- From: Takayoshi Kochi <kochi@google.com>
- Date: Tue, 1 Aug 2017 12:35:23 +0900
- To: Lars Solberg <lars.solberg@gmail.com>
- Cc: public-html@w3.org
- Message-ID: <CADP2=hr5KpWSNBb3jJpemJuXA_yLoTY22nNRGOEL=MV4JUyZ-A@mail.gmail.com>
FYI, these days discussion on GitHub issues could get more attention than in this mailing list, and I found existing one at the whatwg html tracker https://github.com/whatwg/html/issues/1626 On Mon, Jul 31, 2017 at 10:41 PM, Lars Solberg <lars.solberg@gmail.com> wrote: > Correct. Fields viewed on mobile should be able to give hints about which > keyboard to use, without forcing the datatype. > Mobile could use the hinted keyboard as default, but still give the user a > way to switch the layout. > > The problem is basically that "there are no current way using a numeric > pad on keyboard, without giving (often) too much constraints on the field." > > The spec contains all sorts of things nowadays, but it feels like the some > of the most basics have been forgotten :) > > On Mon, Jul 31, 2017 at 5:06 AM, Takayoshi Kochi <kochi@google.com> wrote: > >> So you want a more hint-like attribute, rather than enforcement of the >> type, right? >> I think "inputmode" attribute originally introduced for such purpose, but >> has not gained >> enough interest from browser implementors, unfortunately. >> >> The original one was introduced before the mobile-first era, so we may >> reconsider the spec >> from the current architecture of input for modern mobile environment. >> >> A bit of history... >> >> As far as I know, once Chrome tried to implement some of Japanese >> IME-related modes in inputmode, >> but the effort was abandoned due to low interest and feasibility for >> non-Windows platforms, especially >> mobile platforms. >> https://bugs.chromium.org/p/chromium/issues/detail?id=248482 >> https://bugs.chromium.org/p/chromium/issues/detail?id=244688 >> https://bugs.chromium.org/p/chromium/issues/detail?id=642800 >> At least we should remove the script-specific types (e.g. japanese-kana) >> from the spec. >> >> Other modes have overlap with type= attribute (e.g. type=number), and >> type= is currently >> the preferred way of specifying the type of the input (also the spec is >> saying so). >> IIUC Android Chrome implements type=number. >> >> I'd think this kind of modality should not be specified via CSS, as it >> has nothing to do with styles. >> >> (FYI, IME related mode settings were available in IE and Firefox via CSS. >> https://developer.mozilla.org/en-US/docs/Web/CSS/ime-mode >> https://bugs.chromium.org/p/chromium/issues/detail?id=4490 >> which I think, is a wrong approach, either) >> >> >> On Fri, Jul 28, 2017 at 5:56 PM, Lars Solberg <lars.solberg@gmail.com> >> wrote: >> >>> Hello >>> >>> I am trying to figure out the story of the inputmode attribute for the >>> <input> element. >>> I'm seeing discussions about that attribute long back (5-10 years), and >>> based on https://caniuse.com/#feat=input-inputmode, it was supported >>> back in firefox 17-20. >>> >>> But what happened to it? Why are no browsers adopting it? Giving hints >>> to mobile platforms about which keyboardtype to use seams very powerful. >>> Using type=number sets a lot of other limitations on the field. >>> >>> The best way would be to set hints like this using css. >>> >>> Is there a work in progress here? >>> Is, and will it be type=inputmode that is the way to do this, including >>> it's limitations? >>> Will there be a bigger standard for controlling things like this on >>> mobile devices, as it becomes more popular? >>> >>> Best regards >>> Lars >>> >> >> >> >> -- >> Takayoshi Kochi >> > > -- Takayoshi Kochi
Received on Tuesday, 1 August 2017 03:36:27 UTC