- Omar Bryan Khan
- Dean Allemang
- Mike Bennett
- Jim Logan
- Cory Casanave
- Jacobus Geluk
- Randy Coleman
- Pete Rivett
- Karthikeyan Giriloganathan
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.
5) For next week.
20170912 FIBO FPT
DA has a couple of things to discuss. Slave is working fine (for the headless thing). Question for JL or JG: The CL for the glossary generation requires you send username and password to TWC (by plain text!). Can I have a r/o account for this? JG you can register that user ID and pw in Jenkins under credentials in the home page. Those then inserted as environment variables in the [cut out] itself. JL can set up the account. What name? EDMC-Jenkins pw - let DA know it. What are the limits / rules for passwords? Also send DA the casing version for the username. Jim Logan: New account is created! User name: EDMCJenkins. Password: xxxyyyzz. Permissions: “Project Reviewer” of “FIBO-Master”
Next: Values ontology. Not clear what we originally agreed. Where is that information. Our process should not allow us to lose this information. In terms of workflow - it was not a policy everyone agreed since Elisa did not and was not on that call. How have we misplaced our version of the policy for he Values ontology. Per process Cory should have created a JIRA branch with his version of the Values ontology. In FLT do things only get agreed unanimously? DW: No. DW decides on policy given what people say, even if some people don't agree.
What was lost was one edit done to edit the bugs in CCM last year. CC did not know where to put the Values ontology in terms of the infrastructure. MB thinks we hand copied Cory's Values ontology in CCM with additions and changes, and this will have gone from CCM into the Feb 14 entry into GitHub. At this point it was used all and only for non Release ontologies where they have datatypes. There may have been a decision to extend the use of this to Release. That's where the memory hole is. Right now CCM uses Values for all datatypes.
Round trip - we probably need to never export these (both BFTs and Values) into CCM. References will be found. JL: the OWL Ingest does not include ontologies imported by ontologies that are directly imported - so it's possible to isolate Values and BFTs on ingest, as well as to not emit them. DA: there is no Cory branch in Git. CC is coordinating with MU on this. CC had not done the last step because he did not know where to put it. Where would this go.
Git: click on /fibo (that's the main EDM Council repository). Cory forks that. The new ontology goes into FND per its URL (FND/UtilitiesExt/Values/ ) Dean talks through how to use the fork per our process, using DW fork as an example on screen. What you do is create a new branch of your own fork.
CC we do need to revisit the notion of Business Facing Types. MB these were never intended for OWL, they were synonyms for datatypes, for business presentation. We need to nail down whether these are in and out. MB: for OMG submissions we would not be able to make this change in FIBO 1 since it would break backwards compatibility. So this is a FIBO 2 thing. PR: Bear in mind his detailed comments. For example, we already have DateTime classes. PR: we need to change the DateTime ontology to use the Values approach. MB that's a FIBO 2 thing. MB: meanwhile FCT leads need to know that they need to change references from TimeExt/Time to DatesAndTimes. This is completely separate to the matter of bringing DateTime into line with the Valules thing.
ACTION: MB go to the page that DA has started, for the Developers Guide, and put a note in to outline the above comment. Make a comment on this page to the effect of what needs to be done to bring Provisional ontologies that make references to BFTs so that they are up to date with current practices (and outline those practices). https://wiki.edmcouncil.org/display/INFRA/FIBO+Developer%27s+Guide
Introducing Karthikeyan Giriloganathan who is in Chennai. Working with BNY Mellon. Heads up the VOWL stuff internally. Karthik also knows OWL very well and is developing into an ontologist. Would recommend we have Karthik participate in the FIBO Process Team permanently. Karthik introduces himself. Has been working on semantics and ontologies during his Masters. Then worked on this in Atlanta. Then moved on to Java, working on transaction processing. Returned to India and has started working on this. Thesis was on converting OWL restrictions into RuleML and executing them. Now these are built in to StarDog and others, which were not available when he started.
Action item: TC to work on build errors. TC: there were more build issues; JG fixed the actual build parts but there is something in the Jenkins set-up relating to publishing or validation that wasn't there in the past and may have changed the build config since the last successful build. IT is not failing in the normal software parts of the build, don't know what it is, need JG to diagnose.
ACTION: JG will look into that build job. See notes in JIRA.
Is this critical for Sept 30? DA: Don't think so. This is for building the RDF Toolkit. Don't we have a workaround for distributing this that is up to date? TC: Yes. DA so we have a workaround but this is a technical debt to deal with for sustainability.
OK on Linked Data Fragments. OK: Still on target
Anything else needed from this meeting for DA work? DA no. Found the thing he was looking for. May need help from JL on the MD command line.
JG: On Linked Data Fragments, Karthik has also been working on this if Omar needs any help. OK detailed status - is picking up things on the Jenkins server for this. Can show us on Thursday.
JG diagram of what is running where - DW needs this for his re-write of the BTDM document. JG to supply.
Already on JIRA and Wiki. KK should get invites for these.
- Mike Bennett. go to the page that DA has started, for the Developers Guide, and put a note in to outline the above comment. Make a comment on this page to the effect of what needs to be done to bring Provisional ontologies that make references to BFTs so that they are up to date with current practices (and outline those practices). FIBO Developer's Guide
- Jacobus Geluk JG will look into that build job per TC. See notes in JIRA.