W3C home > Mailing lists > Public > public-webevents@w3.org > January to March 2011

Re: First Draft, Touch Interface Spec

From: timeless <timeless@gmail.com>
Date: Tue, 25 Jan 2011 17:24:55 +0200
Message-ID: <AANLkTikMUA4E068JXFUnZt7T8RR2ZjaVkCrprcftpzPe@mail.gmail.com>
To: "public-webevents@w3.org" <public-webevents@w3.org>
interface TouchPoint {
    readonly attribute integer     identifier;
identifier of type integer, readonly
    an identification number unique to each touch point

this seems problematic :) it might take a while but i bet someone
could overflow it.

also, how does integer relate to long used elsewhere? (i could read
webidl spec, but i'm asking more about why you chose one over the
other).

interface TouchList {
    readonly attribute unsigned long length;
    caller getter TouchPoint item (in unsigned long index);
    caller getter object     identifiedPoint (in integer identifier);

i'd almost rather you used id for both this and the previous thing.

2.1 Attributes
identifiedPoint
    returns the first touchpoint with integer identifier from the list

'integer' do you mean 'index'?

interface TouchEvent : Event {
    readonly attribute TouchList touches;

i'm not a fan of 'touches' :)

    readonly attribute TouchList targetTouches;
    readonly attribute TouchList changedTouches;

> 3.2 The touchstart event
> A user agent must dispatch this event type when the user places a touch point on the touch surface.

i don't like this must.

A user agent shall dispatch this event type to indicate when the user
places a touch point on the touch surface.

note that my complaint is really that if i the user want to talk to
the os, a user agent, or an extension, i don't want my useragent to be
obligated to send an event to the web application. the fix isn't
swapping must for shall (they mean the same thing), it's the insertion
of 'to indicate' which should relax the requirement to enable the
useragent to say 'the user didn't want me to indicate this to the
webapp, so i didn't dispatch the event to it'.

the same comment applies to the others:
> A user agent must dispatch this event type when the user removes a touch point from the touch surface.
> A user agent must dispatch this event type when the user moves a touch point along the touch surface.
> A user agent must dispatch this event type when a touch point moves onto the interactive area defined by a DOM element on which it did not originate. Events of this type must not bubble.
> A user agent must dispatch this event type when a touch point moves off the interactive area defined by a DOM element. Events of this type must not bubble.
> A user agent must dispatch this event type when the user places a more touch points on the touch surface than the device or implementation is configured to store, in which case the earliest touch point in the list must be removed.

Again, a useragent needs to be able to have a legal opt out for the
case where the user agent is being nice to the user.

if i have hand trouble (i'm getting this), i need a conforming
useragent to be able to not send out events when i don't really want
to send out events, but am merely resting my pained hands.
Received on Tuesday, 25 January 2011 15:25:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 25 January 2011 15:26:24 GMT