See also: IRC log
<fsasaki> see admin page http://www.w3.org/2008/WebVideo/Annotations/admin
<dsinger> by the way, Eric Carlson is also here on the same line and IRC with me
introduction eric carlson (Apple)
<fsasaki> close ACTION-20
<trackbot> ACTION-20 Add a link from WG page to the "still alive" MMSEM wiki pages and to make sure that everybody can edit the pages closed
<frank> hello everybody, Veronique and I call in now (we share a phone)
<fsasaki> close ACTION-21
<trackbot> ACTION-21 Create a liaison to Geolocation WG closed
<fsasaki> ACTION-22 Go back to Karen and check about " IPTV metadata specification" - what is it, is it available for us? - pending
<fsasaki> close ACTION-23
<trackbot> ACTION-23 Explain XMLSPEC to wonsuk closed
<fsasaki> close ACTION-24
<trackbot> ACTION-24 Get CVS accounts for veronique and wonsuk closed
<fsasaki> close ACTION-25
<trackbot> ACTION-25 Review XMP basic schema closed
<fsasaki> close ACTION-26
<trackbot> ACTION-26 Review XMP Dublin Core schema closed
<fsasaki> close ACTION-27
<trackbot> ACTION-27 Review XMP Rights Management schema closed
<fsasaki> close ACTION-28
<trackbot> ACTION-28 Review XMP Media Management schema closed
<fsasaki> close ACTION-29
<trackbot> ACTION-29 Tell Wonsuk to review XMP Basic Job Ticket schema and Paged-Text schema closed
<fsasaki> close ACTION-30
<trackbot> ACTION-30 Review XMP Dynamic Media schema closed
<fsasaki> close ACTION-31
<trackbot> ACTION-31 Contact tom baker about dc liaison closed
<fsasaki> action-32 Make a liaison with MPEG using information from the XP homepage - pending
<fsasaki> action-33 Evaluate if contact to IPTV Japan is valuable - pending
<fsasaki> action-34 Contact Open IPTV forum - pending
<fsasaki> close ACTION-35
<trackbot> ACTION-35 Check contacts to OMA closed
<fsasaki> ACTION: Felix to make a liaison with OMA [recorded in http://www.w3.org/2008/11/11-mediaann-minutes.html#action01]
<trackbot> Created ACTION-36 - Make a liaison with OMA [on Felix Sasaki - due 2008-11-18].
ruben: jpeg in process of defining metadata for digital image
... specific type of search/retrieval uc
felix: keep use case and discuss in future call
<fsasaki> or via mail
<cgi-irc> maybe it can be an interesting use case for checking how the interoperability across media can be achieved
<fsasaki> veronique = cgi-irc
<pchampin> +1 veronique
veronique: consider in second stage of ontology dvelopment
thierry: dublin core: everything could be adopted
felix: some fields (eg data) quite general, might conflict with fields from other standards
werner: +1, I think for most dc fields we need to specify how we use them
ruben: granularity of fields (eg name, more structured eg in mpeg-7)
david: string is general, everything else might cause eg internationalisation problems
<dsinger> it may be solved, but it can be a tarpit
goal is to keep simple
veronique: pointing to vocbulary (eg artist names)
<pchampin> wouldn't that be interesting, nevertheless, to allow the URI of another resource (possibly defined elswhere), insread of a plain string?
felix: need to decide whether internationalistation of names is an issue or we just map to simple strings
<dsinger> it's intriguing to think that anything can be a value or a URI
<dsinger> but what happens if I say "my name is" "URI to jpeg image"
<pchampin> I was rather suggesting "the creator is (string|URI)"
david: uri issues brings us back to foaf
<pchampin> I'm not suggesting that we should embed the descrition of that URI...
<dsinger> friend of a friend?
<dsinger> you mean? <http://xmlns.com/foaf/0.1/>
<frank> A string is not less ambigious than a URI (I could write name = sjdhfqflf
<pchampin> I would say a string is *much more* ambiguous
joakim: foaf is used in imdb
<pchampin> But we can not assume every person to have a URI with appropriate FOAF properties
<dsinger> in terms of presentation, it's very clear; a URI ("gopher:" ?) is much less obviously presentable
<js> foaf is used by linked movie database
felix: string may be meaningful or not, additional uri (need to ensure that uri points to suitable resource)
<rtous> also a problem with "address" field
felix: going through elements
dave: even simple copyright can become complex (eg composition of
... simple copyright cannot be ignored
felix: does that require to include properties into ontology
dave: simple copyright, global to media
<frank> I thought we had decided in the last f2f meeting that we simply go for one slot
joakim: dynamic temporal metadata support (for real time apps)
<dsinger> I am tempted to say that ALL metadata is temporal, but the first version only allows you to ask "right now" what is...?
<dsinger> which covers a radio station where it varies by song...
veronique: 3 levels: content - realisation -instance
... still to be decided whether 3 or 1 level
<dsinger> yes, use cases are essential: we'll never decide what we want if we don't know what we want it to be useful for
<rtous> Functional Requirements for Bibliographic Records (FRBR) : http://en.wikipedia.org/wiki/FRBR
ruben: different levels could simplify things (e.g. content description such as title vs tech metadata like encoding)
<cgi-irc> we can also check out how many models go for 3 levels and how many go for 1
<cgi-irc> VRA has also 3 levels
<js> mpeg-7 has two
<pchampin> I'm still not sure there should be a *constant* level of models...
<cgi-irc> two is good enough
<dsinger> what are the 3 levels? (sorry...)
content - realisation -instance
<pchampin> what about the photo of a painting?
felix: volunteers for looking at how standards handle this issue?
<pchampin> or a movie making of?
<rtous> I can do that
<scribe> ACTION: ruben to survey how other standards handle the issue of levels of description [recorded in http://www.w3.org/2008/11/11-mediaann-minutes.html#action02]
<trackbot> Sorry, couldn't find user - ruben
<scribe> ACTION: rtous to survey how other standards handle the issue of levels of description [recorded in http://www.w3.org/2008/11/11-mediaann-minutes.html#action03]
<dsinger> well, basically "content: beethoiven's 5th; realisation: berlin phil, 1962; instance: cassette tape from DG 1964"???
<tobiasbuerger> who do these three levels relate to the levels defined in FRBR (work-expression-manifestion-item)? content:==work, expression==realisation, manifestation==instance; item is not existing?)
<js> you can also look at the music ontology who has adopted the idea
<tobiasbuerger> yes, the music ontology adapted FRBR I guess
<tobiasbuerger> I guess we do not need it in our current use cases
<tobiasbuerger> (I meant the basic job ticket information)
joakim: many detailed technical aspects included
<fsasaki> xmpDM:logComment User’s log comments.
veronique: does this refer to tagging?
felix: little information in xmp spec
veronique: slot for tags needed - could this field by used?
all to look at xmp reviews, discussion continues next week
veronique: more ideas/comments about general structure should be sent to list
<fsasaki> ACTION: Felix to write a mail about the general structure of the use cases / req document [recorded in http://www.w3.org/2008/11/11-mediaann-minutes.html#action04]
<trackbot> Created ACTION-37 - Write a mail about the general structure of the use cases / req document [on Felix Sasaki - due 2008-11-18].
all to review and comment
felix: formal definition of ontology, complexity of api, etc
<dsinger> thx, bye!