W3C home > Mailing lists > Public > whatwg@whatwg.org > August 2010

[whatwg] input type=ink proposal

From: James May <whatwg@fowlsmurf.net>
Date: Mon, 9 Aug 2010 21:55:32 +1000
Message-ID: <AANLkTinhkmaByHhN5ouoB8xta+-eFROk_RxYP_+Ea5c4@mail.gmail.com>
Couldn't "Ink" be implemented in client script with a combination of
canvas, multitouch events and XHR anyway?

Unless I'm missing something...?

-- James

On 9 August 2010 18:27, Mikko Rantalainen <mikko.rantalainen at peda.net> wrote:
> 2010-06-08 19:26 EEST: Charles Pritchard:
>> Has there been prior discussion about an input type="ink" form element?
>>
>> This element would simply capture coordinates from a mouse (or touch/pen
>> input device),
>> allowing the user to reset the value of the element at their discretion.
>>
>> InkML is in last call:
>> http://www.w3.org/TR/InkML/
>>
>> Use Case:
>>
>> As part of a web form, a user signs their digital signature to confirm
>> acceptance of terms.
>
> Uugh.. When I see "digital signature" or "digitally signed", I expect
> public key decryption of cryptographic hash of the signed content, not a
> digitized image of one's signature. The former cannot be easily forged
> as opposed to the latter, which can be copied at will after receiving
> the first copy of digitized image for any purpose.
>
> In my opinion, WHATWG or W3C shouldn't provide any tools to make it any
> easier to create fake sense of security through digitized images.
>
> If you want digital signatures, do it correctly or don't do it at all!
>
>> Use Case:
>>
>> While filling out an online profile, a user submits a simple doodle as
>> their avatar.
>
> Has there ever been anybody asking for such a feature? If browsers
> supported capturing an image from a web camera, it would be much more
> sensible feature for this use case.
>
>> Use Case:
>>
>> To quickly log into an online system, a user scribbles a password,
>> which their server tests for fidelity to their prior scribbled input.
>
> I believe this is borderline sensible use case, but I think this should
> be interal user agent feature: use normal passwords for the login but
> use scribbled input for opening password keyring in the user agent (or
> operating system).
>
> In short, I don't agree that input type="ink" should be supported.
>
> --
> Mikko
>
>
Received on Monday, 9 August 2010 04:55:32 UTC

This archive was generated by hypermail 2.3.1 : Monday, 13 April 2015 23:08:59 UTC