Wikidata:Property proposal/HMS

HMS ship edit

Originally proposed at Wikidata:Property proposal/Transportation

   Not done
Descriptionidentifier for a ship of the British Royal Navy, qualify with "start date" (P580) and "end date" (P582) when it was in use. Vessel may have existed before or afterwards under a different identifier.
Data typeExternal identifier
Example 1HMS Scipio (Q5634244)Scipio qualified with start time (P580): 1782, end time (P582): 1798
Example 2Scipio (Q5317432)Scipio qualified with start time (P580); 1807, end time (P582): 1808
Example 3HMS Resolution (Q1898494)Drake qualified with start time (P580); 1770, end time (P582): 1771
Example 4HMS Resolution (Q1898494)Resolution qualified with start time (P580); 1771
Planned useexpand Help:Import NBD from enwikisource/lists/ships by creating items for 18th/19th century vessels (see Wikidata_talk:WikiProject_Ships#HMS_for_A_Naval_Biographical_Dictionary_(Q16055052))
Number of IDs in source13,000+ items
Distinct-values constraintyes

Motivation edit

The assumption is that the above identifies uniquely a ship for a given point in time. Soon phab:T277855 with allow to use the distinct value constraint for these.

See template above for the planned use.--- Jura 12:17, 30 August 2021 (UTC)[reply]

  Notified participants of WikiProject Ships (Add your motivation for this property here.) --- Jura 12:17, 30 August 2021 (UTC)[reply]

Discussion edit

  • I don't think we should start introducing properties like HMS or USS, HCMS, HNZMS for that matter. We can already use Property:P1448 for the duration of an official name. Likewise, if a ship changes its name it will have different links to Wikipedia and Commons requiring a new WD item anyway, so there is no need for a dated HMS property. De728631 (talk) 21:38, 30 August 2021 (UTC)[reply]
    • @De728631: It's a stable identifier for 13000+ items. There is no way to do the unique value check on the alternative you suggest. The proposals has nothing to to with sitelinks to Wikipedia and/or Commmons. I don't really understand what make you think that. --- Jura 22:01, 30 August 2021 (UTC)[reply]
      • Well, I read you proposal as a means to identify the period of a ship being named HMS Something. However, if it is intended to identify Royal Navy ships in general, can't we use "operator: Royal Navy" instead? That together with any subclass of "ship" should also be unique. I'm just trying to avoid redundancies. De728631 (talk) 22:10, 30 August 2021 (UTC)[reply]
        • The "something" part is the unique identifier for these. The idea is to have reliable coverage of HMS vessels, not to identify the operator (who is known). --- Jura 22:32, 30 August 2021 (UTC)[reply]
          • Alright, I think I'm getting what you mean. So, on the item side this could look like
HMS: Scipio
start time: 1807
end time: 1808
with constraints of "operator = Royal Navy" and ideally subclass of "ship". De728631 (talk) 22:56, 30 August 2021 (UTC)[reply]
Ideally with several other statements, but yes. From Help:Import NBD from enwikisource/lists/ships I got the impression that much is still to do. --- Jura 23:03, 30 August 2021 (UTC)[reply]
  •   Oppose This looks like a quite messy way of structuring something that really should be a list object instead. /ℇsquilo 15:35, 31 August 2021 (UTC)[reply]
    • @Esquilo: can you elaborate? What part do you consider "messy" and what is a "list object" in Wikidata? Never heared of, Wikidata uses queries on properties, such as the one proposed here. --- Jura 15:59, 31 August 2021 (UTC)[reply]
      • The intention seems to be a way to create a list of ships that has had the name "HMS Scipio", for example, by adding a new property to every ship that has who has borne that name. That is messy. Better then to create a list object "list of ships named HMS Scipio" and add the ships to that list with start time (P580), end time (P582) or any other qualifier you like. /ℇsquilo 16:09, 31 August 2021 (UTC)[reply]
    • @Esquilo: I don't see how unique identifiers are ever messy, it's just that some are more useful to some people than to others. If phab:T277855 works out as planned, the new feature should simplify keeping it consistent. The approach you suggest may be an alternative, but adding lists on Wikidata items isn't really popular, nor is it necessarily less "messy" or easier to maintain. In any case, it doesn't seem to have been done and the overall situation for HMS is somewhat "suboptimal". If you plan to work on the suggested alternative, I'd obviously don't want to stop you. --- Jura 18:14, 31 August 2021 (UTC)[reply]
      • Several reasons. First of all, it is not unique. It is just unique in a certain timeframe and for a certain operator. Royal Navy is not the only navy that use the prefix HMS. The Swedish navy also use it. Sweden is not a part of the commonwealth, so they do not use any additional letters like HMAS or HMNZS. (HMSwS is a bastard that is not recognized in Sweden). Secondly, since this property is specific for Royal Navy, it does not cover the same need for other navies. Thirdly, ship items are simply the wrong place to store this information. /ℇsquilo 18:28, 31 August 2021 (UTC)[reply]
    • Have a look at phab:T277855. The new uniqueness constraint should address point 1. I inserted "British" in the description in case it wasn't clear. It's in the nature of identifier properties that they don't cover everything, but 13000 is a good set to start with. Depending on how it goes we can try to do something similar for other navies (or switch to whatever superior solution people come up with there). I wonder where we should store it if not in ship items. --- Jura 18:35, 31 August 2021 (UTC)[reply]
    • @Esquilo: DANFS ship ID (P7910) and Norwegian war sailor register ship-ID (P5156) seem to have a similarly limited scope. --- Jura 11:55, 1 September 2021 (UTC)[reply]