2017-06-15 Meeting notes
- Dennis Wisnosky
- Dean Allemang
Date
Attendees
Agenda
1) Where we are on our road map.
2) Open Action Items
3) JIRA Issues Review - https://jira.edmcouncil.org/projects/RDFKIT/issues
4) Todays content discussion.
Deans plan as a result off 20170612 FLT on 30 June content
CCM Glossary publishing - DA, JL
How do we deliver Linked Fragments - Omar
DN says we should remove history before 30 June pub. FPT and KT seems to think it should be there. If yes, then in a separate place so not to confuse casual users?
UML for 30 June - MB, CC, JL
5) For next week.
Proceedings:
20170615 FIBO FPT RDF ToolKit
https://spec.edmcouncil.org/fibo/ontology/master/latest/tree.html works now
MB: We need to show both a type hierarchy and a topic (ontology containment) hierarchy, based on recent feedback. DA has done a namespace hierarchy view.
Actions
ACTION: DA Link CCM glossary to spec glossary button (and call OWLDoc something else)
Based on CCM meeting recently. DA has all the pieces and hopes to have this done soon. There is a bug in MD that keep this from being convenient. We needed a frig list. DA will maintain by hand, a set of things needed (frigs) to keep this working. This is a known bug, but we didn't know that. This is why this action is open.
Action for people to look at what JG did on specsite on GitHub
JG had a meeting yesterday - what happened? No conclusions on implementation. JG: the content negotiation team - has been asked to come up with more user stories. JG also create a community that are able to interact with this.
NQuads output - action to run these by KT. DA: now has the NQuads available on spec - will send that to KT now for him to look at it.
Action: Matthew Horridge email. DA did forward this and TC has responded. TC has what he needs .
Action: instructions for on line and off line working. DA had to redo some of that since spec now contains both prod and dev stuff. It all worked in DA's own Dev structure,
Action on yesterday call was to open up the RDFKit wiki. This has ow been opened up.
Action: Dean change spec main page to markdown - not done. Currently in HTML. The spec main page is actually 2 main pages: spec top level and /fibo. Which one are we talking about? DW not the launching page but the landing page. DA the landing page now has the form stuff, does not know how to do that in Markdown. Can that be done (captcha etc.)? JG on the new specsite project the approach was to get some markdown fragments for the main body of the home page and the FIBO page, and show them on that new site The feedback form would be hidden under a button. DA cancel this task as it now conflicts.
Action: DA document modes for RDF ToolKit to operate in Git, e.g. stand-alone. DA low priority - get rid
Action: DA catalog files document. DA not done, low priority but will do. Action: DA not to get hit by a bus.
Action: DA + TC work out how RDF ToolKit enable FIBO developers to generate catalog files on demand. This is first part DA thinks now this is a JG action. Remaining text in this Action is a separate task.
Correct the About one not the Catalog one, was what JG will do.
DA as for the about files, we need a stand-alone capability for a FIBO Developer to be able to read in and create both catalog files and About (All) files. Recall discussion from Tuesday re Protege messing with files. Current status: can only happen in Publish context, not stand-alone. JG reiterates the current state.
DA this is great for users, but developers need to be able to pull sources and have a catalog file that works on those. Either they push the catalog into Git which is not a good idea. Or have them be able to create it. DA currently uses a script. DA: Leave this open, DA will prioritize this action,
Action: write what JG says will work. Context: DA knows what this is about. Has done the 'try it' part. Did what techs always do with documentation, had someone teach him how to do it so he can write it down. But has not written that down while still able to recall what it was not to know this. Will do this.
Action (DW): JL did a good job on the napkin slide. DW changed this to a PowerPoint slide. Needs to change to less CCM specific. DW will do this.
