Re: storing EARL in annotea

>Jim, I had a quick chat with ericP about this, and he says that you need
>to connect the sub query type things with each other with the variable
>names (correct me if I've got this wrong eric)

We're working it out on IRC now.  I think I understand now how to structure
a query, but I'll need to do some exploring.  I'm not sure I entirely
understand what's going on with what I get returned, though.

(ask
     '((http://www.w3.org/1999/02/22-rdf-syntax-ns#type ?a
          http://www.w3.org/2001/03/earl/0.95#Person)
          ) :collect '(?a))

Returns:

<earl:Person
    xmlns:earl="http://www.w3.org/2001/03/earl/0.95#">
      <a:Attribution
r:resource="http://iggy.w3.org/annotations/attribution/1014886708.379036" />
   </earl:Person>

along with a bunch of empty earl:Person tags.

I'm assuming all the empty Person tags come from every report that matched,
but why is there only one with the annotea Attribution?  It's always the
same attribution, too.  Eric told me that that one attribution would point
to a combination of the reports that matched, but it doesn't seem to.

>> > A few warnings from my experience playing with annotea:
>> >
>> > 1. The documentation at
>> http://www.w3.org/2001/Annotea/User/Protocol.html
>> > is wrong.  Annotest gives errors if you try to just post straight XML
>> to
>> > it.
>>
>> I have two clients which both post straight XML without problems, the
>> above page is wrong, but only in that it omits the <?xml ... > with that
>> included it works fine for me - what problems do you have?

Ahh... I had left that off.  It works fine now with straight XML, thank you.

>> > Look at the source for
>> > http://annotest.w3.org/annotations?explain=false for examples of how to
>> > submit things - for example, to submit an annotation you have post your
>> > content as w3c_annotate=<url-encoded rdf>.  Same idea for algae
>> queries.
>>
>> I definately don't like this for submission - url length is getting long,
>> I don't trust proxy's even if I can trust annotest to be okay with it.

I'm still using POST.  The only difference is the URL-encoding.  So
apparently the script accepts both methods.  I'm glad to not have to
url-encode now.

>> Also I believe it's important we agree on a namespace to use, algae
>> queries really need to know what namespace you are using and it would be
>> nice if we could all use the same one.

Yes... unfortunately, EARL looks like it's going to continue to change.  Is
it time to decide on a 1.0 now?

Nadia

Received on Friday, 1 March 2002 10:30:42 UTC