- From: Thierry MICHEL <tmichel@w3.org>
- Date: Tue, 27 Mar 2012 19:38:09 +0200
- To: "Bailer, Werner" <werner.bailer@joanneum.at>
- CC: "public-media-annotation@w3.org" <public-media-annotation@w3.org>
Dear all, I have included Werner's proposal in the API, PR version for review by the WG. http://www.w3.org/2008/WebVideo/Annotations/drafts/API10/PR/Overview.html#Implementation-Notes If no objections, it will remain as is. the document will be frozen on friday. Thierry. Best, Thierry. Le 27/03/2012 15:39, Bailer, Werner a écrit : > Dear all, > > please find below a draft text concerning the issue of missing/multiple identifiers to be included as informative text in the API document. > > Looking at the API document, I suggest adding it as a section "Implementation Notes" after the current section 5 (Usage examples). > > Best regards, > Werner > > ----------------------------------------------------------------------- > > This section contains recommendations for implementators for handling missing or multiple identifiers of media resources/fragments. > > x.1 Multiple identifiers of media resources or fragements > > In some source formats, it could be possible to identify the resource or one of its fragments in multiple ways, e.g. by one or more identifiers, fragment name or temporal/spatial fragment URIs. For example, there could be a temporal media fragment, which can be addressed by the time range, that also has an assigned ID. > In the RDF representation of the Ontology for Media Resources, this can be represented (as recommended in the guidelines) by using owl:sameAs. To ensure a similar behaviour in the API, an implementation SHOULD return all such identifiers in a response. If queries to properties of a fragment with multiple are made, the implementation SHOULD accept each of the alternative identifiers and return the same response for each of them. > > x.2 Missing fragment identifiers > > There are source formats, which may contain metadata about a fragment (e.g. a track) without specifying any kind of identifier for it. For the RDF representation this is not a problem, as blank nodes can be used. In an API implementation, a client requesting the list of fragments cannot query properties of a fragment in case there is no identifier. > An implementation SHOULD generate an identifier for the fragment in such a case and SHOULD ensure that it is valid for a sufficiently long time so that the client can use it in subsequent queries to properties of fragments. The identifier is not guaranteed to remain permanently valid. > This can be implemented in different ways, including the following: > - In a session-aware environment (e.g., in the user agent, in a web service environment with session handling), the identifier could be bound to the session and remain valid at least for the duration of the session. > - In a stateless environment, the identifier could be the same for all clients and remain valid a defined time after it is last used (i.e., part of a query or response). > - The identifier could be defined to be unique and permanent. In that case the implementation has to manage the assignment of identifiers to metadata sources. > > ----------------------------------------------------------------------- > Werner Bailer > Audiovisual Media Group > > DIGITAL - Institute of Information and Communication Technologies > > JOANNEUM RESEARCH Forschungsgesellschaft mbH > Steyrergasse 17, A-8010 Graz, AUSTRIA > > phone: +43-316-876-1218 personal fax: +43-316-876-91218 > mobile: +43-699-1876-1218 general fax: +43-316-876-1191 > web: http://www.joanneum.at/digital > e-mail: mailto:werner.bailer@joanneum.at > ----------------------------------------------------------------------- > > >
Received on Tuesday, 27 March 2012 17:38:32 UTC