W3C home > Mailing lists > Public > w3c-wai-ua@w3.org > July to September 2000

Re: Checkpoint 4.6: question about "rewind"

From: Al Gilman <asgilman@iamdigex.net>
Date: Thu, 31 Aug 2000 11:18:04 -0400
Message-Id: <200008311507.LAA273453@smtp1.mail.iamworld.net>
To: w3c-wai-ua@w3.org
[Note that on the substance of the proposal, Harvey and Ian are in
agreement.  I support that agreement.  This blast is just an attempt to
eliminate the appearance of any opportunity for disagreement by expanding a
little on the media vs. HCI theoretical view of these controls.  - Al]

At 01:50 AM 2000-08-31 -0400, Ian Jacobs wrote:
>Harvey Bingham wrote:
>> At 2000-08-29 13:58-0400,Ian Jacobs <ij@w3.org> wrote: <snip>
>>  > I propose that "rewind to the beginning" be changed
>>  > back to "rewind", since I believe the intent of the WG
>>  > was to have VCR-type controls.
>> HB: The VCR-type controls imply a linear medium. Stopping before finished
>> works,
>> but it is a nuisance to accurately position. I do not believe the intent
of the
>> WG was to only consider linear medium.
>What other media do you think were meant?


Hold on a sec.

Harvey, the requirements to support certain minimum linear-metaphor control
functions does not limit the medium overall to be linear.  The digital
talking book has that requirement.  You have to be able to play a whole
book with one button press.  But not every resource will be that way.  

On the other hand, every multimedia presentation boils down to voxels in at
most four dimensions as it is being presented.  Only one of these is time
and time always has an understood forward and backward direction to it.

The time model of SMIL always resolves to a single time axis against which
all the components are lined up as they cross the human:computer interface.

So inside any bubble of content that has a continuous-play mode of
presentation, there is a time axis and the controls manipulating the
location on this axis relate to a linear space of virtual times.

The linear metaphor embedded in these controls is in no way a restriction
on the multimedia content that is not imposed by physics and human
perception.  The requirement to have controls that relate to this
smooth-topology linear model of virtual time is just part of the whole
"direct manipulation" strategy -- providing controls in a model as close to
the way the user perceives the experience as possible.

So, yes; we explain how the controls work by analogy to an audio tape
player or a VCR if we must.  But they don't impose a restriction on the
multimedia domain; the capitalize on a feature of it.

Nothing in the control requirements says the physical medium is
serial-access or the content is streaming from the source.

It just talks about the virtual world in which the controls act.


By the same token, the medium of "structured audio" the way George Kerscher
talks about it is indeed richer than "one linear program."  But this is
more like there is a database that generates lots of query answers or
views, each of which is a continuous-play program.  The linear-time,
continuous-play bubble is the multimedia equivalent of the 'page' that we
know in the hyper-TEXT web.  You can at any time activate controls that
change the premises and construction of the continuous-play stream.  But in
the context of a continuous-play stream, there is only one time axis and it
is smooth and linear.  And the user needs, where the information is not
intrinsically and necessarily real-time in nature, to be able to say things
like "Hold on, there.  Back up a little, like so..."


> - Ian
>Ian Jacobs (jacobs@w3.org)   http://www.w3.org/People/Jacobs
>Tel:                         +1 831 457-2842
>Cell:                        +1 917 450-8783
Received on Thursday, 31 August 2000 11:03:46 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:49:27 UTC