Action: Elie update FIBO Product ontology. DA has done part of this - for the pieces that are done as part of the ontologies themselves (the maturity levels) and has socialized it with MB as chair of FND FCT. That went into production last night. DA - this is working fine. Does not influence what Elie is doing on the other stuff about other product lines and their content. EAL has a couple of actions on this, and domains and modules per Mike A document. Updates as discussed with JG and DA last month. Hope to meet Jun 30 deadline. DA we have hard coded a good deal of the product stuff as a back-stop for June 30 Would be happy to migrate in a functionality neutral way but that can wait to after June 30. But EAL can still work on this stuff but DA has the safety net in place. EAL expects to make progress in the background. DA: Elie should look at the current stuff as hard coded, as a reference point for this work. Is this in the INFRA repository or the RDFKit repository? DA look in the repo called fibo-infra. There are scripts there that are essentially hacks. The branch called master has all the things that are tested and working. fibo-infra/jenkins/bin/publishhedFIBO. There is a thousand line hack. These pieces will eventually be driven by the other thing.
Next actions are JG: Provide resource to generate user specific FIBO content. JG doing that himself - trying to motivate colleagues to work on this. Have their own UI for the knowledge graph. Using the same technology stack as we are using. Expect some cross pollination from them.
Action: JG commit what he has to github for others to use. Is this what we saw Tuesday? JG yes. This is the new repository in the EDMCouncil called specsite. JG will work in parallel in specsite with what is in fibo-infra.
JG describes how this works to find all the files and so on. Once specsite does that it will be moved over. Also needs to refactor the feedback form. Will then have one codebase.
JG to document the architecture and ideas behind the new specsite thing - so people can add their own components. All the objects in our artifact ontology should have their own page. Would include a module taxonomy. Also taxonomy of the Business Domains. Maybe FIBO-Vocabulary is one such. Support multiple views into the artifacts we have. May have new taxonomies of new kinds of artifacts in the future e.g. some they are working on at BNY (business capability taxonomy). Architecture allows for adding these. When you click on a concept in any of these, the concept generates its own page with its own URL on the client site. Then when you land on a business term - what components do we have that can view a business term e.g. CCM components. Incentivizes vendors to create such components. End users can locally determine what components they like to see, and save that as a setting locally.
Action: establish new https thing. This is now called Fragments not Linked Data. That exists (Omar did it). https://fragments/edmcouncil.org
Action: JG Dev architecture picture. Open
Action: MB did this. Will post this to close the action.
Action: OK (next 3) OK to as Ghent people something – done. Second OK action done. 3rd OK action -
This is to pick up artifacts from Jenkins server, do something with the generator, do something with the publisher. No one knows.
Action: fragments server (done as above); 2nd part of this action not known
PR action to send something to JG - did he do that? Not done. Was about markdown
Next action: perpetual candidate hygiene policies. DW this is in JIRA so will remove the action on checking this is in JIRA correctly.
New actions?
JG - the work we did with DN about ontology maturity. This is now a proteted page on RDFTookit. Would like everyone on the call to look and comment. This is also in JIRA.
Next: June 30 progress and pieces.
Glossary: this is OK using the CCM NL glossary. DN may do OWLDoc also.
UML/OWL (CC/JL/MB). This leaves the work DA is doing with respect to EK list on the CP. DA will take on the OWLDoc in synch with the Glossary etc. The big issue is the round tripping with CCM for diagrams. DA and MB discussed the detailed requirements for unifying the single master version, this week.
DA/MB actions to rearrange the CCM package structure in CCM to it is more rational. Repository in TWC called FIBO-Master has a branch called Jims Test. DA sees Jims Test was successful, Want to turn that with the new master. Then the stuff we talked about on Monday will put in the module structures we need to put in.
ACTION: MB to verify against checklist for the origin file. Coordination needed but the pieces all work.
Back to Agenda: what's next? The History Thing: This relates only to FIBO-V. DA can do this as part of the FIBO-V publication. Involves removing one line of code.
ACTION: DA: UML for 30 June after verifying round tripping. Does MB need anything from CC and JL? MB no.
Decisions:
Action items
- Dean Allemang Link CCM glossary to spec glossary button (and call OWLDoc something else)
- Mike Bennett verify against checklist for the origin file. Coordination needed but the pieces all work in Jims Test.
- Dean Allemang UML for 30 June after verifying round tripping. Does MB need anything from CC and JL? MB no.
- .