- From: Stephen Crawley <uqscrawl@uq.edu.au>
- Date: Tue, 30 Jun 2009 09:52:09 +1000
- To: Seno Akira <aqira@cs.ie.niigata-u.ac.jp>
- CC: www-annotation@w3.org
Seno Akira wrote: > Hello,www-annotation members. > I'm looking for strict Annotea server specification. > As trying to implement annotea server,I read "Annotea Prtocols",but I > cannot figure out whole of requests sent by client and replies shoud > be returned for them ,e.g. server should interpret requests about > bookmark ,some query languages like RDQL or SPARQL ,etc. > > You are correct. The "Annotea Protocols" document is not a proper specification. If you look through the recent postings on this list, you will see a number of messages from me that point out "holes" in the document; i.e. points where the document does not adequately describe the required behaviour of an Annotea server (and client). Those messages are based on what I learned in implementing "Danno". What Urs said is also true. The document is "out of date" in the sense that we want to use Annotea to do far more than the "Annotea Protocols" document envisages. Support for the bookmarks schemas, SPARQL queries and so on are some of the examples. > I think that my implementation should be able to support all features > which supported by the server on http://annotest.w3.org/annotations. > But, I could not find documents which tell about strict server > specifications on annotea still now. > There are none. > I want to know strict specifications or server implementation policy > about annotest.w3.org/annotations. > Again, there are none. Actually, it is worse than that. The "annotest" server is known to NOT correctly implement the protocols as described in the Annotea Protocols document. IIRC, annotest requires an incorrect query name to fetch replies to an annotation, and it uses the DC 1.0 namespace rather than DC1.1 which is used in the examples in the Annotea Protocols document. (And there may be other issues too.) > Please tell or suggest me about annotea server specifications. > The best I can suggest is to read the Annotea Pro carefully, read my messages (and others) on this list, and try to figure out something that works for your particular use-case. Also, if you care about interoperability with other people's clients and/or servers, test your clients against their servers and vice-versa. I'd pay most attention to client/server implementations that people are actively working on. For example Annozilla and Danno/Dannotate. Finally, you can always ask questions on this mailing list :-) -- Steve
Received on Monday, 29 June 2009 23:52:54 UTC