are there any objections to the proposed minutes?
<azaroth> RESOLUTION: 21st January 2015 minutes are approved: http://www.w3.org/2015/01/21-annotation-minutes.html
- status of the WBS Poll & Update on the LDP co-location idea
discussing LDP co-location
shepazu to update on the poll status
- 11 people said they would attend the face-to-face
- 8 people responded with regrets
- most cited personal reasons, busy, or otherwise unavailable to attend
- 4 requested a dial-in line
shepazu suggesting Skype or WebRTC
shepazu is the only vegetarian
<tbdinesh> i am too (veg)
the same 11 are also interested in coming to both I Annotate & the LDP co-located f2f
some are unable to attend due to funding reasons
3 of the 11 are interested in funding support
topic correction: WBS poll (not the LDP co-location)
status of the LDP co-location idea
from azaroth
LDP discussed it last week.
there was interest. some may join, but less interest in a formal W3C-style face-to-face
some may join the annotation face-to-face or join for dinner
<shepazu> +1
but not enough interest to support a formal face-to-face with LDP
attendees to F2F: shepazu, azaroth, dwhly, TimCole, RayD, tilgovi
<Kyrce> I responded as well. And also asked for vegetarian.
scribe: some help with additional names would be great
apparently shepazu is not the only vegetarian
<tbdinesh> bigbluehat: am too. but not on poll as i could not edit
other attendeeds nickstenn, bigbluehat, Csillag, ujvari
shepazu: please toss the remote attendee names here (if you don't mind)
shepazu is checking on some survey form issues
scribe: apparently there are some issues
if anyone has trouble with the survey form, please contact shepazu
any remaining questions about the F2F?
no? moving on.
<shepazu> Please check that your status is correct: https://www.w3.org/2002/09/wbs/73180/annotation-
- ob's use case
sorry...
- Jacob's use cases
- how to increase coverage
(...did we loose azaroth ? )
scribe: apparently we lost me :(
tnx
paoloC: Progress with use cases. Should be able to say
something with multiple targets to
...
... Question as to how to use composite target, and how to signal
comparison
... Use cases in medical, xrays etc.
... Second use case annotation discovery -- how do I find annos about
something
... definitely a use case. Example is finding annotations in a
particular context
... Could be able scope of the annotation, or more broadly
<Jacob> yes, via hasScope
paoloC: Third use case is annotation of annotations for discussion.
<Jacob> exactly!
paoloC: Question about multiple targets -- reply targets an
annotation, but is also about the it be both?
... Added to the wiki. Another two that have been sent in that include
automatic annotations
... For example to find tweets about a document and create annotations
that link the tweet to
... Other is more manual, a researcher that finds information about a
resource and annotates ogheter
... nature of the connection should be described
<RayD> +q
Jacob: in CG there was a project to track discourse with
annotations
... expected use was that professor would give a lecture and the
annotations would capture the
... model wise target could just be the annotation, but more accurate
might be to have specific
PaoloC: Would it make sense to point to the prototype or description of it?
Jacob: We have some reports, will send them to you [PaoloC]
PaoloC: That would be great
<Jacob> Sounds good!
<bigbluehat> thanks Jacob !
RayD: Motivations, I'm wondering what the views are on
whether we should have use cases for
... Everything in the spec should be justified by a use case
... Section on motivations seems to be list of types of annotations that
we have imagined
... subject to extension. Should we have use cases
<bigbluehat> bye paoloC
<Jacob> +1 for a use case for each motivation
<bigbluehat> +1 to RayD
<TimCole> +1
<azaroth> azaroth: +1
TimCole found some motivations missing for various use cases
scribe: plans to submit them to the list
azaroth mentions that `describing` aggregated many of other motivations originally proposed
scribe: may be a need to revisit that and break them back out
any further use case discussions?
<Jacob> IIRC, motivation was also intended to be an extension point for the model, where folks could eir use case
<RayD> +q
Bill_Kasdorf: wonders if there should be uses cases for distinguishing the types
<Jacob> I thought that there was a mechanism that used skos to define new motivations...
?
