Provenance Working Group Teleconference -- 14 Mar 2013

<trackbot> Date: 14 March 2013

<pgroth> Scribe: Stian Soiland-Reyes

<pgroth> @stain you ready?

I'm joining

<pgroth> great

sorry

<Luc> time change only affecting europeans

Admin

<pgroth> http://www.w3.org/2011/prov/meeting/2013-03-07

pgroth: Votes on minutes, please

0 (not present)

<dgarijo> +1

<KhalidBelhajjame> +1

<jcheney> +1

<GK> +1

<pgroth> accepted: Minutes of March 7, 2013 Telcon

<Luc> :-)

pgroth: Hoping http://www.w3.org/2011/prov/track/actions/122 will be done eventually by end of WG ..

Documents are published

<pgroth> http://www.w3.org/blog/SW/2013/03/12/call-for-review-prov-family-of-documents-published-as-proposed-recommendations/

<Luc> 4!

pgroth: Congratulations everyone, here is Ivan's blogpost.

<Dong> +1

pgroth: not that much of a hazzle, got it published in the end
... one thing we did in the rush (13 documents) was to not do our standards of writing blog posts and PR as for other releases
... and so I wanted to see if people would be willing to write a blog post and/or send emails to particular mailing lists
... announcing specially the working drafts
... we still have 9 Working Drafts we would like to have reviewed
... can I get some volunteers to write blog posts or email lists?

<KhalidBelhajjame> I can send mails to mailing lists: DataOne, DbWorld, and others

<GK> I guess I could send email to the LDP group as king for review of prov-aq

<dgarijo> I write a blog post, similar to Ivan's: http://linkingresearch.wordpress.com/2013/03/13/the-prov-family-of-specifications-is-released/

<dgarijo> *wrote

pgroth: Khalid, yeah, great, could you do that

stain: volunteers to do a blog post about PAQ

pgroth: and GK if you could write the LDP group

<dgarijo> ok

pgroth: Dani, I will write an overview blog post; perhaps you could write something about the DC Note?
... and did you/Kai notify the DC people abou tthe draft?

dgarijo: yes, it was announced on the web page (?) ; but not sure if was announced on public list, will ping him abou tthat

pgroth: that's everything
... will send emails to group..

<pgroth> https://www.w3.org/2002/09/wbs/33280/provpf/results

pgroth: with the proposed REC, we need to give ?C refs to vote for our documents. They have to vote that it goes to recommendations

<pgroth> https://www.w3.org/Member/ACList

pgroth: we would like you to contact your AC representative to vote for the recommendation
... here you can find your representative
... we need as many votes as possible can to push to REC

<Zakim> stain, you wanted to write PAQ blog post

I'll ask the Manchester one

<jcheney> i've mentioned it to henry

<Luc> i did it for soton

PROV-AQ

pgroth: What we want to do is to resolve some of the core issues
... sent around a reminder to look at these issues
... GK to add some context?

GK: It makes sense to talk about the individual issues.. added a couple of small thigns to the agenda
... do we want to confirm that the issues that I propose to close without further discussion are OK?
... in the order you proposed, or different order?

pgroth: does not matter order, but want to talk about the core issues (which could take time) - the pending ones we can go through fast

GK: it might make sense to do the PENDING ones first...

pgroth: no, at the end
... first issue is ISSUE-618 http://www.w3.org/2011/prov/track/issues/618

Should pingback be described in PROV-AQ?

GK: since we've put the revised pingback proposal in the document (which was published), which to my mind looks more like a provenance discovery mechanism, .. the question was put to the mailing list eariler ; do we want to include or exclude this
... Luc commented to exclude it.. there was statements of inclusion support from at least 3 people - with me that is 4
... as I see it, there is a good reason to include it in that there is reasonable, if not overwhelming support to keep it in place
... there was no co??? reason to exclude it, it was brought within the general scope of the document
... but many did not like the name "Forward provenance" - we do need a better name

<Luc> what about provenance?

