See also: IRC log
<renato> https://www.w3.org/2016/poe/wiki/Scribes
<scribe> scribe: phila
<scribe> scribeNick: phila
<renato> https://www.w3.org/2016/05/30-poe-minutes
PROPOSED: Accept last week's minutes https://www.w3.org/2016/05/30-poe-minutes
<victor> my tiny comment to last minutes' is that I regretted my absence in advance
renato: Speak now if you have any objections to those minutes
RESOLUTION: Accept last week's minutes https://www.w3.org/2016/05/30-poe-minutes
renato: Take naming issue to next week
ben: I still need to come up with some ideas
<James> I added a use case last night
victor: Is it between ODRL and POE?
Ben: I#ve been asked to make a proposal
renato: I'll hand over to the
co-editors: Simon Ben and Michael
... Any discussion we need to have on this call?
michaelS: Compared with last
week's, we only have 1 additional UC, no. 17
... In the meantime, the three editors have discussed how to
work on the UCs
... Proposal is to split the UCs between the three of us
ben: Yes, agree with that, and good that we're not looking at our own use cases.
michaelS: In the meantie, Simon
confirmed that he would take the final step of moving the
content from the wiki to GitHib, transforming them along the
way.
... I have retrived thge requirements from 'my' use cases - and
this is what you find in the requirements page
<renato> https://www.w3.org/2016/poe/wiki/Requirements
michaelS: I will be away next
week
... One deatil - while working on the reqs, I made two
proposals to extend the categories by adding 'processing rules'
- we haven't had much about that in ODRL
... How to process a policy properly and let policies interact
with each other
... So that makes sense to me
... And then we also had 'implementation guidelines' - how we
shoud encourage people
renato: Any comments so
far?
... On that list of reqs, we probably need to annotate... we
need to discuss those reqs before they move to GH and become
part of the Note
... is that the plan?
michaelS: The last discussion
should be on 27 June but any earlier is welcome of
course.
... I circulated a note on what I've done so far
<victor> I raise my hand. I also have some questions
victor: I have a question
regarding the vocabularies.
... In previous versions of ODRL we have a voc adequate for
multimedia content... we have actions like display, distribute,
but this is a possible application domain?
... Will it be media centric again? For e.g. we could derive
some terms from the language terms from UC 1
... and from UC 2 we could have some related to data domain -
create, update, merge etc.
renato: Let me try and answer
that question, Victor. I don't think wewant to be domain
specific. We shouldn't have a a complete set of terms per
domain
... It was about what were common terms from common
assets
... I think we even had translate at one time but we took it
out because no one was using it
... Translate happens in multiple domains so we could add that
back in.
... we don't need to cluster things into domains
... we can have the terms any way we like
... a group could create a Note of new terms. The core specs
should be agnostic I'd say.
victor: The actions from UC 2 are very generic and there aren't many of them.
renato: UC2 is the Linked Data
one. That's not in the requirements set yet.
... In the e-mail from michaelS that wasn't reflected yet.
michaelS: That's Simon's use case
simonstey: I haven't done a lot yet but I was part of the LD Profile work. We came up with those requirements.
renato: So it's a timing issue. OK.
ivan: A completely editorial
thing. It so hapopens that another group is working on a UCR
doc so I played with a script that can be used with ReSpec that
makes it easy to have references to link between use cases and
reqs
... What usually happens is that one UC generates 2 or 3
differnet Reqs that may be shared by several UCs so one decent
way of doing it is that each UC has a list of Reqs
... And then a separate section gives details of the reqs
... So I came up with some JS that can handle that. It's a bit
shaky but it's getting there.
renato: Please send it to the editors
ivan: I'll send a copy of the mail I sent to the other WG's editors.
<Zakim> phila, you wanted to talk about domains
phila: If we list actions, we must define an extension mechanism for adding more
benws: A concept like real time
data - is that domain specific? Does it exist in its own?
... My way around it was to see whether I could think of 3
domains where a term could be used.
... I don't think ODRL is domain-specific. Additional profiles
can be added
... That was my feeling anyway
... And on the issue of profiles
... I'm looking at creating a profile for stock exchanges which
have a very specific language
phila: Talked about a likely workshop on the topic of vocab management and profiles
benws: Are profiles sufficient for an extension mechanism
phila: I believe so, yes
michaelS: From my activity area
of action vocab in ODRL, I was aware that it has to have a
shared view of what an asset it
... Which we don't actually define
<Zakim> ivan, you wanted to comment on purely editorial
michaelS: If people feel that
their domain is not covered, then we need to know.
... We can only discuss things when we have an issue
victor: Regarding hte
implementation of software based on ODRL, there should be
levels of compliance. Which profiles am I complyign with
... So we can check that the syntax is right, but with
additional validation
... Will a profile only define new terms, or can it also define
structural extensions
... In the current ODRL, there is a non-normative anne that
defines how to use Boolean operators but it's not
official
... So I wonder whetehr a profile can define these structural
ideas
... I support using SHACL for validation, but it doesn't extend
to XML
phila: (For XML you'd use XSD)
renato: Two quick ones. For
profiles, we need to more work on how we handle profiles
... In ODRL alll the onus is on the external community.
... On conformance, we need to discuss that more. ODRL is only
an expression language
<michaelS> The current ODRL Profile specs https://www.w3.org/community/odrl/model/2.1/#section-212
renato: What I wanted to get back
to is the requirements.
