Meeting minutes
Admin
jo: minutes of last meeting
… Ben - call a meeting to discuss FESE terms
ben: still to do
markd: came across investor.gov which has some terms
… also reaching out to FISD members
jo: actions on caspar - done - will come back to later
Resolution: Accept minutes of last meeting
Report from FISD Webinar
link to webinar: https://
markd: all went well
ben: feedback was positive, 90 mins was the right amount of time
… people seemed to stay for the whole time
… level of discussion was good and candour of the panel was remarked upon
… would be good to have a glide path
… discussion on adoption was interesting
… had a discussion with a bank wondering if they wanted to be an innovator or if they were happy following
… can we avoid this?
markd: points raised about under-representation of exchanges was important
… call to action needs to be top of mind
Versioning
caspar: Nigel and I sent over high level recommendations and points for discussion
… ideally use semantic versioning ... also support in OWL for deprecating terms
… difficult to test for compatibility ... difficult to test for compatibility at the domain level
… in some cases action is backwards compatible ... but not in others, for example in a prohibition
… looking for prior art, found a couple of papers ... but they don't seem to be complete
… diff needs to done at semantic level not document level
… if version is 0 then doesn't matter about breaking changes
ben: let's say you have version 1.0.1 ...
… if someone is using 2.... then you have a problem
… if you have 1.0.2 then you can carry on
… problem lies in the number in the middle
… take the point on addition of prohibition or constraint
… is there a way of calling out if an unknown terms was found
… so that would be fail gracefully
caspar: like the idea of a workflow
jo: could we not say that we'll edit the document only according to rules that are predictable
caspar: most minor number - fixes etc.
… middle number is used for non-breaking incremental changes
ben: if we changed the definition of duty that would be catastrophic
… but if we add a term then we can localise the problem.
jo: change version number differently according to the nature of the addition, breaking and non-breaking additions
caspar: if you hit a term and you don't understand it, then you fail if the rule is to evaluate all the prohibitions
jo: implementation rules make perfect sense as long as we think it rmeains robust
ben: if this works we can maintain simple versioning system to help adoption
jo: anyone else we can reach out to?
caspar: no, academic work is detached
ben: I'll ask someone who is an ontology expert
… next step is to write it in to the standard
… will liaise with nigel and caspar to reflect their work
Action: Ben to draft versioning text
markd: differences between exchanges - can we accommodate the differences in derived data
ben: we are taking the lego block approach - we have some properties ...
… which in combination will hopefully explain the differences between exchanges
… but if they come up with new models we may need additional terms
Identifiers
caspar: sent a long email, crux is that URN NID is sufficient for every purpose. Some concern around versioning. e.g. currency codes are static LEIs grow over time
… looks like all roads lead to ISIN ... so if you will allow multiple different types of standards for IDs then there could be a problem
… seems like there is always a temporal component, whereby the id is checked as of the dat of the contract
jo: so there is a bi-temporal model at work here
ben: having a short-hand whereby the standard is at the date of the contract
… is a neat solution
caspar: with a perm url the problem doesn't arise
… unbounded sets will need to be a permid
ben: I can't believe ISO codes would reallocate, such a naff thing to do
… all the codes we are interested in would be ISO codes
Caspar: I'll check for what the deprecation policy is for any of the referenced standards
ben: good news is that we can referenced via URN
Action: Caspar to look into deprecation policy of ISO standards
ben: I'll put the prefixes into the standard
Action: Ben to put the identifier prefixes from Caspar into the standard
ben: M49 set is superset of ISO - do we lose anything if we use it
caspar: couple of weird quirks, like gb for uk
Action: Caspar to verify that M49 codes with a URN
GLEIF
ben: LEIs can be dereferenced, they are enforced by regulators, anyone in the market will have one
… GLEIF will create a URN, and are applying for that schema right now
… we connected them with W3C who hopefully will help
… so we can make the recommendation to use LEI and to use URN syntax to do so
Feedback on Document
ben: feedback is drying up, and attendance is decreasing, does this mean that people are happy enough ...
… I will write a new version, tidy a few things up
… so shall we move on to implementation and interoperability ...
laura: makes sense to me
… are there any parts of the standard that need more review, happy to help with that
… transition to monthly, good to know, other people then might be involved
(polling floor for views)
michelle: seems like a good point to move forward
mark: what are the targets we are trying to hit?
… cadence should fit wheat we are trying to achieve and when by
ben: good point, let's focus on what we mean by interoperability, and specify some projects
ben: if I see some areas that need further review I'll contact Laura/Michelle/Mark
… versioning and IDs need to go on
… I'll try to do that in the next fortnight
AOB
jo: hearing none, meeting closed
rrsagent: draft minutes