GK: one possibility - was "downstream provenance"
... Two issues: i) Does anyone have any reason not to include it? ii) Alternative names - here throwing in "Downstream provenance"

Luc: I still maintain the view expressed in the email
... feel that the requirements as (??) should be, never really been agreed by the WG
... as for the rest of the document it was guided by the scenario
... it seems to be that what is proposed is (?) solution; (?) there was another one.. there could be others
... but nowhere can I (?) them, we did not agree what are the rquirements
... so that is an issue with it
... But if the group decides to include it, we should discuss the name

<Luc> agreed by the Working Group!!!!

GK: Disagree with the characterisation of not having requirements; in the wiki what Tim initialyl proposed was a requirement that was guiding the design
... would have to dig to find the URI.. This was mentioned in my email response to Luc

<pgroth> http://www.w3.org/2011/prov/wiki/Provenance_ping-backs

<Luc> is there a group resolution endording this document?

GK: there are 3 scenarios in this wiki page; while we did not go into the same level of detailed analysis; I don't think this is fair to say it was not proposed without requirements

<Zakim> GK, you wanted to disagree about requirement: Tim proposed one which guided the design

Luc: there was no set of requirements which was agreed by the WG - there is no resolution deciding this

GK: these were up for discussion, but nobody disagreed with them ; but agree there was no formal resolution
... from what Ivan said I did not believe a formal resolution as needed

pgroth: I think we did agree to try it out when we talked about it at last F2F
... I can try to dig that up and confirm.
... what I actually asked on the mailing list was what is the role (?)

<GK> My recollection concurs with Paul - we did agree to look at pingback

pgroth: agree in a sense with Luc; the.. (?)
... whether the requirement ...(?) design requirements. Tim had a go at it, and proposed a solution, then Stian proposed a solution. I did not.
... If we say that this is a note.. but if this was a recommendation; then I would agree we shoudl not do this; too experimental.
... and we have not look ed at it as long as the other thrings. But this is a note.
... in other cases, like PROV Links, or other things we thought were useful, but preliminary, we just publish them as NOTE
... It would come under the status of a note

<GK> Agree. Nothing more to add.

pgroth: I could propose one solution, is that we are currently in a WD phase; and this is included now in Stian's proposal. And so we would like to particularly get review on the pingback.

<GK> +1

<Luc> i don't think there is time

pgroth: and based on that decide to remove or keep
... but there's a review period now

Luc: do you mean internally or externally?

pgroth: right now we are in a phase were we have released a WD for external feedback

<GK> (That is: go forward with it as is, and pull it if there are significant problems i review)

pgroth: and we are in a phase doing internal reviewing
... and so if we get comments from external reviews.. or if internal reviews show issues..

<GK> I plan to ask LDP to review: I could draw their attention to this area.

Luc: that (?) is already there..
... we are in that situation.. it's time to make a decission

GK: are there others than yourself that said it was not fine

Luc: No.. and I am not going to vote on this

pgroth: so you don't want to re-review ?

Luc: when I sent my email this week; I had just read the text for the first time
... perhaps we can have an informal vote on if there are other objections
... then it could be a resolved matter

<pgroth> straw poll: include ping back in the paq

Luc: +1 is to keep, and -1 if you want to remove

<jcheney> +0 (haven't reviewed but don't object)

<smiles> +1

<GK> +1

<dgarijo> +0

<KhalidBelhajjame> +0

<zednik> +0

<Dong> +1

<SamCoppens> +0

<CraigTrim> +0

<jcheney> It seems to me that a note is an appropriate place for a preliminary design, as long as it's clearly marked as such

I would call that luke warm calling for more review..

Luc: would like to hear pgroth's view as an PAQ editor

pgroth: my view is that I kind of like it; but needs more review
... because it's the newest thing in the PAQ it needs another round of review

<Dong> +1

pgroth: if there is errors I would want it out because of time

<GK> I'm with Paul's view here - if problems are exposed pull it.

pgroth: if that is the concensus.. GK are you OK to proceed like that?

