08-mediafrag-minutes.html 12 KB
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<html lang='en'>
<head>
  <title>Media Fragments Working Group Teleconference -- 08 Sep 2010</title>
  <link type="text/css" rel="STYLESHEET" href="http://www.w3.org/StyleSheets/base.css">
  <link type="text/css" rel="STYLESHEET" href="http://www.w3.org/StyleSheets/public.css">
  <link type="text/css" rel="STYLESHEET" href="http://www.w3.org/2004/02/minutes-style.css">
  <meta content="Media Fragments Working Group Teleconference" name="Title">  
  <meta content="text/html; charset=iso-8859-1" http-equiv="Content-Type">
</head>

<body>
<p><a href="http://www.w3.org/"><img src="http://www.w3.org/Icons/w3c_home" alt="W3C" border="0"
height="48" width="72"></a> 

</p>

<h1>Media Fragments Working Group Teleconference</h1>
<h2>08 Sep 2010</h2>

<p><a href='http://lists.w3.org/Archives/Public/public-media-fragment/2010Sep/0001.html'>Agenda</a></p>


<p>See also: <a href="http://www.w3.org/2010/09/08-mediafrag-irc">IRC log</a></p>

<h2><a name="attendees">Attendees</a></h2>

<div class="intro">
<dl>
<dt>Present</dt>
<dd>Yves, Raphael, Erik, Davy, Silvia</dd>
<dt>Regrets</dt>
<dd>Jack</dd>
<dt>Chair</dt>
<dd>Erik </dd>
<dt>Scribe</dt>
<dd>Silvia</dd>
</dl>
</div>

<h2>Contents</h2>
<ul>
  <li><a href="#agenda">Topics</a>
	<ol>
	<li><a href="#item01">ADMIN</a></li>
<li><a href="#item02">SPECIFICATION</a></li>
<li><a href="#item03">PUBLIC COMMENTS</a></li>
<li><a href="#item04">Media Fragments in RTSP</a></li>
<li><a href="#item05">TEST CASES</a></li>

	</ol>
  </li>
  <li><a href="#ActionSummary">Summary of Action Items</a></li>
</ul>
<hr>
<div class="meeting">

<p class='phone'>


</p>

<p class='phone'>

<p class='irc'>
&lt;<cite>trackbot</cite>&gt; Date: 08 September 2010
</p>

</p>

<p class='irc'>
&lt;<cite>erik</cite>&gt; scribe: Silvia
</p>

<p class='irc'>
&lt;<cite>erik</cite>&gt; ScribeNick: Silvia
</p>

<h3 id="item01">ADMIN</h3>
<p class='irc'>
&lt;<cite>erik</cite>&gt; Date: 08 September 2010
</p>

<p class='phone'>
PROPOSED to accept the minutes of the 07 July 2010 telecon:
</p>

<p class='phone'>
<a href="http://www.w3.org/2010/07/07-mediafrag-minutes.html">http://www.w3.org/2010/07/07-mediafrag-minutes.html</a>
</p>

<p class='irc'>
&lt;<cite>erik</cite>&gt; +1
</p>

<p class='irc'>
&lt;<cite>davy</cite>&gt; +1
</p>

<p class='phone'>
+1
</p>

<p class='phone'>
accepted
</p>

<p class='phone'>
LC Period: extension?
</p>

<p class='phone'>
<cite>erik:</cite> should be extended till the end of the month <br>
... we have to close everything for TPAC
</p>

<p class='phone'>
<cite>Yves:</cite> extending till the end of the month beyond TPAC is good to get feedback
</p>

<a name="action01"></a>
<p class='irc'>
&lt;<cite>scribe</cite>&gt; <strong>ACTION:</strong> troncy to send reminders to all relevant groups [recorded in <a href="http://www.w3.org/2010/09/08-mediafrag-minutes.html#action01">http://www.w3.org/2010/09/08-mediafrag-minutes.html#action01</a>]
</p>

<p class='irc'>
&lt;<cite>trackbot</cite>&gt; Created ACTION-183 - Send reminders to all relevant groups [on Raphaël Troncy - due 2010-09-15].
</p>

<p class='phone'>
<cite>erik:</cite> there will be a meeting at TPAC <br>
... Davy, Raphael and Yves will be there - Yves can only make it on Tuesday
</p>

<p class='irc'>
&lt;<cite>Yves</cite>&gt; <a href="http://www.w3.org/2002/09/wbs/35125/TPAC2010reg/">http://www.w3.org/2002/09/wbs/35125/TPAC2010reg/</a>
</p>

