still quite a lot of gray area that makes this SC 
problematic as a whole.


-- 
Patrick H. Lauke

www.splintered.co.uk | https://github.com/patrickhlauke
http://flickr.com/photos/redux/ | http://redux.deviantart.com
twitter: @patrick_h_lauke | skype: patrick_h_lauke

Received on Sunday, 24 December 2017 11:19:03 UTC