ARIA membership and role="password"

[Jonathan Kingston]

<mz...@mozilla.com>
Hey,

So I was just informed that Mozilla isn't a member of the ARIA working
group which shocked me, we have however had a hand in the spec over the
years (I have cc'd those mentioned).

I notice over the years some disappointment with the spec as it being a
separate module to semantics in HTML itself however I don't see a real
opposition not to be in the group. This seems more of a formality when the
group split into two working groups.

It appears that ARIA 1.1 is moving to be resolved in the next few weeks so
if we had any objections we would need to move now I have been told.

*role="password"* has been added to the spec:
https://rawgit.com/w3c/aria/master/aria/aria.html#password and I jotted my
objections in a post:
https://jotter.jonathankingston.co.uk/blog/2016/05/16/role-password-is-not-wise/

My post was largely dismissed in this mail:
https://lists.w3.org/Archives/Public/public-aria/2016May/0126.html

Is it worth us joining? Can we discuss the wider use of ARIA itself?
Rushing to standardise features seems a shame.

Thanks

[Alexander Surkov]

Hi, Jonathan.

As far as I can tell, no one in Mozilla is super active in ARIA
nowadays; I personally (among others I think) skim through ARIA email
threads and provide occasional feedback on feature by feature basis.

One of reasons keeping me off I think is I'm worried about direction
taken in ARIA 1.1, which looks with me as to turn ARIA into universal
markup language for web page semantics description, so that ARIA
vocabulary can be used to describe whole HTML and more. I feel like
every single feature from HTML gets added to ARIA eventually, and I
not always convinced that each requested feature has a real use case
on the web. I expressed these worries, but my voice probably wasn't
strong enough to turn the wheel back.

So I'm not confident about role='password' too. While I don't have
strong objections, but I never heard the web authors complaining about
the missing feature. I would say if we do have strong concerns, and
cannot negotiate with the group to keep the feature off the spec, then
it shouldn't mean we have to implement the feature in Gecko. After all
if the feature doesn't get two implementation, then it gets removed
from the spec afaik.

Thanks.
Alexander.

On Thu, Jun 2, 2016 at 8:55 AM, Jonathan Kingston <jkin...@mozilla.com> wrote:
> Hey,
>
> So I was just informed that Mozilla isn't a member of the ARIA working group
> which shocked me, we have however had a hand in the spec over the years (I
> have cc'd those mentioned).
>
> I notice over the years some disappointment with the spec as it being a
> separate module to semantics in HTML itself however I don't see a real
> opposition not to be in the group. This seems more of a formality when the
> group split into two working groups.
>
> It appears that ARIA 1.1 is moving to be resolved in the next few weeks so
> if we had any objections we would need to move now I have been told.
>

> role="password" has been added to the spec:

[David Bolter]

Hi Johnathan,

Our lack of direct W3C ARIA involvement recently is mainly due to
time/resource constraints and we have influence via proxy.

I think ARIA is best scoped as passive semantics used to describe the
reality of how web developers (ab)use HTML. It is not unreasonable to argue
that ARIA, by existing, implicitly condones this (ab)use. That said, we
made the decision a long time ago to advocate using proper semantic HTML
(and evolve HTML where it falls short), while accepting the need for a
solution/hack for the reality of what web developers are doing.

The password case is interesting though I've not followed it closely. I'm
happy to connect you with people (including Rich) if needed (feel free ping
me offline)...

Cheers,
D

On Thu, Jun 2, 2016 at 8:55 AM, Jonathan Kingston <jkin...@mozilla.com>
wrote:

> <mz...@mozilla.com>

> Hey,
>
> So I was just informed that Mozilla isn't a member of the ARIA working
> group which shocked me, we have however had a hand in the spec over the
> years (I have cc'd those mentioned).
>
> I notice over the years some disappointment with the spec as it being a
> separate module to semantics in HTML itself however I don't see a real
> opposition not to be in the group. This seems more of a formality when the
> group split into two working groups.
>
> It appears that ARIA 1.1 is moving to be resolved in the next few weeks so
> if we had any objections we would need to move now I have been told.
>

> *role="password"* has been added to the spec:

[Marco Zehe]

An update to this: I just re-joined the ARIA-WG on Mozilla's behalf.

Marco