Jacob: points out that there is a way to define new motivations that would provide for wth
azaroth: a use case for extending the motivations would be valuable
<Jacob> I think so, so I
RayD: there are already use cases for finding annotations. Use cases for finding by specific
azaroth: +1 for using motivations for discovering
annotations
... for instance tagging
next topic?
azaroth: sent straw person proposal based on lengthy mailing list discussions
<azaroth> http://w3c.github.io/web-annotation/protocol/wd/
link: http://w3c.github.io/web-annotation/protocol/wd/
contains thoughts about protocol ^^
intended as a straw person proposal *not* has azaroth as chair person stating the new
scribe: going through the table of contents now
... if you have further thoughts this week, comment on the mailing list
... and discuss on future calls
... ivan to chair next week
no objections to reviewing the protocol proposal
please disagree if you disagree
section 3
3.1 is about containers from LDP
scribe: discusses what a container is in general terms
... posits 3 different types of containers
... an annotation, containers related to a particular annotation, and
other resources that heets, etc)
... last group is not part of the annotation, but necessary for its
interpretation
next section is about CRUD
scribe: creation is a POST
... body of the POST is the JSON-LD of the annotation
... updating: you can replace (via PUT) or sending partial updates via
PATCH
... PATCH uses an RDF format
... delete does what it says
... it's possible to state a name for the annotation via a Slug header
... server may or may not use that header for identification
... Retrieval
... GET on the URI of the annotation
... listing is via interaction with the container
... a big green TBD for search
... not yet written up events from the Social Web WG
any comments? does this sound like something you could get behind? or should we keep looking
+q
<RayD> I like this approach
+1 from bigbluehat
TimCole: asks about the suggesting of a URI for a resource
azaroth: for example, you want to post an image as a body
and you want to give the image a lient needs a way to ask for a certain
name be given to it
... otherwise the server will do what it likes (incrementing number,
UUID, etc)
... it's not mandatory. even if you support it, you do not have to do
what the client is
<TimCole> got it.
azaroth: server may use whatever scheme it uses
... reason to do it is that if you know the server will allow it, is
that you can reference ed in the system
... otherwise, you'd have to clean up your client-side annotation after
the server responds
shepazu: nothing azaroth said didn't make sense...well...
... azaroth, you are proposing a sort of LDP path forward for the
protocol?
... azaroth: yes.
azaroth: could you repeat that? you were echoing
<azaroth> We can't hear you if you're talking doug :(
LDP but less firmly baked?
ivan: it's been a while sinse I looked into LDP.
... most of what is here is basic HTTP verbs
<azaroth> My point was: we should constrain the features of LDP. A profile of LDP rather than just you should do that
ivan: LDP doesn't add too much to that. It comes in with the concept of the container, and correct?
azaroth: that's pretty much it. A few additional
requirements around signaling--assuming
... via the Link header
ivan: if we decided to define a protocol ourselves, then is it correct that we would end up very close to LDP?
azaroth: if we followed the Web Architecture then we would end up with something very similar
shepazu: is back.
... what form would our suggestion of LDP take?
... a specific spec? here's how you do annotations?
... or could it be, if you're using something by the Social Web WG,
here's how.
... I assume this isn't at the data model?
azaroth: correct.
... (chair hat off) my perspective is that we should find one solution
whenever 6 different protocols at once
... if the social web working group comes up with something very
different, then I think we n the two
<ivan> +1 to Rob
azaroth: in order to further adoption. otherwise, you'll have optimizations for specific servers
shepazu: don't completely agree. we're not at the point
where we need to pick a course yet
... social web has not yet determined how they will deal with
annotations
... are we going to have a spec that says "when you use LDP, this is
what we expect you to
... how are we going to say what the relationship is?
... is it an advisory? or a normative note?
azaroth: it's normative. when using LDP, you MUST do this
or that
... you must use the context from the annotation WG
np
<dwhly> I'm about to join our design meeting
ivan: trying to find some examples. allow for turtle, but we might restrict to a specific
ivan: a simple, but still normal spec
back...but I think it's over :(
azaroth: sorry I had to drop off.
<azaroth> no problems
<tbdinesh> bigbluehat: you there? FYI, i will also be attending the face to face. But I could not edit sue - they are looking into it)
yep
tbdinesh: good to know. I'd ping shepazu about that
<shepazu> tbdinesh, you already got the response... you're using the wrong account