Observation Types

[KathiSchleidt]

Looking for O&M Observation Types, I come up with the following options via the following paths

1. Starting with OGC Definitions, (strangely enough NOT the page that http://www.opengis.net/def resolves to), where the entries are marked as "Vocabulary", I have 2 options for observation types:

• OM observation types: http://www.opengis.net/def/observationType - an interesting list of anything to do with O&M - according to this resource, Samples are observation types. We've tried a few times to have this error fixed, but yet to succeed.
• Observation type: http://www.opengis.net/def/observation-type

2. Starting with the base page, so what http://www.opengis.net/def actually resolves to, I again can access observation types under the ConceptScheme Hierarchy (interesting is that I cannot find the Vocabulary page in this hierarchy), I also find:

• OM observation types: http://www.opengis.net/def/observationType/ - same link as above, but with a trailing /, now a Collection, providing the following:
  • O&M Observation types: http://www.opengis.net/def/observationType/OGC-OM/ - Another Collection, containing:
    • O&M Observation types 2.0: http://www.opengis.net/def/observationType/OGC-OM/2.0/ - here we find some observation types!
  • Observations and Measurements definitions: http://www.opengis.net/def/ogc-om/ - A Collection of observation and sample types
  • Observations and Measurements XML definitions: http://www.opengis.net/def/ogc-omxml/ - A collection with three other observation types initially missed
  • OMXML Observation types: http://www.opengis.net/def/observationType/OGC-OMXML/ - a collection containing:
    • OMXML Observation types 2.0: http://www.opengis.net/def/observationType/OGC-OMXML/2.0/ - another collection with the three missed types

What one would really need:

• Observation Types: All Observation types. The 3 extended types could to my view be included. While not formalized in the ISO Spec (reason for exclusion to my understanding), having these additions in the same namespace as they pertain to OGC types would not cause issues with ISO, would make things more consistent
• Sample Types: currently missing, in some cases erroneously grouped under Observation Types. Unclear where this came from, error is in the original ttl available on GitHub, has been addressed a few times, but always gotten lost :(
• O&M Definitions: Not sure we need such a grouping if we have observation and sample types cleanly sorted

ToDo:

• Fix the Sample isA Observation
• Decide on a naming scheme (pretty impressed how many versions of camel-kebab-case are possible!)
• Maintain all existing bits just in case somebody is using them, but mark them as deprecated and a link to the cleaned scheme

[rob-metalinkage]

A lot of this content was manually curated in the past - and has not (intentionally) been changed so its totally appropriate for the SWG to take over quality control and updates!

There is another level of complexity around equivalent forms - "observationType" vs "observation-type" and version equivalences. At this stage there is no policy guidance on this and so the Definitions Server just exposes these as best it can - but it does need a discussion and clarify of both policy and display requirements. I've been dreading this :-) I'm guessing there was historical inconsistency in use and the problem has now been propagated to a complicated and ill-defined set of upstream user requirements.

The TTL source file [1] can be updated via a pull request (or email to me) any time, and I can propagate the updates to defs-dev for review then to production.

[1] https://github.com/opengeospatial/NamingAuthority/blob/master/definitions/conceptschemes/om.ttl

[KathiSchleidt]

@rob-metalinkage while I can try and sort om.ttl, we'd also need some sort of battle plan on how to deal with the diversity of O&M resources - once we've agreed on the true set of resources, how to deprecate the myriads of pseudo-twins!

And many thanks for admitting that you've been keeping your head in the sand on this issue, confirms my impression!

Btw - is anybody aware of an open source editor with which I can sanely edit the ttl? Protege doesn't cut it, I can't afford TopBraid, WebVOWL makes me dizzy, and somehow Notepad++ (my go-to when all else fails) is missing essential tooling! ;)

[dr-shorthair]

Vs-code with the Stardog plugin gives good syntax colouring at least

[ghobona]

@KathiSchleidt The tasks on the ToDo list above are appropriate. The O&M SWG is the expert group for the O&M standard, therefore, I will move this GitHub Issue to the O&M SWG's repository. I am anticipating that the O&M SWG will create a branch or fork of this repo and make the proposed modifications to om.ttl. Once the O&M SWG has completed the proposed modifications, the OGC-NA can review the proposed modifications and either approve them or request clarifications/changes.