<h3 id="item02">SPECIFICATION</h3>
<p class='phone'>
Discussion (Philip+Yves): media fragment syntax grammar (why segment production rule is not normative?), extensibility, pseudo algorithm code
</p>

<p class='phone'>
<cite>scribe:</cite> has not yet happened
</p>

<p class='phone'>
close ACTION-175
</p>

<p class='irc'>
&lt;<cite>trackbot</cite>&gt; ACTION-175 Update the grammar parsing automatically generated code for the URI syntax closed
</p>

<p class='phone'>
<cite>Yves:</cite> we need to discuss extension points <br>
... that will trigger what we define as valid media fragments <br>
... will have to wait until next week when we can talk with Philip
</p>

<p class='irc'>
&lt;<cite>Yves</cite>&gt; related also to <a href="http://lists.w3.org/Archives/Public/public-media-fragment/2010Sep/0000.html">http://lists.w3.org/Archives/Public/public-media-fragment/2010Sep/0000.html</a>
</p>

<p class='phone'>
<cite>erik:</cite> we'll discuss it next week
</p>

<h3 id="item03">PUBLIC COMMENTS</h3>
<p class='phone'>
Image sprite in HTML5: <a href="http://lists.w3.org/Archives/Public/public-media-fragment/2010May/0061.html">http://lists.w3.org/Archives/Public/public-media-fragment/2010May/0061.html</a>
</p>

<p class='phone'>
<cite>erik:</cite> this is related to a discussion on the WHATWG
</p>

<p class='phone'>
<cite>Yves:</cite> this is about what to do in the UI with spatial media fragments
</p>

<p class='phone'>
<cite>erik:</cite> the point made is that cropping is requested as the UI effect of media fragments
</p>

<p class='irc'>
&lt;<cite>raphael</cite>&gt; please, consider the use case bring by Europeana and sent yesterday
</p>

<p class='phone'>
<cite>Yves:</cite> we may need to support both, highlighting and cropping, and need to differentiate between them
</p>

<p class='phone'>
<cite>davy:</cite> I think it's application-dependent <br>
... I think we should leave that open
</p>

<p class='irc'>
&lt;<cite>raphael</cite>&gt; ... they are very keen to implement the spec and they really would like to have a keyword in the URI that can refer to a description of any arbitrary spatial shape
</p>

<p class='phone'>
<cite>Yves:</cite> in that case we are not controlling what is presented to the user
</p>

<p class='irc'>
&lt;<cite>raphael</cite>&gt; ... the usage would not be retrieval as for the current spatial media fragment, but address and highlight on client a particular shape
</p>

<p class='phone'>
<cite>silvia:</cite> the WHATWG doesn't want to specify the presentation of media fragment URIs in the HTML spec, but wants it to be in our spec <br>
... that's where it belongs in their mind
</p>

<p class='irc'>
&lt;<cite>raphael</cite>&gt; ... what Europeana wants is a way to address any spatial region, with a media fragment URI
</p>

<p class='phone'>
<cite>silvia:</cite> yes, what raphael says - we may need to introduce another means of specifying arbitrary shapes which can then only be used for highlighting
</p>

<p class='phone'>
<a href="http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-July/026937.html">http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-July/026937.html</a>
</p>

<p class='phone'>
<cite>Davy:</cite> they should just use the query parameter
</p>

<p class='phone'>
<cite>silvia:</cite> no, they don't want the cropping to happen on the server, they just want the effect to be specified
</p>

<p class='phone'>
<a href="http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-August/027581.html">http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-August/027581.html</a> &lt;- the key thread
</p>

<p class='phone'>
<a href="http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-August/028163.html">http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-August/028163.html</a>
</p>

<p class='phone'>
<cite>davy:</cite> do you agree with Ian?
</p>

<p class='phone'>
<cite>silvia:</cite> I didn't but I think he has a point <br>
... maybe we have to specify it and it may even apply to all applications <br>
... or we may need to put an extra mechanism in to specify if we want cropping or highlighting
</p>

<p class='phone'>
<cite>davy:</cite> or we can just specify it as cropping for the browser
</p>

<p class='phone'>
<cite>Yves:</cite> it depends on the application and we need to explain that better
</p>

<p class='phone'>
<cite>silvia:</cite> but the problem is that the researchers in Europe also want to use it in the browser, so it's still the same application
</p>

<p class='irc'>
&lt;<cite>Yves</cite>&gt; effect=highlight (with a default of cropping)
</p>

