W3C home > Mailing lists > Public > public-media-fragment@w3.org > April 2009

Review of section 8

From: Silvia Pfeiffer <silviapfeiffer1@gmail.com>
Date: Sun, 12 Apr 2009 21:53:24 +1000
Message-ID: <2c0e02830904120453x1db31dd3l6d5ae597a8d7da6f@mail.gmail.com>
To: Media Fragment <public-media-fragment@w3.org>
Hi all,

Here goes review of section 8....


Section 8.1

* (minor text change:) first paragraph: replace "There exists already"
with "There exist already"

* Can we structure the list given in section 8.1 a bit more nicely? I
would suggest:
  1. General specification of URI fragments
      -> URI fragment RFC 3986
  2. Fragment specifications not for audio or video
      -> HTML named anchors (I would refer to HTML4.01
http://www.w3.org/TR/REC-html40/)
http://www.w3.org/TR/REC-html40/struct/links.html#h-12.1
      -> XPointer named elements
      -> text/plain RFC 5147
      -> SVG
  3. Fragment specifications for audio/video
      -> temporal URI
      -> MPEG-21
      -> should MPEG-7 be added?

* in the HTML part, it should not be refers "to a specific location
within the resource" but "to a specific named anchor within the
resource"

Section 8.2

* if we cannot provide a syntax example for Metacafe, Dailymotion, or
Microsoft IIS, we should leave them out, because they do not
contribute anything

Section 8.3

* "Temporal URI" covers the temporal addressing - track comes through
ROE and named comes through CMML; it would be better to use "Temporal
URI/Ogg technologies" for this
* it should be "Non-URI-based" rather than "None URI-based"
* we should remove "HTML5" from this list, since it has removed all
its start and end attributes in favor of going with media fragment
URIs that we define - so, HTML5 will do whatever we come up with

Section 8.3.1.2

* call it "Temporal URI/Ogg technologies"

Section 8.3.1.2.1

* I do not understand the separation between "out-of-context" and
"in-context"; instead, I would prefer that the second half of this
section be moved to "8.3.1.2.3. Named" because CMML does indeed enable
named addressing of fragments; that it also provides annotations for
Ogg files and outgoing hyperlinks is not of importance here
* last paragraph should start with "Note how there is.." rather than
"Not how there is.."
* should get a reference to temporal URI spec,
http://annodex.net/TR/draft-pfeiffer-temporal-fragments-03.html

Section 8.3.1.2.2

* the examples run over my right page border - I think they need to be
moved from <eg> to <example> - like a lot more such sections further
down in section 8
* should get a reference to ROE spec, http://wiki.xiph.org/index.php/ROE
* last sentence: replace "(at least virtually)" with "(at least in theory)"  :-)

Section 8.3.1.2.3

* the first example is not an example for a named URI, but one for a
combined time and track URI
* move the second half from 8.3.1.2.1 here - I think it makes a lot
more sense here to have CMML defined
* should get a reference to CMML spec,
http://annodex.net/TR/draft-pfeiffer-cmml-03.html

Section 8.3.1.3 MPEG-21

* can we have a reference to a Web page that explains this stuff here?
* what are the "iloc" and "tkhd" boxes?
* in the "ffp()" paragraph, remove "for identifying tracks" (already
stated earlier)
* in the "offset()" paragraph remove "Here we locate 100 bytes
starting at the 11th byte within myfile.mp4" - it seems to refer to an
example that is not present in this document
* in the "mp()" paragraph, somethings seems to be missing after the
colon. Or maybe the next paragraph starting with "Note that
hierarchical.." is in the wrong place and should be moved after the
two dot-points
* in the "mask()" paragraph, I do not understand what is meant by a
binary (1-bit deep) mask - can you explain? This also applies to
section 8.3.1.3.4 which I do not understand either (and the sentence
in the <eg> example is cut off).