GK: ok, that is entirely reasonable. If there are problems we don't have time to rework it.

pgroth: and we just released the WD, so encourage people to re-read it properly. Also contacting the LDP group etc

GK: not technical reworks.

pgroth: and then the renaming issue.. we should be able to deal with it without changing technical bits

Luc: if we keep it in, I am of the view that we shoudl not qualify (?) prospective... don't think it is suitable. It is provenance.
... no kind of qualification of the provenance

pgroth: OK, suggest to leave the renaming issue out
... some concerns about the name

<Zakim> GK, you wanted to say I'm OK with Luc's position from a technical perspective, but would be good to have some motivation.

GK: technically it's fine to not commit to upstream or downstream.. but would be useful with motivation for what the mechanism is there for. Just editorial.

stain: thinks that there should be an editorial motivation on the most typical usecase of notifying-upstream-pingback; but not technically limit one way or another ; I should be free to pingback some provenance about how King Richard III was found under a parking lot

<pgroth> proposed: to ask for more review of the ping back mechanism and if there are technical problems then remove, otherwise keep pingback

<GK> OK

<pgroth> accepted: to ask for more review of the ping back mechanism and if there are technical problems then remove, otherwise keep pingback

pgroth: should we recommend RDF for provenance? http://www.w3.org/2011/prov/track/issues/428

GK: the mechanisms that are presented in indepdendent from provenance format (as requested)
... there is still a weak recommendation that PROV-O in a "standardized RDF format" is suggested, but not required for the mechanism
... Luc had a comment abou tthat.. if we want to change it I would be fine
... but many of the feedback we have got is to use PROV-O and RDF, and perhaps nudging people in that direction

(I've had people asking me today about PROV-JSON via PROV-AQ)

<GK> Curent text: "Most mechanisms described in this note are independent of the provenance format used, and may be used to access provenance in any available format. For interoperable provenance publication, use of PROV-O represented in a standardized RDF format is recommended. Where alternative formats are available, selection may be made by content negotiation."

Luc: lots of REST services out there just XML and JSON. It would be fine if they could export provenance. It would be good to help them indivdiually.. (?)
... that I would think they want to export provenance in the same formats
... so my view is to not promote RDF in this case
... just recommend the use of PROV serializations
... we have already media types for PROV-N, there's PROV-XML (media type?), and then RDF

<GK> I'm OK with this change if that's the group's view.

pgroth: also in favour of that personally. Just say "Use PROV" should be good enough
... PROV-O will rpobably win the day anyway.. I don't think PAQ needs it

stain: was initially pushing for the "Should be RDF" bit so that there would be some kind of promise or recommendation of what kind of serialization you would find; but is buying into Luc's argument to just go for any PROV serialization

<GK> Suggest "For interoperable provenance publication, use of PROV represented in any of its specified formats is recommended. "

<pgroth> proposed: "For interoperable provenance publication, use of PROV represented in any of its specified formats is recommended. "

<Luc> it's good, thanks

+1 (RECOMMENDED)

<smiles> +1

<KhalidBelhajjame> +1

<SamCoppens> +1

pgroth: any objections?

<dgarijo> +1

<zednik> +1

<jcheney> +1

<GK> +1

<pgroth> accepted: "For interoperable provenance publication, use of PROV represented in any of its specified formats is recommended. "

is this 425?

<Luc> issue-425?

<trackbot> ISSUE-425 -- Why does the service description need to be rdf? -- pending review

<trackbot> http://www.w3.org/2011/prov/track/issues/425

thnx

trackbot is clever :)

<trackbot> Sorry, stain, I don't understand 'trackbot is clever :)'. Please refer to <http://www.w3.org/2005/06/tracker/irc> for help.

