Wikidata:Property proposal/Landgate Object ID

Landgate Object ID

Originally proposed at Wikidata:Property proposal/Place

   Withdrawn

Description	Landgate (Western Australian Land Information Authority)
Data type	External identifier
Domain	electoral district of Western Australia (Q5356189), local government area of Western Australia (Q55557858), regional park (Q6063204), locality (Q3257686), town (Q3957), suburb (Q188509), suburb/locality (Q7632426)
Example 1	Landsdale (Q85758876) → 20/1
Example 2	Shire of Wiluna (Q2051063) → 14/1
Example 3	Wooroloo Regional Park (Q21972405) → 4/1
Example 4	Rocklea (Q55448888) → 16/1
Example 5	Marble Bar (Q1025115) → 15/1
External links	Use in sister projects: [ar] • [de] • [en] • [es] • [fr] • [he] • [it] • [ja] • [ko] • [nl] • [pl] • [pt] • [ru] • [sv] • [vi] • [zh] • [commons] • [species] • [wd] • [en.wikt] • [fr.wikt].
Planned use	Identify State electorates, LGAs, regional parks, localities, towns and suburbs
Number of IDs in source	There is currently 137 Local government areas of Western Australia, 59 Electoral districts of Western Australia, 14 Regional Parks and an undetermined number of localities, suburbs and towns in Western Australia
Expected completeness	eventually complete (Q21873974)
Implied notability	Landgate (Q6484527)
Formatter URL	https://services.slip.wa.gov.au/public/rest/services/SLIP_Public_Services/Boundaries/MapServer/$1

Motivation

A recent discussion on the English Wikipedia highlighted the problem that Electoral districts LGAs, towns and localities in Western Australia often share a common name. Especially with the latter two, identification of what an item on Wikidata is about can sometimes be ambiguous. As an added benefit, where applicable, this identifier also provides a reference for land area covered. As can be seen in this list, further uses past the examples above is possible. Calistemon (talk) 08:07, 9 January 2023 (UTC)

Discussion

• Thierry Caro 99of9 Canley Mattinbgn Samwilson JarrahTree ScottDavis Oronsay Pru.mitchell Bidgee Harryboyles Dhx1 SJK Nickw25 Danadl MargaretRDonald Nicolekearney Calistemon Eothan Evolution and evolvability REMerrick Atroche The-Pope Bjenks hfordsa SHB2000 AmandaSLawrence Chris.sherlock3
  Notified participants of WikiProject Australia --99of9 (talk) 22:39, 9 January 2023 (UTC)
•  Question @99of9: I'm very keen to have a Landgate ID! Thanks for looking into this. It looks like this is a combination of the layer ID and the feature object ID. Are these stable, do you know? Are they used anywhere to refer to these entities? I guess I'm just wondering if these are incidental internal database IDs that are liable to change at some point. If no one's using them as identifiers, I don't know if we should. Also, the features are discrete geometries and don't represent the entities themselves (where there are multiple geometries per entity) — for example, Wooroloo is nine features: 4/1, 4/2, …, 4/9. Would we add all of those as values? Sam Wilson 22:54, 9 January 2023 (UTC)

  @Calistemon: Sorry, I meant to ping you! Sam Wilson 22:55, 9 January 2023 (UTC)

•  Support thanks for offering this. --99of9 (talk) 22:46, 9 January 2023 (UTC)

• I defer to Dhx1 on this. --99of9 (talk) 00:20, 10 January 2023 (UTC)

•  Oppose I see that boundary_id and feature_number fields are also present--are these more stable than using the ObjectID field built into ArcGIS and should one of these fields be used instead? If Landgate were to switch from ArcGIS products to an alternative, the ObjectID field would seemingly be lost. If keeping with layerid/objectid, this would be generic for all ArcGIS geodatabases and thus this property proposal should instead be made generic too? The property name would be "Geodatabase ObjectID" (integer quantity) with a new qualifier property of "Geodatabase Layer Name" (monolingual text) and perhaps another "URL" qualifier or "Geodatabase" qualifier to denote where the geodatabase resides. ArcGIS state that layer IDs constantly change based on default drawing order unless a user has specifically changed a setting to allow manual assignment of static identifiers to each layer. Behind the scenes, a feature layer is generally a table in a database, and it therefore seems likely the table has a string identifier rather than an integer identifier. I'm not sure however if a layer name and table name correspond with each other. --Dhx1 (talk) 23:37, 9 January 2023 (UTC)
• I’m going to see if I can get Andrew Owens to respond. He is hugely involved in Commons but has only just really found out about Wikidata but he is an extremely active participant in WA and knows all the ins-and-outs of WA and has an incredible knowledge of electoral and LGA boundaries. - Chris.sherlock3 (talk) 10:36, 10 January 2023 (UTC)

  Actually, just spoke to Andrew, he was at (and helped organise) the inaugural 2016 Wiki conference where Wikidata was explained in detail. Apologies to Andrew! - Chris.sherlock3 (talk) 10:52, 10 January 2023 (UTC)

•  Comment Unfortunately the discussion above is far too technical for my understanding and I, as the nominator, can't provide any clarification whatsoever. Thanks for the opinions offered however as it is clearly preferable to kill off the proposal now rather than make all the effort of adding identifiers that may become dead links in the near future. Calistemon (talk) 14:51, 11 January 2023 (UTC)