Section 8.3.2.1 SMIL

* this section has a lot of ~jack examples that should be replaced
with the correct w3c url
* this section further has a lot of <eg> examples that print over the
right page border
* 8.3.2.1.1 last paragraph has a spare "do" in "It is also possible to
(do) have..."
* 8.3.2.1.1 needs a URI example at the end, which explains that you
can link to http://example.com/fragf2f.mp4#tic2area
* 8.3.2.1.2 fourth paragraph claims that the coords attribute is
compatible with HTML - might be worth stating CSS as well
* 8.3.2.1.3 last sentence should read "Yet, no such implementations
are known." to make the sentence neutral

Section 8.3.2.2 MPEG-7

* 8.3.2.2.1 first paragraph third sentence should read "The
MediaTimePoint (has) is (build up) defined as follows ..."
* next sentence should read "The MediaDuration is (build up) defined
as follows..."
* 8.3.2.2.1 should add at the end what the example actually describes.
Something like: "This MPEG-7 example describes a 'shot1' starting at
6sec 2002/2500 sec and lasts for 9sec 13389/25000 sec." (if I am
correct...)
* 8.3.2.2.2 should add after the example what it actually describes.
Something like: "This MPEG-7 example describes a spatial mask called
"speaker" which is given by the coordinates (40, 300), (105,210),
(???), (320,240)." (and can somebody please clarify what that middle
bit is?)
* 8.3.2.2.2 should also clarify what the <box> definition is - what do
those numbers actually mean?
* 8.3.2.2.3 should add how to address such a thing - would it be
http://example.com/fragf2f.mp7#id=speaker ?

Section 8.3.2.3 SVG

* I think the whole temporal section should merely state that SVG
relies either on SMIL or HTML5 as a foreign object to introduce
temporal media fragmentation. I don't think it has ay temporal
fragmentation of its own.
* 8.3.2.3.1 last word in first paragraph should be "either"

Section 8.3.2.4 TimedText

* this is very bare and does not show any hyperlinking at all; unless
we have an idea how hyperlinking works in TimedText we should either
drop this or make an editorial note that we require clarification on
this

Section 8.3.2.5 HTML5

* as stated above, this should be removed; the current example is
temporal URI, so not HTML5, and the Date and Time descriptions given
there are not relevant.
* it might be possible to write something like this:
  "Currently, HTML5 relies on the abilities of the used media format
for providing media fragment addressing. Here is an example using Ogg:
(enter Ogg examples). In future, HTML5 is planning to adopt the
fragment URI specifications of this document for providing fragment
addressing. Input from the WHAT and HTML working groups is requested."

Section 8.3.2.6 TV-Anytime

* 8.3.2.6.1 In contrast to all other section here, the TV-Anytime
section does not include examples of how to use a
temporal/spatial/named document example and URI address it. Instead,
this section contains extracts from the specification of the
TV-Anytime language. This section is basically unreadable and unusable
because of that. I found it hard to wade my way through from
tva:TimeBaseReference, tva:TVAMediaTimeType, and
tva:fragmentIdentification to the specification of
TimeBaseReferenceType through to mpeg7:mediaTimePointType and
mpeg7:MediaRelIncrTimePointType and their specifications. Can we
please rather have a TV-Anytime example document with some actual time
points in them? Or can we just ditch this whole paragraph an say that
TV-Anytime uses the MPEG-7 specification to define media fragments?

* 8.3.2.6.2 is missing an example

Section 8.3.2.7.1 Image Maps

* The "Discussion" part does not really fit the rest of this section.
What is the intention with this list? Does anything in this list have
an influence on what we will specify for spatial fragment URIs? If
not, it should be ditched, IMHO.


OK, enough for today. :) Tomorrow I will work on fixing my sections!!
Happy Easter!!

Cheers,
Silvia.
Received on Sunday, 12 April 2009 11:54:16 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 21 September 2011 12:13:32 GMT