GK: one comment has been that Why does Service Description have to be RDF in
... and the other by Stian, was to mention JSON-LD as an explicit format
... My response ; Similar to the provenance format.. the mechanism would work with any service description format; but the only one we are actually describing is one based on RDF. There is more bias towards RDF in this case. but does not precelude the use of alternative formats
... the main reason here was that the RDF one was the easiest one to expecify. The format we use use the RDF linked data properties (?) - allow us to have multiple serve descriptions in the same document.
... would think it was too late to define anything else.. but what we have is a service description based on RDF.. but left open in the document to use other formats.
... as a final comment.. the idea to use other formats came out strongly from LDP group as well (Linked Data Platform)

using content-negotiation to get different service description is common in world of XML web serices

GK: and so remain compatible; but taking it further in our use of RDF

pgroth: to summarize - we allow any service description format using conneg ; we give one example of how it is described in RDF

GK: ok, but stronger than example

pgroth: ONE way

GK: not mandatory, but only one we specify

stain: Still think that Luc's argument from before also applies here; my JSON-LD proposal was a way to give a simple JSON format that just happens to also be valid JSON-LD (and therefore correspond to our RDF format)

pgroth: (???) leaving the door open forservice descriptions; specially in terms of REST. There is no common way to do REST service descriptions
... we can give one easy way to do it. But we don't mandate it.

Luc: If I was to write it, I would do it the way you said, Paul
... I noticed how a service description language format... (?) We identified this is the information we expect to find.
... and for illustration, here's an example using RDF
... using content-negotiation to find the representation

<GK> Paul's formulation sounded good. Would be happy to work on that.

pgroth: think my formulation is not too far from what's there.. I can fine-tune it

<Luc> agreed, it's minor fine tuning

GK: happy with that. The way it came over in pgroth's expla=nation was good. So that they can use whatever works in their environment

(sorry I did not get to scribe most of what pgroth said before!)

<pgroth> proposed: work on refining the editorial around the description of service descriptions in the paq

pgroth: any objections?

<pgroth> accepted: work on refining the editorial around the description of service descriptions in the paq

should we make that action?

pgroth: now to go through list of pendingreview

<GK> http://lists.w3.org/Archives/Public/public-prov-wg/2013Mar/0090.html

<pgroth> ACTION: pgroth to update service description editorial [recorded in http://www.w3.org/2013/03/14-prov-minutes.html#action01]

<trackbot> Created ACTION-165 - Update service description editorial [on Paul Groth - due 2013-03-21].

GK: running through the document of MUST and MAY.. done in last editing/review
... Oh, right!
... we just figured out what to do with 425 in this meeting

<Luc> issue-300?

<trackbot> ISSUE-300 -- Quote vs Quotation (Involvement versus Activity) -- closed

<trackbot> http://www.w3.org/2011/prov/track/issues/300

GK: ISSUE-600 PROV-pingback was an old issue to add pingback, now toclose

<Luc> \issue-609?

<pgroth> iisue-600

<Luc> issue-609?

<trackbot> ISSUE-609 -- Specify how to locate a SPARQL endpoint for querying provenance -- pending review

<trackbot> http://www.w3.org/2011/prov/track/issues/609

GK: ISSUE-609 is how to locate a SPARQL endpoint.. now covered by Service Description

<Luc> issue-622?

<trackbot> ISSUE-622 -- Should PROV-AQ bless use of JSON-LD for service description? -- pending review

<trackbot> http://www.w3.org/2011/prov/track/issues/622

GK: ISSUE-622 about JSON-LD.. I think we just agreed how to address that

<Luc> issue-624?

<trackbot> ISSUE-624 -- Should PROV-AQ specify PROV service URI or always use template? -- pending review

<trackbot> http://www.w3.org/2011/prov/track/issues/624

GK: The ISSUE-624 - about specifying service URI or template.. this was going back to an earlier discussion
... where we are we always get the direct access URI by means of a template in service description

<Luc> issue-628?

