ERT WG

Review of action items

<JibberJim> Hi people, sorry I have no phone available :-( there's insufficient internet here for my voip

action item 1 - Shadi and Charlos todo

item 1 - Carlos will post by next Monday (RDF-CL in EARL)

test

<shadi> jim, we are looking at open action items

<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2006Apr/0001

<shadi> i beliebe your 1st action item is partially closed, right?

<shadi> and your second one, you need to repind the annotea folks

EARL pointers proposal by Jim

<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2006Mar/0003

<JibberJim> yes, I need a little finishing up on the 1st, if everyone is happy with what I posted, and the 2nd yes I need to chase the annotea folk again

<niq> also for the minutes: The proposal makes a lot of sense as far as it goes, but that's just a first reaction

JK - content needs to be converted to text. Need another property that that describes encoding.

<JibberJim> Do you mean the original encoding? would this not come from the HTTP vocab?

<JibberJim> the encoding in snippets would be in the encoding of the RDF file - ie it could be transcoded from the original

<niq> c.f. my comment on encoding wrt HTTP-in-RDF from last night (EU time)

<JohannesK> Jim: if you have e.g. images/PDFs you only have bytes not characters, but you need to transform these bytes into text (e.g. with Base64) because EARL is a text format

NIQ no need to record info that is already in HTTP info

<shadi> jim, is it ok to give you an action item to clean up the proposal and provide RDFS that can be plugged into the schema document?

<JibberJim> ah, JK I wasn't considering the snippet etc. methods would work for non text content

<JibberJim> yes shadi it's fine

<shadi> ACTION: jim to clean up the proposal and provide RDFS that can be plugged into the schema document [recorded in http://www.w3.org/2006/04/05-er-minutes.html#action01]

SZ - wonders if we should use something like byte offest length?

JK - doesn't make sense to mix. Should be char offset.

NIQ - simple tools use byte offsets easier.

SZ - problem with unicode (chars can use more that one byte)

SZ - some times bytes instead of chars makes sense

<scribe> ACTION: JK to send comment to list [recorded in http://www.w3.org/2006/04/05-er-minutes.html#action02]

URI-in-RDF request by CarlosV

<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2006Mar/0021

SZ - we should be cautious about encoding everything in RDF.

JK - was CarlosV's idea and he had use cases in mind but I don't know what they are.

JK - If don't want to use it in EARL should not be required.

SZ - URI is important for EARL so needs to be properties

SZ - many items can be put in URI so what is benefit of seperating out?

CV - consistency

SZ - What is problem with EARL URI?

SZ - there are minor drawbacks so we need a solid use case

SZ - more efficient to record just header info and not content

CV - would like a single property/class to describe, currently have too many things

JK - need use cases to describe where split is necessary, else use single property

SZ - good suggestion

<scribe> ACTION: CV will document and elaborate benefits/drawbacks to using single/multipe properties [recorded in http://www.w3.org/2006/04/05-er-minutes.html#action03]