W3C home > Mailing lists > Public > public-canvas-api@w3.org > October to December 2014

Re: hit region list

From: Richard Schwerdtfeger <schwer@us.ibm.com>
Date: Mon, 6 Oct 2014 13:47:53 -0500
To: Rik Cabanier <cabanier@gmail.com>
Cc: Jatinder Mann <jmann@microsoft.com>, Canvas <public-canvas-api@w3.org>
Message-ID: <OF387304FE.6E2C54B2-ON86257D69.0066F72A-86257D69.006742C4@us.ibm.com>

OK. We have tests now for manual tests. They may require some cleanup but I
think we are covered. We should discuss text for "notifying the user" to
refine it for the spec. in an earlier email this weekend.

So, if everyone is happy with that the bug can be closed.

Mark, you are good to go on writing the test.

Rich


Rich Schwerdtfeger



From:	Rik Cabanier <cabanier@gmail.com>
To:	Richard Schwerdtfeger/Austin/IBM@IBMUS
Cc:	Jatinder Mann <jmann@microsoft.com>, Canvas
            <public-canvas-api@w3.org>
Date:	10/06/2014 01:37 PM
Subject:	Re: hit region list





On Mon, Oct 6, 2014 at 9:26 AM, Richard Schwerdtfeger <schwer@us.ibm.com>
wrote:
  The main reason for this bug filing was testing.

  The browser must be able to handle the scrolling situation - per
  Dominic's comment. So, here we are in complete agreement but we need to
  test for it. I have 2 test cases for this. I should note that just
  because you say the browser should handle the scrolling it does not mean
  that it will with respect to the a11y api mappings. We need to test for
  this. Whenever the screen location of the hit region changes the
  accessibility api MUST be updated or it magnifiers will NOT work.


Yes. This should all happen automatically. Canvas hit regions are no
different from other accessible objects.
  So, the ONLY way we can test this is through the use of manual tests. If
  you look at the test cases I put out they are manual tests. So, if the
  group is ok with manual testing then we don't need an API to get the hit
  region list.


I believe manual tests are fine



  Inactive hide details for Rik Cabanier ---10/06/2014 11:05:56 AM---On
  Mon, Oct 6, 2014 at 7:32 AM, Richard Schwerdtfeger <schweRik Cabanier
  ---10/06/2014 11:05:56 AM---On Mon, Oct 6, 2014 at 7:32 AM, Richard
  Schwerdtfeger <schwer@us.ibm.com> wrote:

  From: Rik Cabanier <cabanier@gmail.com>
  To: Richard Schwerdtfeger/Austin/IBM@IBMUS
  Cc: Jatinder Mann <jmann@microsoft.com>, Canvas <public-canvas-api@w3.org
  >
  Date: 10/06/2014 11:05 AM
  Subject: Re: hit region list





  On Mon, Oct 6, 2014 at 7:32 AM, Richard Schwerdtfeger <schwer@us.ibm.com>
  wrote:
        Rik.

        Hit testing is managed by the user agent. When you do things like
        clearRect() it is supposed to delete some of the hit regions. What
        you are suggesting is comparable to developers trying to keep track
        of focus changes. I don't agree with your assertion.

        So, once you set a hit region it is given to the browser and it has
        its own copy of it that it operates off of. What are the odds of
        staying in synch with that over time?

  As Dominic points out, the browser will manage that for you. You can even
  "tear off" the canvas element from the DOM, insert it in a different
  location and it should keep on working. If the only reason for this API
  is testing, we shouldn't add it.

  The a11y infrastructure does have access to the list of hit regions but
  I'm unsure how you can use that for tests.

        Inactive hide details for Rik Cabanier ---10/05/2014 09:12:01
        PM---On Sun, Oct 5, 2014 at 5:14 AM, Richard Schwerdtfeger <schwe
        Rik Cabanier ---10/05/2014 09:12:01 PM---On Sun, Oct 5, 2014 at
        5:14 AM, Richard Schwerdtfeger <schwer@us.ibm.com> wrote:

        From: Rik Cabanier <cabanier@gmail.com>
        To: Richard Schwerdtfeger/Austin/IBM@IBMUS
        Cc: Canvas <public-canvas-api@w3.org>, Jatinder Mann <
        jmann@microsoft.com>
        Date: 10/05/2014 09:12 PM
        Subject: Re: hit region list






        On Sun, Oct 5, 2014 at 5:14 AM, Richard Schwerdtfeger <
        schwer@us.ibm.com> wrote:
              Jatinder, Rik,

              I am working on the testable statements for hit regions.
              There is no programmatic way to determine the hit region
              list. So, when we want to verify that a hit region having the
              same id is replaced in the list we cannot determine if the
              old hit region, associated with the id, is removed from the
              list. A developer will want to be able to look at the hit
              region list.

              I filed a defect on this:
              https://www.w3.org/Bugs/Public/show_bug.cgi?id=26977

              This will need to be a canvas 2.0 item and it does restrict
              our ability to test all of canvas 1.0 hit region
              functionality.

        I don't see why this needs to be a feature in canvas. If a
        developer wants to keep track of the hit regions, he can do that
        himself.










graycol.gif
(image/gif attachment: graycol.gif)

Received on Monday, 6 October 2014 18:48:27 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:31:57 UTC