<trackbot> ISSUE-628 -- Specification of anchor in HTML/RDF vs HTTP is inconsistent -- pending review

<trackbot> http://www.w3.org/2011/prov/track/issues/628

GK: and ISSUE-628 there was an issue raised by others about inconsistency aobut anchor specification in HTML vs HTTP
... there is an inconsistency.. which we discussed earlier.. but one which is of small importance only arrising in edge cases
... the document specifies how to use these things in a way that avoids the inconsistencies
... that means we invent less new mechanisms
... so that is the list of issues that is now PENDINGREVIEW and I propose to CLOSE - given no objections

Luc: to me, I was not asking for a redesign.. just meant a note in the text; there is a difference between what the two approaches (?) could do.

GK: ok, that is a good point. I'll make a note to myself to do that

<pgroth> ACTION: gk to add a bit of text explaining the inconsistency between html/rdf and http [recorded in http://www.w3.org/2013/03/14-prov-minutes.html#action02]

<trackbot> Created ACTION-166 - Add a bit of text explaining the inconsistency between html/rdf and http [on Graham Klyne - due 2013-03-21].

<pgroth> accepted: close the pending review issues listed in the minutes

pgroth: on response to James Anderson.. running out of time

timetable check

pgroth: important topic..

<pgroth> http://www.w3.org/2011/prov/wiki/WorkplanTillFinalPublication

pgroth: we have essentially trying to close shop some time end of April
... we need to stage all documents by 2013-04-23
... asking all editors to put their final proposals for timeline (when things are to be done)
... I think all editors have done that
... are any of the editors concerned about the time?
... have a couple of other tasks on that page to be done
... namespace pages
... updating the FAQ
... making the PROV page on the wiki better
... need volunteers for those other tasks
... I would do the namespace task
... Provenance of Documents
... Luc said he would do that.. Tim would look at PROV-O's prov
... but editors should write their own PROV

stain: should not each of the formats use their own format for their PROV..?

pgroth: to use content-negotiation between formats

<GK> (and a .htaccess to handle the content negotiation?)

pgroth: but need templates?
... running out of time.. could people volunteer for the rest of this?
... yes, deadline would be bout 2013-04-23.. or really 2013-04-30

(personally that's too short for me to help, given easter etc)

<GK> (Difficult for me to commit to more than prov-aq on that timescale)

GLD last call

<pgroth> http://www.w3.org/TR/2012/WD-vocab-org-20121023/

pgroth: Government Linked Data Group has published their last call for their ontology.. and it uses PROV. It would be good for some of us to review it before last call

Luc: we've already reviewed...?
... was published in october
... Jun and I drafter a response from the WG. The document has not changed.

pgroth: so you asked them to change it, but they have not?

<Dong> @Paul, can you give brief descriptions for "Other Tasks" on the wiki page, so I can see what I can help? Thanks.

Luc: right, that is still the bversion in October. It was only announced on..(?)

pgroth: but I was reading this email..

Luc: perhaps we should talk to Ivan. About derivation.. was assuming (?) with activities

<KhalidBelhajjame> Teher is a more recent version that dates of March the 14th: https://dvcs.w3.org/hg/gld/raw-file/default/org/index.html

Luc: there was another one
... not sure what they are doing

<pgroth> ACTION: luc and paul to talk about gld [recorded in http://www.w3.org/2013/03/14-prov-minutes.html#action03]

<trackbot> Created ACTION-167 - And paul to talk about gld [on Luc Moreau - due 2013-03-21].

pgroth: for everyine to think about how we can promote draft and the proposed recommendations
... DO get your AC ref to vote

<dgarijo> bbye

<KhalidBelhajjame> thanks, bye

<Luc> bye

<SamCoppens> Bye

<zednik> bye

<jcheney> bye

<pgroth> trackbot, end telcon

<GK> Bye

- DRAFT -

Provenance Working Group Teleconference

14 Mar 2013

Attendees

Contents

Admin

Documents are published

PROV-AQ

timetable check

GLD last call

Summary of Action Items

Scribe.perl diagnostic output