See also: IRC log
<trackbot> Date: 19 November 2014
<Bill_Kasdorf> Getting tons of static on the phone
<Bill_Kasdorf> Okay now
I am willing to attempt it.
<azaroth> scribenick: Kyrce
Proposal to accept last week's minutes. Any objections?
<azaroth> RESOLUTION: 12 November 2014 minutes approved: http://www.w3.org/2014/11/12-annotation-minutes.html
Reminder: ongoing call for consensus to publish the first public working draft.
Please read the document if you have not yet.
Reminder: first working draft, not in any way final.
Please give any red flags or any way that you would be unwilling for it to be published.
Two topics: 1. use cases.
Directions between systems and a s a means of discovering what people need to do. Requirements so we can move forward
2. The next face to face. Proposal that it be in San Francisco and cohosted.
There are some concerns in that it would be three face to face meetings in a row in the same location.
Use cases should take longer, so discuss face to face first.
ivan: Concerns were stated on email. Next F2F is actually in Japan. For us here in Europe, three F2F that are very far and complicated (flight, finances).
... There are advantages to being close to iAnnotate.
<tbdinesh> can someone tell me the conf code pls
<tbdinesh> to call
azaroth: geographic distribution is important. Any other thoughts about coming back to SF next april
TimCole: Do we have any sense of how many members will attend iAnnotate anyway?
<dwhly> Actually Sloan is sponsoring
<rayd> I am unable to make it.
azaroth: Some travel sponsorship available, cost will be less of an issue but timing and jet lag are.
<tbdinesh> (tnx ivan, am in. 2666 did not work the 1st try)
<azaroth> acl ivan
paoloC: two aspects: location and funding. Location biggest issue. iAnnotate is easy to attend. Location is a problem.
Ivan: A lot of people want to go to iAnnotate but that argument may hold in 2016 which would mean ALL the F2F meetings would be in SF. Not a big issue, if the majority is fine with SF, "I won't lie down on the road for that"
Ben: Recommended moving iAnnotate around. Might be good to have the gathering in different places. More participation and help from other members of the community.
Azaroth: if there were offers of help for organization, would that be acceptable for 2015 or is that already in motion?
Ben: not yet in motion, just started queries to Sloan. Nothing is reserved; just put queries together
How many people here would travel to Europe to an annotation conference?
Ivan: I am worried Sloan will have problems to finance travel outside the US
Ben: Looking into that.
TimCole: What are the other options for Spring 2015? Are there other meetings that people tend to go to on the East Coast US or in Europe?
Can we do this over email before our next call in two weeks?
Ben: the location is incidental to the proposal, I need to find out if it's possible to have a location in Europe
Paolo: Two years ago we had an event centered on annotation at Harvard. One of the Institutes is interested in annotation.
Correct last as Paolo, not Ivan.
<dwhly> x
Azaroth: Paolo investigat hosting into Harvard; Dan find out from Sloan where they would be able to fund travel; Azaroth to get help from list for volunteers and for help with logistics, including hosting and other conferences people might travel to.
Ivan: WWW is slightly later (May) in Florence Italy but submissions for conference deadline has passed. New York Book Fair, and IDPF usually has a conference that week. Publishing Interest Group may have a meeting as well. Might be complicated to squeeze in iAnnotate there. That's also later in the year.
... week of 25 May
Azaroth: Any more comments about F2F?
No more comments.
Azaroth: Proposed workflow for the working group to come to understanding issues between transferring annotations between systems
1. obtain use cases; 2. analyze them; 3. look at existing standards and technologies. Hopefully we find that all the requirements are covered by existing technologies and we can just make recommendations.
Likely there will be requirements that are not covered or that are not covered appropriately
and we will need to create others
Comments?
None.
Two possible things on call: Easy one, discuss how we want to manage step one.
The other: harder for Kyrce, go around and get the use cases on the call itself.
Rather than offline.
Suggestion: do the use case selection offline, use a future call to discuss
<bjdmeest> +1
PaoloC: In the past: use cases are really vague. There are abstract use cases, and there are detailed use cases that have real connotation.
