Chatlog 2012-05-31

See original RRSAgent log or preview nicely formatted version.

Please justify/explain all edits to this page, in your "edit summary" text.

14:47:19 <RRSAgent> RRSAgent has joined #prov
14:47:19 <RRSAgent> logging to http://www.w3.org/2012/05/31-prov-irc
14:47:21 <trackbot> RRSAgent, make logs world
14:47:21 <Zakim> Zakim has joined #prov
14:47:23 <trackbot> Zakim, this will be 
14:47:23 <Zakim> I don't understand 'this will be', trackbot
14:47:23 <Luc> Zakim, this will be PROV 
14:47:24 <trackbot> Meeting: Provenance Working Group Teleconference
14:47:24 <Zakim> ok, Luc; I see SW_(PROV)11:00AM scheduled to start in 13 minutes
14:47:24 <trackbot> Date: 31 May 2012
14:47:32 <Luc> agenda: http://www.w3.org/2011/prov/wiki/Meetings:Telecon2012.05.31
14:47:44 <Luc> Chair: Luc Moreau
14:47:54 <Luc> Scribe: Tom DeNies
14:47:59 <Luc> rrsagent, make log public
14:48:04 <Luc> zakim, who is here?
14:48:04 <Zakim> SW_(PROV)11:00AM has not yet started, Luc
14:48:05 <Zakim> On IRC I see RRSAgent, Luc, pgroth, MacTed, stain, trackbot, sandro
14:48:17 <Luc> Regrets: James Cheney, Paolo Missier
14:50:58 <Zakim> SW_(PROV)11:00AM has now started
14:51:05 <Zakim> +BrendanIAB
14:53:17 <Luc> topic: Admin
14:57:09 <Curt> Curt has joined #prov
14:57:47 <Zakim> +Curt_Tilmes
14:58:01 <GK_> GK_ has joined #prov
14:58:08 <GK> GK has joined #prov
14:58:26 <Zakim> + +1.518.276.aaaa
14:58:50 <tlebo> tlebo has joined #prov
14:59:12 <Zakim> +Luc
14:59:15 <smiles> smiles has joined #prov
14:59:17 <TomDN> TomDN has joined #prov
14:59:28 <Zakim> +??P49
15:00:18 <Zakim> +TomDN
15:00:26 <sandro> [ Luc - regrets from me for today, sorry for the late notice ]
15:00:33 <Zakim> +??P42
15:00:42 <dgarijo> dgarijo has joined #prov
15:00:48 <Zakim> -??P42
15:00:56 <Luc> @sandro, ok, had you emailed the announcements of working drafts to various w3c lists?
15:01:07 <sandro> Yes, I did.
15:01:15 <Luc> thanks, action can be closed.
15:01:18 <sandro> thanks
15:01:21 <satya> satya has joined #prov
15:01:27 <Zakim> +??P53
15:01:33 <Zakim> +??P54
15:01:41 <dgarijo> Zakim, ??P54 is me
15:01:43 <GK> zakim, ??p53 is me
15:01:43 <Zakim> +dgarijo; got it
15:01:49 <Zakim> +GK; got it
15:02:06 <SamCoppens> SamCoppens has joined #prov
15:02:29 <dgarijo> http://www.w3.org/2011/prov/wiki/Meetings:Telecon2012.05.31
15:02:48 <Luc> proposed: to approve Minutes of the May 24 2012 Telecon 
15:02:55 <GK> Thanks
15:03:00 <Luc> http://www.w3.org/2011/prov/meeting/2012-05-24
15:03:02 <dgarijo> +1
15:03:04 <TomDN> +1
15:03:04 <smiles> +1
15:03:09 <SamCoppens> +1
15:03:13 <Curt> +1
15:03:15 <Zakim> +Satya_Sahoo
15:03:18 <satya> 0 (did not attend)
15:03:27 <Zakim> +OpenLink_Software
15:03:34 <zednik> zednik has joined #prov
15:03:34 <MacTed> Zakim, OpenLink_Software is temporarily me
15:03:34 <GK> +1
15:03:36 <MacTed> Zakim, mute me
15:03:40 <Luc> resolved:  Minutes of the May 24 2012 Telecon 
#15:03:45 <Zakim> +BrendanIAB.a
15:04:03 <Zakim> +MacTed; got it
15:04:05 <Zakim> MacTed should now be muted
15:04:11 <TomDN> Luc: review open actions
15:04:15 <Zakim> +??P30
15:04:31 <tlebo> tlebo has joined #prov
15:04:34 <TomDN> Luc: there was an action on Graham to review the constraints, and he did
15:04:35 <SamCoppens> zakim, SamCoppens is with TomDN
15:04:38 <tlebo> zakim, who is on the phone?
