ERT WG

HTTP Vocabulary in RDF comments

<JohannesK> http://www.w3.org/WAI/ER/HTTP/issues

johannesk: reply to issue001 from jo rabin - we don't care if no response
... does anyone think we need to deal with broken connection/no responses?
... do we need to record the reason?

carlosI: thinks this is outside the scope of the language

drooks: it sounds to me like we are trying to define how we handle exceptions
... i.e. unexpected responses/no response
... is that for us to define?

johannesk: vote for a NO to this issue?

<CarlosI> http:Response

<CarlosI> http:responseCode

<CarlosI> 5xx or whatever

carlosI: if there is no response there is nothing to record

<JohannesK> PROPOSAL: 1. Incomplete response: record at least the parts of the response that are received. 2. No response at all: out of scope.

CarlosI: if request doesnt not have a recorded response then it means there was no response so we kind of have a 'no response' covered

johannesk: but we cant recored why there was no response

<JohannesK> PROPOSAL: 1. Incomplete response: record at least the parts of the response that are received. 2. No response at all: only recording the request means there was no response (for whatever reason).

+1

<rruemer> +1

<CarlosI> +1 but maybe it need a good explanation (to avoid confusion with 5xx)

<JohannesK> RESOLUTION : ISSUE-001 1. Incomplete response: record at least the parts of the response that are received. 2. No response at all: only recording the request means there was no response (for whatever reason).

johannesk: we can deal with issue 003 and 002 together
... should we use existing properties or invent our own?

CarlosI: thinks we can use dcterms

<JohannesK> http://dublincore.org/documents/dcmi-terms/

<JohannesK> http://www.w3.org/Protocols/rfc2616/rfc2616-sec13.html#sec13.2.3

carlosI: would prefer them to be an extension

<JohannesK> PROPOSAL: ISSUE-002/-003 Put a note to "Vocabulary Scope": Timestamping is not part of the HTTP/1.1 spec, so it is out of scope for HTTP-in-RDF, it may be handled outside of this vocabulary

+1

<rruemer> +1

<JohannesK> RESOLUTION: ISSUE-002/-003 Put a note to "Vocabulary Scope": Timestamping is not part of the HTTP/1.1 spec, so it is out of scope for HTTP-in-RDF, it may be handled outside of this vocabulary.

<CarlosI> +1

johannesk: we've run out of time