See also: IRC log
<TimCole> PROPOSED RESOLUTION: Minutes of the F2F are approved: https://www.w3.org/2016/05/27-annotation-minutes.html
TimCole: any remarks about the minutes from last week?
RESOLUTION: Minutes of the F2F are approved: https://www.w3.org/2016/05/27-annotation-minutes.html
RESOLUTION: Rob completed a large amount of work on the drafts, progress has been posted
azaroth: update-- two issues opened by
Europeana folks, normalization discussion needs to agreed upon, other
than these, other issues have been closed
... should be able to quickly close the remaining issues
... ready to go to CR
... vocab still needs some examples added
<TimCole> https://github.com/w3c/web-annotation/issues/227
azaroth: recommendation from
internationalization group was that the normalization para be removed
unless specific requirements make it necessary
... normalizations might be applied due to our requirements but not
necessary to mention internationalization
ivan: propose to close the issue as they (internationalization folks) have suggested
<TimCole> Proposal: for #227 remove paragraph on normalization and close (move to editorial)
<azaroth> http://w3c.github.io/web-annotation/model/wd2/#text-quote-selector
azaroth: want to look at the 2nd para after
the table
... can try to lop the paragraph into two in order to separate the
normalization from the part describing selection
<tilgovi> Mostly just removing the DOM Strings part, then? And splitting the rest around it?
azaroth: suggesting splitting the para so that the first part discusses normalization and then rephrase the second para so that it doesn't discuss normalization at all
<TimCole> applications SHOULD implement the DOM String Comparisons method. This allows the Selector to be used with different encodings and user agents and still have the same semantics and utility.
azaroth: so first 2 sentences become a para, next sentence is deleted [?], and remainder of paragraph has all mentions of normalization removed
<TimCole> Note that this does not affect the state of the content of the document being annotated, only the way that the content is recorded in the Annotation.
TimCole: suggest we preserve the sentences
above and delete everything else
... so that no one thinks that the underlying content should be changed
... opinions? ok, with removing the paragraph altogether, but also ok
with preserving stuff about string comparisons as long as we don't
provide details on how those comparisons are to be made
azaroth: will make quick changes now, then
...[garbled]
... come back to issue before end of call
ivan: need to close remaining issues, then make a resolution to freeze features, give WG a week to review the documents, so that they can note any glaring problems, with the goal to officially request to go to CR by the end of next week
TimCole: so ok to vote next week so long as WG has been notified to review the documents by the end of today
ivan: yes
TimCole: plans for testing, need to be finalized by next week?
ivan: documents, pubs must be ready, call
with director must be set up, that period should be used 100% on testing
... so if possible plans for testing should be finalized by next week
<azaroth> Okay, new version at: http://w3c.github.io/web-annotation/model/wd2/#text-quote-selector
ivan: plans must be written down, as agreed in Berlin
TimCole: link posted to issues not yet
marked postponed or editorial
... most recent issue is for resource previews in annos
ivan: let's close issue 227, rob has made the changes
azaroth: moved selection to preceeding para, deleted everything else except the dom string mention
tilgovi: dom string comparison
recommendation was one of things called out by the internationalization
folks as something causing problems
... should not introduce normalizations there
TimCole: so drop the mention
azaroth: should we go ahead and delete the subsequent para which mentions the dom apis
<tilgovi> Shouldn't that next para read "Text Quote"?
<tilgovi> It says Position.
TimCole: deletion doesn't change the substance of the section, just means we aren't giving any help to implementers
tilgovi: would leave in the DOM api's para, otherwise people will use the selector api
TimCole: my sense is that the less we say, the better; whole thing is in flux (as discussed in Berlin)
<tilgovi> Fine for me.
<azaroth> PROPOSED RESOLUTION: Remove DOM string comparison, UTF-8, and avoid implications that comparison should be part of the normalization routine
<TimCole> +1
+1
<azaroth> +1
<ivan> _+1
<ivan> +1
<bigbluehat> +1
<tilgovi> +1
<Kyrce> +1
RESOLUTION: Remove DOM string comparison, UTF-8, and avoid implications that comparison should be part of the normalization routine
ivan: is #249 postponed?