15:04:41 <Luc> topic: Schedule for internal release and review  
<Luc>Summary: Editors reported the following. prov-primer is ready for review. prov-dm/prov-n are ready, except for role and hasProvenanceIn currently under discussion. prov-o should be ready by mid-next week, except for role and hasProvenanceIn. There is currently no schedule for prov-constraints. We agreed that reviews will start when all issues are tackled to decide whether we go for last call.  We will revisit this agenda item next week. 
15:04:57 <Zakim> +SamCoppens; got it
15:04:58 <TomDN> Luc: all other open actions closed as well
15:04:59 <Zakim> On the phone I see BrendanIAB, Curt_Tilmes, +1.518.276.aaaa, Luc, ??P49, TomDN, GK, dgarijo, Satya_Sahoo, MacTed (muted), BrendanIAB.a, ??P30
15:05:02 <Zakim> TomDN has TomDN, SamCoppens
15:05:06 <tlebo> zakim, I am aaaa
15:05:17 <Zakim> +tlebo; got it
15:05:22 <TomDN> Luc: we need to agree on the time when  the documents can be released for review, and for how long
15:05:36 <dcorsar> dcorsar has joined #prov
15:05:50 <TomDN> ... there are still 2 open issues
15:06:04 <TomDN> ... on the agenda today to close them
15:06:23 <TomDN> ... but release of documents would not be held up by these issues
15:06:37 <TomDN> ... It is unclear when the CONSTRAINTS document would be ready for review
15:06:54 <TomDN> SimonM: prov-primer is ready for review
15:07:13 <Luc> q?
15:07:15 <TomDN> Luc: prov-o?
15:07:30 <TomDN> tlebo: About a dozen issues in the tracker
15:08:18 <TomDN> ... most important issues is hasprovenancein and dictionary 
15:08:33 <TomDN> ... most others are editorial issues and shouldn't pose many problems
15:08:44 <TomDN> Luc: When would the document be ready for review then?
15:08:54 <TomDN> tlebo: Could be by early next week
15:09:15 <Zakim> +??P2
15:09:19 <Christine> Christine has joined #prov
15:09:20 <TomDN> ... except hasProvenanceIn
15:09:24 <GK> If PAQ terminology is part of hasPriovenanceIn discussion, I'd move to put it in a separate namespace ... if only to keep the issues distinct in our discussion.
15:09:44 <pgroth> +q
15:09:44 <Luc> q?
15:09:49 <TomDN> Luc: Should we wait until these issues are resolved, or proceed with review?
15:10:23 <TomDN> Paul: It would be nice to have reviewers take a look at it before the last call
15:10:49 <pgroth> ack pgroth
15:11:02 <tlebo> +1, would be nice to have reviewers be for "Last Call" (for real)
15:11:14 <TomDN> ... We already released something recently, so there is no real pressure to release something right now. It would be better to wait until everyone agrees that the document is ready for last call
15:11:17 <pgroth> yes
15:11:23 <dgarijo> i think so, yes
15:11:31 <TomDN> Luc: Should we postpone this to next telecon?
15:11:34 <GK> I shall miss the next two teleconferences .. travelling, meetings, etc.
15:11:50 <Luc> topic: Dublin Core Best Practice  
<luc>Summary: Kai, Daniel, and Simon have worked with Dublin Core members  and produced a draft document explaining how to map some DC concepts to Prov. Tim, Paul, Satya, Stephan and Sam will review the document. Other WG members are invited to provide feedback on the document as well.  The document will be transferred to the W3C infrastructure shortly.
