ERT WG -- 21 Mar 2007

SAZ: several comments about HTTP in RDF

<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2007Mar/0082

SAZ: some quite important, so maybe we should put HTTP in RDF on hold
... try to decide which should we address before publication
... proposal to add text at section 1 to clarify the scope of the vocabulary

CI: could help people to focus comments on scope

RESOLUTION: add clarification of the vocabulary scope at section 1

SAZ: pospone discussion about timestamps

JK: agree

CI: +1

RESOLUTION: pospone discussion about timestamps

SAZ: add some kind of blurb about the conventions for values normalization?

CI: useful, but could wait

<JohannesK> yep

SAZ: no clear idea right now about how it should like
... jump to #4 for the moment, decide on # later
... #4 is something that is not a core feature of the vocabulary

RESOLUTION: pospone discussion about http:transcript

SAZ: #5 feedback requesting unified mechanism to access properties, right now we have differences between named properties and unknow properties
... implies a substantial change on the model
... the change provide more benefits than problems
... don't see any negative impact

JK: message header with two properties: name and values
... with the new approach the name is no necessary any more
... you have to create a new class or subclass for unknow properties

SAZ: don't think so

<JohannesK> e.g. foo:X-Fun owl:subClassOf http:MessageHeader

SAZ: the field name should be fixed

JK: how to model a fixed field name?

SAZ: are you saying it's not possible?
... we need to wait for alternatives

<JohannesK> JK: proposal is ...

<JohannesK> JK: http:header -> http:MessageHeader

<JohannesK> JK: http:MessageHeader http:fieldName http-h:accept

<JohannesK> JK: http:MessageHeader http:fieldValue "text/html"

RESOLUTION: change the name properties of header to resource values using the owl:oneOf approach

<scribe> ACTION: JK to take his proposal about owl:oneOf to the mailing list for further discussion and refinement [recorded in http://www.w3.org/2007/03/21-er-minutes.html#action01]

JK has just done his action

<JohannesK> JK: fieldValue can have Literal or (collection of) HeaderElement

<JohannesK> JK: HeaderElement has properties elementName (required), elementValue (optional), several param

<JohannesK> JK: Param has properties paramName and paramValue

<JohannesK> elementName is Literal, elementValue is Literal, param should be Param

<shadi> header = [ element ] *( "," [ element ] )

<shadi> element = name [ "=" [ value ] ] *( ";" [ param ] )

<shadi> param = name [ "=" [ value ]

RESOLUTION: adopt the Jakarta method for header element values

SAZ: clarification of what XML mean in the algorithm
... already disscused but not transmited

RESOLUTION: refine description of the algorithm

SAZ: right now fixed list of response codes
... response code class right now is oneOf a fixed list
... we can add another oneOf thing for generic responses

RESOLUTION: provide mechanism for extensible response code

SAZ: if we have a sequence of request we may want to record them as they came

JK: a Collection is ordered

SAZ: in that case it's a simple fix

RESOLUTION: describe the sequence of request in the Conection class

SAZ: back to #3
... proposal to try to address it editorially

RESOLUTION: (editors) add blurb on normalization conventions

SAZ: could you incorporate changes?

JK: yes

ERT WG

21 Mar 2007

Attendees

Contents

Summary of Action Items