Intent to Implement and Ship: autocapitalize attribute

[Mounir Lamouri]

* Contact emails:
mlam...@chromium.org

* Spec:
Base on Apples' proprietary API:
https://github.com/mounirlamouri/html-autocapitalize/blob/master/proposal.md

* Summary:
Enables web pages to give hint with regards to the autocapitalization
behaviour to use on text fields.
(See proposal.md above if you want more information, it's fairly short.)

* Platform Support:
The feature only makes sense on platforms that use Blink's IME flags.
These are Android and ChromeOS.
I would like to first ship on Android (including WebView) where this
feature is the most important.
The ChromeOS team will only have to use the Blink flags for their own
virtual keyboard.

* Demo link:
http://mounirlamouri.github.io/sandbox/autocapitalize.html

* Compatibility Risk:
Apple, Microsoft and Mozilla expressed support. Apple's proprietary
attribute is commonly used on the Mobile Web. There are millions of
pages using it.
Shipping this feature will _improve_ compatibility.

* OWP launch tracking bug
https://crbug.com/466930

* Link to entry on the feature dashboard
https://www.chromestatus.com/features/4529989986811904

-- Mounir

[Rick Byers]

Yes, lets do this!  Thank you!

[Chris Harrelson]

LGTM

To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+...@chromium.org.

[Dimitri Glazkov]

LGTM2.

:DG<

[Philip Rogers]

LGTM4

[Philip Jägenstedt]

Hi Mounir,

Can you file a bug to integrate your proposal into HTML? It's likely that Hixie would have useful technical feedback, and it will ensure that the autocapitalize spec doesn't get out of sync with the HTML spec's terminology over time.

This is the discussion on the WHATWG list:

https://lists.w3.org/Archives/Public/public-whatwg-archive/2015Mar/thread.html#msg22

Where did Microsoft leave feedback, and what did they say?

As for the proposal itself, the only issue I can see is in the minutiae of the states and reflection. Making the invalid value default depend on the input element state is atypical, a quick grep of the HTML spec and I can't find any other such case. Judging by WebKit's source, they ignore the input type for purposes of reflection. Can we do the same, and just ignore the autocapitalize attribute for input types other than text or search?

Philip

[Mounir Lamouri]

I've reopened the spec bug that Apple opened in 2011:
https://www.w3.org/Bugs/Public/show_bug.cgi?id=12409

I'm fine with having the attribute apply only to Text and Search states
given that Android (at least default keyboard) will only take into
account capitalization hints for text keyboards. However, I don't think
it's a decision that blink-dev should make. Let's discuss that in
whatwg@ or the bug. I will make sure to ship the final decision.

-- Mounir

On Mon, 16 Mar 2015, at 04:14, Philip Jägenstedt wrote:
> Hi Mounir,
>

> Can you file a bug <https://whatwg.org/newbug> to integrate your proposal

> into HTML? It's likely that Hixie would have useful technical feedback,
> and
> it will ensure that the autocapitalize spec doesn't get out of sync with
> the HTML spec's terminology over time.
>
> This is the discussion on the WHATWG list:
> https://lists.w3.org/Archives/Public/public-whatwg-archive/2015Mar/thread.html#msg22
>
> Where did Microsoft leave feedback, and what did they say?
>
> As for the proposal itself, the only issue I can see is in the minutiae
> of
> the states and reflection. Making the invalid value default depend on the
> input element state is atypical, a quick grep of the HTML spec and I
> can't
> find any other such case. Judging by WebKit's source

> <https://trac.webkit.org/browser/trunk/Source/WebCore/html/Autocapitalize.cpp?rev=160733>,

[Philip Jägenstedt]

Thanks Mounir!

To ensure that we get feedback from Hixie and that it ends up in both specs if deemed sane, I also filed a separate WHATWG bug:

https://www.w3.org/Bugs/Public/show_bug.cgi?id=28228

I'll also copy me spec feedback to that bug.

Definitely LGTM to implement right away, and I hope that there's some feedback on the spec bug(s) before it ships. However, I think the details of how reflection works and what happens for input types other than text and search are small enough that shipping with them unresolved would be OK.

Thanks for driving this in spec and implementation!

Philip