15:11:50 <TomDN> ... By then we will have agreed on outstanding issues.
15:12:19 <dgarijo> https://github.com/dcmi/DC-PROV-Mapping/wiki/Mapping-Primer
15:12:25 <TomDN> danielG: I have sent an email this morning with the mapping links
15:13:47 <TomDN> ... We have divided the mapping in a primer, a specialization of prov-o, a direct mappings doc and finally, a complex mappings doc
15:14:04 <Luc> q?
15:14:07 <pgroth> +q
15:14:15 <TomDN> ... next steps are to complete the round trip
15:14:38 <Zakim> +??P13
15:14:39 <GK> OCLC?
15:14:58 <TomDN> didn't hear the question properly, sorry
15:15:09 <TomDN> Zakim, mute me
15:15:09 <Zakim> TomDN should now be muted
15:15:15 <Luc> q?
15:15:18 <Luc> ack pg
15:15:18 <pgroth> ack pgroth
15:15:27 <stephenc> stephenc has joined #prov
15:15:31 <Luc> q?
15:15:41 <TomDN> danielG: if we have more problems, we will move it to the W3C wiki
15:16:17 <TomDN> Luc: I think it is important that we move this document to the W3C website (for intellectual property reasons)
15:16:44 <TomDN> ... My suggestion is: move to wiki (or most appropriate location)
15:16:51 <Luc> q?
15:16:51 <TomDN> danielG: OK, will do this
15:17:10 <TomDN> Luc: Who would be willing to review the DC mapping document?
15:17:16 <tlebo> I'll try.
15:17:17 <dgarijo> It is not very long i promise :)
15:17:20 <pgroth> +1
15:17:31 <satya> +1
15:17:33 <zednik> +1
15:17:34 <SamCoppens> +1
15:17:43 <dgarijo> thanks to all.
15:18:10 <Luc> topic: definition of role
<luc>Summary: We discussed the property provo:hadRole (the attribute provdm:role ). The  object of the discussion was the domain of this property, and the corresponding English definition. There is a preference to have a domain as broad as possible, the challenge is now to find an English definition for this notion.  We didn't reach a conclusion. Graham volunteered to draft a definition. Discussion is to continue by email. In the interest of time, chairs will seek to put forward a consensus proposal.
15:18:19 <Luc> http://www.w3.org/2011/prov/track/issues/384
15:18:25 <TomDN> Luc: thanks to Daniel for putting the effort into this document, it is great that we can connect to the Dublin Core community
15:18:38 <GK> The documet looks good to me on cursory glance.  Exposes and analyses issues nicely
15:19:29 <TomDN> Luc: As I was reviewing the PROV-DM, I came across this issue http://www.w3.org/2011/prov/track/issues/384
15:19:48 <TomDN> ... The challenge is getting the right definition for the DM
15:20:20 <TomDN> Luc: Suggestions are welcome. 
15:20:25 <tlebo> q+ to respond to role
15:21:06 <TomDN> GK: my initial inclination here is consistence with the DM
15:21:41 <TomDN> ... It is difficult to find a definition that is both technically correct, and suits all the intuitions
15:22:21 <TomDN> ... I would be inclined to choose the technical definition, and then provide further illustration
15:23:15 <TomDN> GK: All of these roles are subtle variations of existing relations of the DM
15:23:26 <TomDN> Luc: is the word "role" appropriate?
15:23:40 <TomDN> GK: In many cases, yes. 
15:23:48 <Luc> q?
15:23:56 <TomDN> ... In most of the useful cases, it fits.
15:24:16 <TomDN> ... It invokes intuition, without being too committing
15:24:28 <TomDN> Tim: I have some comments
15:24:38 <TomDN> ... I'd rather try to keep it broad.
15:24:54 <TomDN> ... and avoid constraining it too much, to where it can be used.
15:25:39 <tlebo> * prefer to keep it broad - avoid pinning down and constraining where it can be used.
15:25:44 <tlebo> * having role on Involvement seems to make sense.
