RE: Mikes request that we identify an upper limit on the number of digits

Thanks John, I was going to bring that up as well.

Dredging up memories from my ancient psychology degree, the 7 +/-2 thing is very ‘over applied<https://twitter.com/alastc/status/932650988506411009>’ in web design (articles), but for the scenario of reading a 4-8 digit number (or password) and typing it in, Miller’s law is very applicable.

And from there assuming that people with cognitive issues struggle to remember, struggle to perhaps read the correct line in time before the damn things changes… I buy the need.

Finding more than one or two characters to transcribe is probably fairly common, in fact that’s probably why there is little research on it because it is so obviously a problem for some people. It is just an on/off thing, rather than X number of people can remember more than Y characters.

So I agree with the way the SC doesn’t specify a cut off of X characters, it is all or nothing. (This conversation about cut-off is odd, it isn’t coming from the SC, but a number Mike found in certain research.)

Having said that, I don’t see it as feasible. The mechanisms that have been raised so far are fine for large (very large) companies, but then you are pricing out accessibility conformance for smaller companies.

More on that on another thread: https://lists.w3.org/Archives/Public/w3c-wai-gl/2017OctDec/0722.html


-Alastair