TimCole: should be marked editor action or closed
<azaroth> http://w3c.github.io/web-annotation/model/wd2/#motivation-and-purpose
azaroth: #257 -- want to include info in the
anno that allows the client to display a snippet or preview to the end
user
... ivan has suggested this be postponed, [rob] agrees, don't know what
clients actually need to do this yet
ivan: not sure it even needs to be in the model at all
TimCole: doesn't need to be in v.1 of the model
<azaroth> PROPOSED RESOLUTION: Postpone #257, lacking information as to what is appropriate for previews of resources
<ivan> +1
<azaroth> +1
<TimCole> +1
+1
<tilgovi> +1
<bigbluehat> +1
RESOLUTION:
Postpone #257, lacking information as to what is appropriate for
previews of resources
... that leaves us with #247
azaroth: can add a sentence saying that if
you have contradictory information from external resources, believe the
external resources and not the annotation
... e.g., external resource claims target is html, and anno claims
something different, believe that it is html
TimCole: discussion?
<azaroth> PROPOSED RESOLUTION: Add to the note in 3.2.1 that information from the resource should be considered authoritative, not the Annotation's properties
<Kyrce> +q
Kyrce: is this a question of content of the resource or its format?
azaroth: not content, just its metadata
<ivan> +1
+1
<TimCole> +1
<Kyrce> +1
<azaroth> +1
<bigbluehat> +1
RESOLUTION: Add to the note in 3.2.1 that information from the resource should be considered authoritative, not the Annotation's properties
TimCole: that seems to be it; 2 issues for testing and 2 issues for pending
<azaroth> Yay! :D
<Loqi> woot
<azaroth> Thank you all :)
ivan; all done, as testing issues don't need to be addressed at this time
scribe: everything else is editor's actions
... my impression is that all of them have been addressed but they need
to be closed
ivan: features freeze, call for one week review, then CR next week
<ivan> PROPOSED RESOLUTION: The model, vocab and protocol documents are now in feature freeze, and in a one week WG review period to propose them for CR next week (06-09)
<azaroth> +1
<PaoloCiccarese> +1
<ivan> +1
<TimCole> +1
+1
<bigbluehat> +1
<Kyrce> +1
TimCole: discussion?
<ShaneM> +1
<tilgovi> +1
RESOLUTION: The model, vocab and protocol documents are now in feature freeze, and in a one week WG review period to propose them for CR next week (06-09)
TimCole: will take the testing issues out of order and
start with the new one
... ivan noted a need for interoperability testing
ivan: yes, would be nice to demo interoperability via server, in practice -- 1 client pushes an anno into a server and another client fetches that anno and displays it in its own way
<ShaneM> hmm.... I don't think that is a CR requirement.
ivan: whether we have enough implementations to do that, don't know, but would be a good extra
TimCole: like this idea in general, but
concerned that the implementations don't have much overlap w/r/t domain,
community, or topicality
... e.g., emblem annotation are kind of unique, is testing if they're
interoperable artificial?
ivan: understand, shouldn't be a formal
active criteria, but would be very nice to demonstrate
... if it can be done, it will strengthen the interoperability of the
standard
<ShaneM> protocol testing should ensure that each client sends and retrieves annotations correctly...
azaroth: seems like we should try to have 2 clients and 2 servers where client 1 makes a anno on server 1, copy it to server 2 and have it read by client 2
ivan: testing a singular implementation for the protocol is not the same as testing across multiple clients
ShaneM: however, if have multiple clients and they all pass the protocol tests, then hasn't interoperability been tested?
ivan: need a server independent from the clients
ShaneM: assuming this for protocol tests
ivan: discussed in Berlin to use scenarios
ShaneM: the web platform test infrastructure is a server, so be pointing the protocol tests at that server, then an independent server will have been provided to/for them
TimCole: to be clear, the protocol requires accepting/responding to LDP exchanges, so it didn't seem clear to us that that capacity existed at this tinme
ShaneM: the platform is capable of modeling any protocol that is desired
<azaroth> ShaneM++
<Loqi> ShaneM has 2 karma
<bigbluehat> ShaneM+++++++
<Loqi> ShaneM has 3 karma
TimCole: sounds like we have a better way to
do protocol testing than we thought in Berlin
... seems that we still need to test if anno looks the same in two
different implementation environments, e.g., Rob's implementation v
Europeana implementation
ShaneM: protocol test is next for me, sounds like vocab/model tests are there, so will move on to implementing protocol tests
TimCole: have built some schemas to test the model, is the infrastructure to trigger a run in place?
ShaneM: will post the instructions for how to do the testing to the lists
TimCole: will spend next friday on discussing testing after taking the CR vote
ivan: everyone should take some time to read through the documents
TimCole: will be very good, many small typos
lurking
... adjourn
<ivan> trackbot, end telcon