15:25:48 <tlebo> * the name of the role should prefer the "object" instead the "subject" (it's the role of the prov:involvee)
15:25:52 <TomDN> thanks :)
15:26:07 <TomDN> Luc: Not sure it works. 
15:26:40 <TomDN> ... In some cases the role applies to the subject, in others to the object. For example in usage-generation
15:27:00 <pgroth> +q
15:27:04 <GK> But isn't the role a *relation*, applying to the combination of subject *and* object?
15:27:04 <tlebo> q-
15:27:19 <TomDN> ... Finding something that is intuitive for the user is the point, and maybe "role" is not the right word.
15:27:32 <TomDN> Paul: I wouldn't go with over-specification
15:27:41 <GK> @paul +1
15:28:21 <TomDN> ... We want to provide syntax for people to use when describing their provenance. I don't know if further specifying "role" would affect the semantics much. 
15:28:33 <TomDN> ... Better to wait to see how it will be used.
15:28:48 <tlebo> @paul, but concerned about the "subject vs. object" of swapping - can lead to confusion.
15:28:48 <MacTed> q+
15:28:52 <Luc> a role is the function of an entity, agent, or activity with respect to another entity, agent, or activity
15:28:56 <TomDN> Luc: I just want to find a good English definition
15:28:57 <MacTed> Zakim, unmute me
15:28:57 <Zakim> MacTed should no longer be muted
15:29:05 <GK> I would offer to provide some word, but I have very limited time in the next two weeks.
15:29:06 <pgroth> ack pgroth
15:29:46 <TomDN> MacTed: My sense of "role" in this space is that there are some primitives which that some entity may be acting as
15:29:53 <TomDN> ... in a provenance description
15:30:15 <TomDN> ... An entity is acting as an entity in a defined process
15:30:22 <Luc> @macted, current definition is:  A role is the function of an entity or an agent with respect to an activity
15:30:43 <TomDN> ... I don't think that people will use this kind of terminology
15:30:52 <GK> @macted ... actually one place where roles do come up is in inputs to workflow elements, which ism exactly like divisor, etc.
15:31:02 <Luc> q?
15:31:03 <tlebo> so @macted just wants agents to be able to have roles, and not entities?
15:31:05 <Luc> ack macted
15:31:26 <TomDN> GK: One of the areas where roles are important in is workflow provenance
15:31:42 <tlebo> @gk, yes, role = input and role = parameters is useful to describe the Entity's Usage in an Activity.
15:32:03 <TomDN> ... This is exactly the place where the object-subject problem is posed
15:32:13 <pgroth> yes
15:32:15 <Luc> @macted, current definition is:  A role is the function of an entity or an agent with respect to an activity
15:33:08 <TomDN> Luc: in the model, we have relations that don't mention an activity
15:33:39 <TomDN> Luc: Do we go for the current definition, or with a broader domain? (but no definition yet)
15:33:40 <satya> But, why is role only applicable in context of activity
15:33:56 <TomDN> MacTed: examples of these relations?
15:34:03 <tlebo> thats meta-provenance, @macted.
15:34:03 <TomDN> Luc: Attribution for example
15:34:16 <TomDN> ... Derivation
15:34:44 <Luc> q?
15:35:30 <TomDN> Luc: We have a choice. Either decide what the domain is and work out a definition. or take the definition and adapt the domain to fit it
15:35:35 <pgroth> +q
15:35:50 <TomDN> ... What would people prefer?
15:36:25 <TomDN> pgroth: Its hard to define, but maybe we can get around it by being more open
15:36:35 <Luc> q?
15:36:38 <Luc> ack pg
15:36:45 <TomDN> GK: Broadly, I agree with Paul. 
15:37:25 <TomDN> ... Would like to take a shot at a definition, but will be travelling the next week, so it is possible that I don;t have the time
15:38:17 <TomDN> Luc: Let's do that. We make the domain as unrestricted as possible. and try to work out a definition
15:38:18 <Luc> topic: Provenance Locator (hasProvenanceIn)
<Luc>Summary: Luc summarized issues related to the proposed relation hasProvenanceIn(subject, bundle, [alias]). The debate seemed to converge on the question of whether we define a new qualified relation to allow optional alias.  We didn't reach a conclusion either. Discussion is to continue by email. In the interest of time, chairs will seek to put forward a consensus proposal.