... Some the requirements are a bit ambiguous.
... For example, under model, usage facet of an action.
Distinguishing between academic, commercail, etc. I'm not sure
what that means so I don't know what to do with the
requirement.
... I'll go through the requirements but they need to be as
unambiguous as possible.
benws: I take your point. My
answer is that when all threee editors have gone over the doc,
hopefully the clarity will be a lot better.
... I don't understand all the reqs. We can iterate until we
all understand the reqs.
michaelS: This is a good example
where I came to the conclusion that I can feel I undersdtand
it, but I can't alweays prove that I understanbd
... My decision is whether to include it as it is, or add in
some assumptions that might change the req
... I think it sould be good to have a face to face to Skype
call to understand.
benws: E-mail not good enough?
michaelS: Not really
... It's a case of terminology too. What is a condition?
... It's not a term we've used in ODRL, for example
<simonstey> +q
<simonstey> -q
benws: If Simon and I get involved then hopefully these things will become clearer
phila: It's normal for editors to need to get together
victor: I'm available for extrea Skype calls if needs be
<renato> https://www.w3.org/2016/poe/track/actions/open
action-7?
<trackbot> action-7 -- Benedict Whittam Smith to Provide use cases on financial data -- due 2016-04-18 -- OPEN
<trackbot> https://www.w3.org/2016/poe/track/actions/7
benws: I provided 8 use cases for this
close action-7
<trackbot> Closed action-7.
<simonstey> action-5
victor: I'd like to be release from actrions 5 and 8
<trackbot> action-5 -- Víctor Rodríguez-Doncel to Provide 2 use cases from upm -- due 2016-04-18 -- OPEN
<trackbot> https://www.w3.org/2016/poe/track/actions/5
<simonstey> action-8
<trackbot> action-8 -- Víctor Rodríguez-Doncel to Add more example use cases for poe.uc.01 -- due 2016-04-25 -- OPEN
action-5?
<trackbot> https://www.w3.org/2016/poe/track/actions/8
<trackbot> action-5 -- Víctor Rodríguez-Doncel to Provide 2 use cases from upm -- due 2016-04-18 -- OPEN
<trackbot> https://www.w3.org/2016/poe/track/actions/5
action-7?
<trackbot> action-7 -- Benedict Whittam Smith to Provide use cases on financial data -- due 2016-04-18 -- CLOSED
<trackbot> https://www.w3.org/2016/poe/track/actions/7
action-8?
<trackbot> action-8 -- Víctor Rodríguez-Doncel to Add more example use cases for poe.uc.01 -- due 2016-04-25 -- OPEN
<trackbot> https://www.w3.org/2016/poe/track/actions/8
close action-5
<trackbot> Closed action-5.
close action-8
<trackbot> Closed action-8.
phila: My problem is the deadline
benws: I'm expecting to iterate
phila: Would like to leave them open but I have not been ignoring them
benws: Any other issues
renato: last month I posted a
link to the Licence Picker from the ODI
... And Serena posted a similar tool from INRIA
... What's common across both is that they have mapped the
common licences and mapped them to ODRL terms and others
... That seems a useful service that the community might want
to use
... I want a licence that does XYZ and see what it looks like
in machine terms
benws: Would that be a W3C service?
renato: I see it as a Note, on
how we've mapped terms
... If we brought them together that might be useful to the
open data world
phila: Woiuld like that very much, yes, and could work on it.
benws: So you could say, this is
what a CC-By looks like
... What would scare me is the potential number of such
licences. 30? 40?
renato: I think there are 2030 at the moment
James: On our platform, which
uses ODRL, we're doing some work on creating offers
... we'd like to make those intelligible. We'd like to use the
relevant elements of ODRL and use icons along the way.
benws: Do you allow people to apply CC licences or is it all about specific offers?
<renato> License Picker licenses: http://data.open.ac.uk/licence-picker/?controller=picker&action=index
James: We have a simple model.
We've slightly extended ODRL 2.1
... We do have CC modelled within that.
<renato> ... and ontology: http://data.open.ac.uk/licence-picker/?controller=ontology&action=index
James: We want to experiment with differnet software licences
benws: Do you think it would be useful to provide a Note of how to describe ODRL versions of common licences.
James: The issue is where would the vocab come from?
benws: I think we could use ccREL but you had to extend it?
James: Yes. Common licences so far have things like 'commercial' but that means different things to different people
<victor> Just as a reminder, since 2014 we maintain a set of nearly 200 licenses at http://rdflicense.appspot.com/
<jo> my regrets, I have to drop off the call
victor: Our list implements ODRL,
conng, for software etc. I worked with Serena on that
... I never looked for the sanction of W3C as mappings are
arguable
... But I actively maintain that dataset
James: An authoritative list would be useful.
benws: It is always *our*
interpretation
... It's who gives the authority of the interpretation
renato: So it may not be useful
benws: It would be very useful!
victor: We include who provides
the mapping
... It's very useful and practical but has no legal value
... I like it because it has multilingual support
q/
<James> https://tldrlegal.com
James: As a demo of the
issues...
... These are crowdsourced definitions of what the constraints
are on GPL3 for exammple
... People may miss bits that i think are important
phila: Rambles about what otehrs have said and, unusually, offers to help (under BDE)
smyles: I was going to express enthusiasm for this idea.
renato: if you are planning to
come, or not, please let us know
... If you're still trying to come, please let us all know.