<p class='phone'>
<cite>silvia:</cite> maybe we should create a new scheme for arbitrary selections and they would just be for highlighting
</p>

<p class='phone'>
<cite>Yves:</cite> I was thinking of adding a parameter
</p>

<p class='phone'>
<cite>Davy:</cite> how does that work with multiple dimensions
</p>

<p class='phone'>
<cite>silvia:</cite> we could do that in the temporal domain, too
</p>

<p class='phone'>
<cite>Yves:</cite> I want Jack's input on this
</p>

<p class='phone'>
<cite>silvia:</cite> this discussion is different to what we looked at before: it's only about what is being presented in the application <br>
... we could do something like add an extension to the region, e.g. xywh=x,y,w,h_crop or xywh=x,y,w,h_highlight
</p>

<p class='phone'>
<cite>erik:</cite> we need to discuss in a larger group <br>
... this is only about how to render something, nothing to do with what goes over the wire <br>
... this in fact relates to all three of the comments we received
</p>

<p class='phone'>
Robert Sanderson: <a href="http://lists.w3.org/Archives/Public/public-media-fragment/2010Jul/0010.html">http://lists.w3.org/Archives/Public/public-media-fragment/2010Jul/0010.html</a> &lt;- also
</p>

<p class='phone'>
<cite>scribe:</cite> but the Chris Double comment is different
</p>

<p class='phone'>
Chris Double: <a href="http://lists.w3.org/Archives/Public/public-media-fragment/2010Jun/0099.html">http://lists.w3.org/Archives/Public/public-media-fragment/2010Jun/0099.html</a>
</p>

<p class='phone'>
<cite>Yves:</cite> answer to Chris is: it should always send the header, because in the worst case it will send the whole resource
</p>

<a name="action02"></a>
<p class='irc'>
&lt;<cite>scribe</cite>&gt; <strong>ACTION:</strong> Yves to email Chris the reply on when to send the range header [recorded in <a href="http://www.w3.org/2010/09/08-mediafrag-minutes.html#action02">http://www.w3.org/2010/09/08-mediafrag-minutes.html#action02</a>]
</p>

<p class='irc'>
&lt;<cite>trackbot</cite>&gt; Created ACTION-184 - Email Chris the reply on when to send the range header [on Yves Lafon - due 2010-09-15].
</p>

<p class='phone'>
<cite>Yves:</cite> Robert wants to be able to extend stuff <br>
... the main issue is that you have no control of what you see on the other end, since the reference URI changes all the time <br>
... this may be an issue <br>
... apart from this, I think it's ok
</p>

<p class='phone'>
<cite>silvia:</cite> I think adding extra name-value parameters to the URI is ok - that's how we built it <br>
... but they want to change the spatial fragment spec, so that's not ok <br>
... they should just add another name-value parameter
</p>

<p class='phone'>
<cite>erik:</cite> seems we need to discuss this next week
</p>

<h3 id="item04">Media Fragments in RTSP</h3>
<p class='phone'>
close ACTION-182
</p>

<p class='irc'>
&lt;<cite>trackbot</cite>&gt; ACTION-182 Include summary into wiki page closed
</p>

<p class='phone'>
<cite>erik:</cite> philip cannot make it next week - let's have the grammar discussion in two weeks <br>
... next week we will focus on the public comments
</p>

<h3 id="item05">TEST CASES</h3>
<p class='phone'>
nobody here for it
</p>

<p class='phone'>
<cite>erik:</cite> let's close it here
</p>


</div>
<h2><a name="ActionSummary">Summary of Action Items</a></h2>
<!-- Action Items -->
<strong>[NEW]</strong> <strong>ACTION:</strong> troncy to send reminders to all relevant groups [recorded in <a href="http://www.w3.org/2010/09/08-mediafrag-minutes.html#action01">http://www.w3.org/2010/09/08-mediafrag-minutes.html#action01</a>] <br />
<strong>[NEW]</strong> <strong>ACTION:</strong> Yves to email Chris the reply on when to send the range header [recorded in <a href="http://www.w3.org/2010/09/08-mediafrag-minutes.html#action02">http://www.w3.org/2010/09/08-mediafrag-minutes.html#action02</a>] <br />
&nbsp;<br />


[End of minutes] <br>
<hr>

<address>
  Minutes formatted by David Booth's 
  <a href="http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm">scribe.perl</a> version 1.135 (<a href="http://dev.w3.org/cvsweb/2002/scribe/">CVS log</a>)<br>
  $Date: 2010/09/08 12:14:37 $ 
</address>
</body>
</html>