15:38:29 <Luc> http://www.w3.org/2011/prov/wiki/ProvenanceLocator
15:38:41 <Luc> http://www.w3.org/2011/prov/track/issues/385
15:38:54 <TomDN> Luc: I drafted a summary of the issue on the wiki (see link above)
15:38:55 <tlebo> FWIW, what about making prov:oHadRole and prov:sHadRole to distinguish between talking about the subject or object of the Involvement?
15:39:12 <tlebo> \ prov: sHadRole and prov:o HadRole
15:39:29 <tlebo> \ prov:    oHadRole
15:39:36 <TomDN> ... It is important to be able to navigate the distributed graph structure we create using the PROV-DM
15:39:46 <Luc> http://dvcs.w3.org/hg/prov/raw-file/default/model/releases/ED-prov-dm-20120525/prov-dm.html#term-hasProvenanceIn
15:39:55 <Luc> hasProvenanceIn(id, subject, bundle, target, attrs),  
15:40:06 <MacTed> (not past tense?)
15:40:48 <TomDN> ... With this relation, we are sayign that a given subject has some provenance in a bundle, and the target is an alias that is known under a different name in the bundle
15:41:03 <TomDN> ... The issue was raised that this is too complicated.
15:41:15 <dgarijo> @MacTed: good point, but it is referring to where the provenance IS located, not where it was located.
15:41:16 <Luc>   hasProvenanceIn(subject, bundle) 
15:41:29 <TomDN> Luc: It was then suggested to simplify to above
15:41:50 <tlebo> "shock" sioc
15:42:06 <Luc>      hasProvenanceIn(subject, bundle, alias) 
15:42:15 <TomDN> ... there are some precedents in literature for this kind of relation
15:42:35 <TomDN> ... Target may be better understood as alias (see above)
15:42:40 <Luc>       hasProvenanceIn(subject, bundle) 
15:42:46 <Luc>       alternateOf(subject,  alias) 
15:42:46 <tlebo> +1 to "alias" instead of "target-uri"
15:43:03 <TomDN> ... It was suggested that 2 binary relations might be better than 1 ternary
15:43:49 <TomDN> ... There was a comment made that it still not precise enough, and it isn't clear what the alias is in the bundle
15:43:51 <Luc>      hasProvenanceIn(subject, bundle, alias, [prov:type="..."]) 
15:43:56 <GK> q+ to comment on not knowing what to look for in the bundle...
15:44:25 <tlebo> @luc @gk, you look up "any" of the alternateOfs?
15:44:28 <Luc>      hasProvenanceIn(subject, bundle, alias) 
15:44:43 <GK> @tlebo yes!
15:44:53 <TomDN> ... When we talk about the alias, we refer to a name, not a resource
15:44:54 <smiles> @tlebo agreed
15:44:59 <Luc> q?
15:45:05 <smiles> q+
15:45:50 <TomDN> GK: In response to issue that it doesn;t help with navigation: you would look for all of them
15:45:56 <Luc> ack gk
15:45:56 <Zakim> GK, you wanted to comment on not knowing what to look for in the bundle...
15:46:45 <TomDN> SimonM: Replacing the target with alternateOf is depending on how the entities are related
15:46:58 <TomDN> ... which is important
15:47:18 <pgroth> +q
15:47:22 <tlebo> FWIW, I think I proposed hasProvenanceIn() can decompose into prov:isTopicOf, prov:alternateOf, and prov:atLocation
15:47:30 <TomDN> ... Even if you have the alias in hasProvenanceIn, you should still have the alternateOf relation
15:47:33 <Luc> q?
15:47:34 <GK> q+ to ask my broader question: why do we need aliases at all?
15:47:36 <TomDN> (or specializationOf)
15:47:38 <Luc> ack smi
15:48:02 <TomDN> pgroth: Aren't we trying to put too much into this hasProvenanceIn?