Azaroth: good point. I was intentionally being vague so people could put in what they wanted and we could try to manage this.
We could be more specific.
paoloC: I have a lot of high level use cases. clients to server, server to server, very high level
"I need client A to communicate to this server and return the annotation to client B"
Azaroth: how about we ask for more user stories rather than use cases
something concrete that people actually need to do.
<tbdinesh> azaroth: +1
Then we won't be trying to do everything all at once.
Requirements for real systems, these are things that people will actually need rather than things that they theoretically desire.
rayd: The use cases are the stories and the higher level are requirements derived from the use cases. Two step process; requirements derived from the stories.
Point out that i submitted three use cases yesterday that are in the form of stories; are these the sort that we are talking about?
Azaroth: I haven't read them but I know the background. They are good but the type we need are about the transfer, rather than using annotations for specific purposes
rayd: don't you need specific purposes to generate the requirements?
Azaroth: Agree, but it was the focus on the purpose rather than on the interaction. We can create additional for other deliverables. We can use them as stories towards examples.
... do we want to have a call about use cases for annotation in general? or do we want to be more specific about the sort of use cases we are looking for at this moment?
<ivan> http://www.w3.org/TR/csvw-ucr/ : uc for the csv on the web wg
Ivan: Making very clear what kind of use cases we are looking for now is a very good idea. We can spend much time on them and lose sight of our other work. My original question: example of something that has worked well, I have the use cases for csv on the web working group
... Stories; and deriving requirements in a separate chapter which was the interesting work. What issues are common? What requirements can we answer, and what requirements are out of scope?
... that approach had worked well in the past.
Azaroth: that is what i had in mind.
paoloC: Server to Server communication, to be more concrete
... I need to retrieve all the specific annotations for a certain URL.
... I need to retrieve specific annotations across multiple URLs. Which one is a specific use case?
Ivan: for the use cases to be plausible and useful, it has to be bounced through some REAL use cases.
... Work where this feature is necessary.
... what i think rayd called a story, has to be included as well.
... can't only come from an academic perspective. Real stories, real cases. Filter: when two use cases are similar we should not duplicate
Ivan; the other part of my answer is that what counts is the use case and not the model. We can say that this group cannot respond to this use case. We should not forget about a use case just because it doesn't apply in today's model.
Ivan: We are looking for use cases for the protocol parts. I would also like to see use cases for the javascript API and for robust anchoring.
Azaroth: do you think we should get use cases for js API and robust anchoring now? or in general?
Ivan: I would be happy if we had the various use cases for they may not be independent of each other. I understand this is more wokr
Ivan; *work but it's better than coming back in six months with ones that impact the use cases already created.
<paoloC> The second use case I mentioned is already implemented and allows integration between existing clients (Utopia for PDF and Domeo): https://www.youtube.com/watch?v=OrNX6Sfg_RQ however it requires extensions that are not in the current Model (Bibliographic objects management).
Azaroth: so we should create use cases for the protocol, the js API and the robust anchoring.
paoloC: The use case is already created. Requires servers to store information that is not part of the model. It's not just protocol, it is also part of a model. is it a use case/
Azaroth: agreed use case.
Ivan: use case.
Azaroth: add name.
TimCole: What are the sources of the use cases? There is a lot of existing work. Especially what Rob did before the WG was formed in analyzing and making more generic use cases in ePub group. You posted use cases in March. Perhaps they need to be selectively enriched? Would they give us a starting point?
<TimCole> http://www.w3.org/TR/dpub-annotation-uc/
Azaroth: we should treat that document as input from the digital publishing working group to this group.
... There wasn't a lot of discussion about it, and there is some conflict of interest since I was heavily involved with each. I'd like to have this validated.
Ivan: I don't think it's a conflict of interest. The document will be published in a final form soon.
<azaroth> +1 to Ivan
Ivan: When we derive all the requirements we should go to dpub and see if there are requirements there we need to reflect here. The important thing is to look at it from a requirement standpoint.
