Editorial progress meeting on serialization specifications (Conference Call) | |
Name | Editorial progress meeting on serialization specifications (Conference Call) |
Time | Monday, 07 August 2023, 10:30am to 12:30pm UTC
(Monday, 07 August 2023, 10:30am to 12:30pm UTC) |
Description | https://www.oasis-open.org/apps/org/workgroup/lexidma/members/action_item.php?action_item_id=3946 |
Minutes | Rollcall 4/7 voters, Miloš to re-gain voting right at the end of the meeting Approve minutes from 31st July https://markmail.org/message/3pzh7i3sxvycivk5 https://www.oasis-open.org/apps/org/workgroup/lexidma/email/archives/202307/msg00009.html Michal seconded, no objections
Core and modules terminology change: Use properties instead of Contents Make it consistent in serialization references
Core change Consensus URI is REQUIRED at lexicographicResource To be implemented in serializations and examples Michal to implement the above
Background to the above decision: Making REQUIRED either Title or URI would be more complicated to express in serializations and potentially confusing for implementers. We don't want to complicate the top level object URI is flexible and can be used for semantic versioning etc. while Title should be kept human readable and thus optional (not critical for exchange)
Discussion: WRT top level URI use: Consider use of the DICT protocol for exchange and URIs? https://en.wikipedia.org/wiki/DICT
WRT potentially standardizing exchange of fragments Consider defining valid fragments? Define optional roots in the core data model? Define a serialization independent fragid mechanism?
If we go for this we need a proposal for the addressing mechanism and exchangeable chunk rules (including omission/projection/linkage rules) by the meeting of 16th August Action: Miloš
Background to the above discussion: Exchanging whole dictionaries would be suboptimal in many use cases. Is considered risky by many potential implementers. Avoid scope creep towards a processing format At the very least we should design an out of scope statement, ideally we should legalize fragment exchange for already implemented use case
Review serialization specs related actions SQL Michal to clean up this week
RDF John to commit this week
NVH Miloš to update according to core terminology agreements
JSON Michal to reformulate: Do not use the term class, strictly use JSON terminology object, property, array (also internally)
Agreed meeting time in the week of 14th August Wed 16th August 1200 UTC dF to schedule AOB Adjourned at 1208 UTC
|
Agenda | Rollcall Approve minutes from 31st July https://markmail.org/message/3pzh7i3sxvycivk5
Review serialization specs related actions SQL RDF NVH JSON Agree meeting time in the week of 14th August AOB Adjourn no later than 1230 UTC |
Submitter | Dr. David Filip |
Group | OASIS Lexicographic Infrastructure Data Model and API (LEXIDMA) TC |
Access | This event is visible to OASIS Lexicographic Infrastructure Data Model and API (LEXIDMA) TC and shared with
|