15:48:15 <MacTed> \ owl:sameAs serves the purpose of saying joe:thing1 and fred:thing2 are aliases for the same entity
15:48:16 <TomDN> ... simpler would be: look in this bundle.
15:48:21 <pgroth> ack pgroth
15:48:25 <tlebo> q+ to say that the only new thing that hasProvenanceIn offers is prov:isTopicOf, so we should reduce it to just that.
15:48:40 <Luc> ack gk
15:48:40 <Zakim> GK, you wanted to ask my broader question: why do we need aliases at all?
15:49:04 <TomDN> GK: My broader issue is that the requirement is to be able to navigate a distributed graph. The hasProvenanceIn is suitable for this
15:49:14 <pgroth> are main purpose is to do provenance of provenance
15:49:16 <Luc> q?
15:49:18 <pgroth> which is why we need bundles
15:49:24 <TomDN> ... The stuff that is being added is causing more confusion than adding useful properties.
15:49:28 <pgroth> and thus hasProvenanceIn
15:49:40 <Luc> q?
15:50:00 <tlebo> I think I proposed hasProvenanceIn() can decompose into prov:isTopicOf, prov:alternateOf, and prov:atLocation
15:50:15 <TomDN> tlebo: I think that hasProvenanceIn can be composed into a new istopicOf relation and alternateOf that we already have
15:50:22 <Luc> q?
15:50:26 <tlebo> the only new thing that hasProvenanceIn offers is prov:isTopicOf, so we should reduce it to just that.
15:50:26 <Luc> ack tlebo
15:50:26 <Zakim> tlebo, you wanted to say that the only new thing that hasProvenanceIn offers is prov:isTopicOf, so we should reduce it to just that.
15:50:31 <TomDN> ... It might be better to reuse the thing we already have.
15:50:43 <dgarijo> @Tim: http://xmlns.com/foaf/spec/#term_primaryTopic?
15:51:04 <tlebo> @dgarijo, need to relax the "primary-ness"
15:51:25 <tlebo> if foaf:primaryTopic had a superproperty foaf:topic, then yes.
15:51:30 <GK> SO why not just use the same name?
15:51:33 <TomDN> Luc: I think that having to search the entire bundle when looking for an alias is not a good solution.
15:51:42 <MacTed> +1 GK
15:51:45 <tlebo> @gk @who?
15:51:47 <TomDN> ... When you assert provenance, you want to release your bundles fast
15:51:55 <TomDN> ... but also connect them properly.
15:52:05 <MacTed> Zakim, who's here?
15:52:05 <Zakim> On the phone I see BrendanIAB, Curt_Tilmes, tlebo, Luc, ??P49, TomDN (muted), GK, dgarijo, Satya_Sahoo, MacTed, BrendanIAB.a, ??P30, ??P2, ??P13
15:52:08 <Zakim> TomDN has TomDN, SamCoppens
15:52:08 <Zakim> On IRC I see stephenc, Christine, dcorsar, tlebo, zednik, SamCoppens, satya, dgarijo, TomDN, smiles, GK, GK_, Curt, Zakim, RRSAgent, Luc, pgroth, MacTed, stain, trackbot, sandro
15:52:09 <pgroth> +q
15:52:12 <TomDN> ... and currently we have no suitable mechanism for this
15:52:33 <MacTed> q+
15:52:40 <pgroth> ack pgroth
15:52:52 <Luc> q?
15:52:59 <pgroth> +q
15:53:12 <tlebo> @gk, waht about "using same name"?
15:53:20 <TomDN> ... The challenge is that you can't just extend an existing relation with an alias, since it is binary
15:53:36 <GK> @tlebo didn't I just say that? :)
15:53:39 <TomDN> Macted: this problem is inherent to a distributed graph
15:53:46 <Luc> q?
15:53:49 <tlebo> @gk, yes, but what are you saying?
15:53:50 <Luc> ack mact
15:53:55 <TomDN> ... At some point asserters will assert in the wrong way.
15:54:08 <pgroth> i think I have a nice way out
15:54:08 <TomDN> ... We can't stop that.
15:54:18 <Luc> q?