Azaroth: I think we have a good understanding of what we want. Summer; Call for use cases about annotations. Ask that respondents tag with type (protocol, js API or third) . Collect over a period of time, and when we feel like we have enough we will analyze. They should be stories rather than requirements. Fair summary?
<rayd> +q
Azaroth: How to do it? we could use documents (unwieldy); google doc (specific tech); email list (needs submitters to access list, not anonymous);
rayd: Email list is the most expeditious way. If you assign someone to consolidate then an anonymous user can send to that person.
<bigbluehat> +1 here
Azaroth: sounds good
Ivan: what about using a wiki?
... github wiki or w3c wiki
Azaroth: maintain in wiki, someone is responsible for transcribing from mailing list to wiki
<bigbluehat> +1 ^^
Ivan; most obvious. Can be difficult in email with formatting, etc.
TimCole: wiki also. Wiki makes it easier to keep discussion focussed.
... Wiki keeps better trail
Azaroth: github vs. w3c?
Ivan: w3c is stable; github is great for repo.
<rayd> github wiki difficult formatting
<bigbluehat> +1 for w3c wiki
Azaroth: transcribe from mailing list to w3c Wiki.
<Bill_Kasdorf> +1 w3c wiki
Rob: I will take it on if no one else can.
<bigbluehat> ^^ sound right?
Azaroth: everyone is responsible to submit, we need someone to manage getting them from the email into the wiki and to manage the wiki.
paoloC and bigbluehat
Azaroth: would you want to send request to mailing list, or should I do that?
<bigbluehat> +1
paoloC: better coming from Chair, and you can explain the mechanism on the wiki and we will take care of formatting and other issues.
Azaroth: any further comments?
dinesh: wondering if I could be paired with someone to build up some use cases?
Azaroth: yes, everyone is new and equal, this is a new working group so no one should feel like there are existing social hierarchies that need to be maintained. Suggestion: discuss them on the mailing list. But if you prefer to discuss offline we can do that as well.
dinesh: I prefer discussing offline if possible.
Azaroth: I can do that if you like, or paoloC or bigbluehat.
... any other use case comments, or any other business?
<dwhly> +q
<azaroth> PROPOSAL: use annotation to get feedback on the specs?
<TimCole> +1
dwhly: possibility about using annotator on… no test URL as yet. Would people be open to using Annotation as a way to get feedback on spec and incorporate that?
<JakeHart> +1
<bjdmeest> +1
<Matt_Haas> +1
<bigbluehat> +1
<azaroth> +1
<paoloC> +1
+1
<ivan> +1
<MGU> +1
Ivan: can you stay on after call to discuss?
paoloC: we update the specs often and they all fall on the same URL. What is the behavior of the hypothesis platform with that.
<azaroth> RESOLUTION: use annotation to get feedback on the specs
dwhly: if the annotation has not been deleted by created, and the text has changed, it should still remain if the text has not changed substatially.
... if you annotate a word and the sentence is deleted, the annotation is not attached but still discoverable.
... if you annotate a word and the sentence is moved, the annotation should move.
... we have things that we are focussed on delivering but are not built yet.
paoloC: there should be guidelines, if the chunk of text is changed and we are having a discussion, where does the discussion go? We can try and set up conventions.
dwhly: if there are obvious ways we need to change and improve we have devs to work on it.
azaroth: no call next week, thanks to everyone!
... reconvene in two weeks.
<bigbluehat> bye all
<tantek> is this channel logged?
<azaroth> tantek: Not other than RRSAgent during calls
<azaroth> as far as I know!
<ivan> tantek: http://www.w3.org/2014/11/19-annotation-irc is the log, and I will clean up the minutes. See also https://www.w3.org/annotation/wiki/Minutes for the minutes in general
<tantek> ivan - is that a daily automatic log? or only during telcons/f2f meetings?
<ivan> I think it is only for the telcon
<ivan> let me see what the script does...
<ivan> trackbot, end telcon