15:54:34 <GK> Ah, ... I mean use the same name when referring to the same thing.  URIs are, after all, a global namespace.
15:54:50 <GK> So use the same name in different budles.
15:55:05 <tlebo> @gk, ok, so you're not referring to my proposed name of isTopicOf ?
15:55:14 <TomDN> pgroth: A way around this is: we have hasProvenanceIn, but we allow it to be extended
15:55:57 <TomDN> Luc: this is a possibility. It would imply that you need a qualified relation for it. But the idea seems to be that we don;t want that.
15:56:03 <tlebo> why do we need a qualified relation on this?
15:56:11 <pgroth> for extension
15:56:12 <GK> @tlebo ... not sure.  I haven't really studied that - or is it just a name change?
15:56:21 <tlebo> gimmie isTopicOf and we're done.
15:56:28 <Luc> q?
15:56:31 <satya> q+
15:56:33 <pgroth> gotta go, sorry
15:56:41 <tlebo> @gk, no, it's a rename and reduction to hasProvnenaceIn - alternateOf - atLocation.
#15:56:41 <Zakim> -BrendanIAB
15:57:05 <tlebo> q+ to say qualifying this property seems like overkill.
15:57:12 <Luc> ack pgr
15:57:28 <TomDN_> TomDN_ has joined #prov
15:57:33 <TomDN_> (connection dropped out)
15:57:53 <TomDN_> GK: I don't see why would need really this alias mechanism
15:58:22 <tlebo> so luc's problem is linking to "already created" bundles?
15:58:32 <tlebo> s/to/two/
15:58:55 <TomDN_> satya: my concern is that it is a very narrow scenario.  I'm not sure we as a WG should meddle in thi
15:59:10 <Luc> q?
15:59:13 <Luc> ack sat
15:59:19 <TomDN> satya: we should avoid making the model overcomplicated
16:00:27 <TomDN> tlebo: If we reduce hasProvenanceIn to hasTopicOf it would solve this problem
16:00:36 <TomDN> Luc: the problem is aliasing
16:00:54 <TomDN> tlebo: When you change the id, use alternateOf
16:01:15 <TomDN> Luc: then you would end up looking for all alternateOf relations, which is unfortunate
16:01:57 <GK> It seems to me that this whole issue is being motivated by implementation concerns, when what we are defining here is a data model.  Premature optimization ... and all that.
16:01:59 <TomDN> ... Then you could en up with multiple specializations for the same agent in one bundle
16:02:03 <MacTed> +1 GK
16:02:05 <TomDN> s/en/end
16:02:21 <tlebo> q?
16:02:23 <tlebo> q-
16:02:34 <TomDN> Luc: we will continue this discussion via email
16:02:47 <tlebo> thanks!
16:02:51 <Zakim> -??P13
16:02:52 <Zakim> -??P2
16:02:53 <Zakim> -??P49
16:02:55 <GK> I am feeling inclined to appeal to the decisoioin from the last F2F -- lacking consensus, drop it.
16:02:56 <Zakim> -tlebo
#16:02:57 <Zakim> -BrendanIAB.a
16:02:59 <Zakim> -Satya_Sahoo
16:03:03 <Zakim> -Luc
16:03:05 <Zakim> -dgarijo
16:03:08 <Zakim> -MacTed
16:03:11 <Zakim> -Curt_Tilmes
16:03:13 <Zakim> -??P30
16:03:15 <Zakim> -TomDN
16:03:19 <Zakim> -GK
16:03:22 <Zakim> SW_(PROV)11:00AM has ended
16:03:23 <Zakim> Attendees were BrendanIAB, Curt_Tilmes, +1.518.276.aaaa, Luc, TomDN, dgarijo, GK, Satya_Sahoo, MacTed, SamCoppens, tlebo
17:58:32 <SamCoppens> SamCoppens has joined #prov
17:58:58 <SamCoppens> SamCoppens has left #prov
18:06:39 <Zakim> Zakim has left #prov
# SPECIAL MARKER FOR CHATSYNC.  DO NOT EDIT THIS LINE OR BELOW.  SRCLINESUSED=00000367