Wikidata:Property proposal/fleet or registration number

fleet or registration number

Done: fleet or registration number (P2802) (Talk and documentation)

Description	the fleet number or registration of the vehicle
Data type	String
Template parameter	"fleetnumbers" in en:template:Infobox Locomotive
Domain	vehicles, events (accidents)
Example	LNER Class A4 4488 “Union of South Africa” (Q6459251) → 4488 (qualifier: issued by (P2378) London and North Eastern Railway (Q1092839)) Eckwersheim derailment (Q21482333) item operated (P121) SNCF TGV Duplex (Q928892) (this property as a qualifier) → 744 FAB 1 (Q5424292) → FAB 1 Bluesmobile (Q1117762) BDR 529 (qualifier: issued by (P2378) Illinois Department of Transportation (Q4925114))
Source	external reference, Wikipedia list article, etc.

Motivation

The number of an individual vehicle is often a noted fact, especially for rail locomotives. There are two use cases for this I can think of, the first is on items about individual vehicles and the second is on items about notable events (e.g. accidents) where that vehicle is significantly involved (where it would be used as a qualifier to item operated (P121). aircraft registration (P426) exists for aircraft and is used in a broadly similar way, and IMO ship number (P458) exists for ships but I'm don't that merging or generalising those would be a good idea - indeed it might be better split this proposal into separate properties for rail vehicles and road vehicles? Thryduulf (talk: local | en.wp | en.wikt) 15:28, 17 November 2015 (UTC)[reply]

Discussion

Support We need this as a super-property for such information, as both aircraft registration (P426) and IMO ship number (P458) are specific and should be left as is. Josh Baumgartner (talk) 20:17, 18 November 2015 (UTC)[reply]
Support Danrok (talk) 22:42, 23 November 2015 (UTC)[reply]
@Danrok: Do you think you can merge your proposal above code letters with this one ? Snipre (talk) 13:43, 26 November 2015 (UTC)[reply]

@Snipre: That would prevent validation of values. Code letters are always letters as far as I know. Plus, there's also another one for ships called Official Numbers [1], a system which came after code letters, and before IMO ship numbers. These unique IDs are very important because they allow for the finding of a ship on say the Lloyd's Register, and other official paperwork, archives, etc. I need to read-up more on official numbers before making a proposal on that. Danrok (talk) 13:56, 26 November 2015 (UTC)[reply]

After no objections for three months I'm marking this as ready. Thryduulf (talk: local | en.wp | en.wikt) 13:33, 25 February 2016 (UTC)[reply]

@Thryduulf: Not done - use inventory number (P217) ("identifier for a physical object... in a collection") for fleet numbers. No prejudice against a separate property, if that is not deemed appropriate, for registration numbers. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:19, 7 March 2016 (UTC)[reply]
- @Pigsonthewing: Why have you set this to "not done" (which smacks of supervoting) rather than putting your opinion as an oppose? I disagree that inventory number (P217) is relevant here as using it in the circumstances intended would violate (almost?) every constraint on that property! Thryduulf (talk: local | en.wp | en.wikt) 23:43, 7 March 2016 (UTC) (resigning to fix ping. Thryduulf (talk: local | en.wp | en.wikt) 11:33, 8 March 2016 (UTC))[reply]
- Because, as I noted, the property already exists. If the constraints are wrong, fix them. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:25, 8 March 2016 (UTC)[reply]
  - But the property doesn't exist! The constraints for inventory number (P217) are not wrong, they exist to enable validation of the accession numbers for items in a GLAM collection and are correct for that job. That is a completely different concept to fleet numbers and registration numbers of vehicles (which may or may not be part of a collection, and when they are will likely have an accession number as well as a fleet and/or registration number). Thryduulf (talk: local | en.wp | en.wikt) 20:11, 8 March 2016 (UTC)[reply]
  - @pigsonthewing: as my above comment doesn't seem to have been spotted. Thryduulf (talk: local | en.wp | en.wikt) 12:44, 11 March 2016 (UTC)[reply]
    - I refer you to my previous response, and my initial comment, by each of which I stand. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:03, 11 March 2016 (UTC)[reply]
      - @pigsonthewing: So where do we go from here? I completely disagree with you that inventory number (P217) is suitable here. @Danrok, Snipre, Joshbaumgartner: Do you have any additional comments? Thryduulf (talk: local | en.wp | en.wikt) 20:59, 11 March 2016 (UTC)[reply]
        [2] Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:00, 11 March 2016 (UTC)[reply]
        Note I have undone this change as there is clearly no current consensus that it is appropriate. Thryduulf (talk: local | en.wp | en.wikt) 17:23, 14 March 2016 (UTC)[reply]
        Reverting valid edits (adding an "identifier for a physical object... in a collection", which matches the property description perfectly) is not a way for you to overturn this closure. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:45, 14 March 2016 (UTC)[reply]
        Except it is not an object in a collection in the sense that p217 means, so the edit is not valid. Thryduulf (talk: local | en.wp | en.wikt) 12:36, 16 March 2016 (UTC)[reply]
        You are imagining the meaning; wrongly. But as I said below, this is not the place to debate such matters; and the use of P217 is now being discussed on Project chat. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:48, 16 March 2016 (UTC)[reply]
@Thryduulf, pigsonthewing: In no way is a vehicle's registration number the same as a museum's inventory number. That is definitely not how this should be done. Boats and planes are held in collections, so clearly the two would become muddled up. Danrok (talk) 21:24, 11 March 2016 (UTC)[reply]
- Which is why I wrote "No prejudice against a separate property, if that is not deemed appropriate, for registration numbers". inventory number (P217) is qualified by the name of the collection to which the inventory number refers; and thus multiple values may be stored. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:57, 11 March 2016 (UTC)[reply]
  - @pigsonthewing, Danrok: a fleet number is not conceptually different to a registration number (it's a unique identifier for a vehicle), but both are very different to accession numbers and there is no collection name with which you can qualify a fleet number if the vehicle is not in a collection. Thryduulf (talk: local | en.wp | en.wikt) 17:20, 14 March 2016 (UTC)[reply]
    - You're being Anglo-centric. In some countries, the registration number changes with the registered owner, or yearly. However, this is not the place to debate such matters. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:45, 14 March 2016 (UTC)[reply]
      - eh? Accession numbers are stable and don't change yearly, whereas fleet and registration numbers can - this is another argument against inventory number (P217)'s suitability. Thryduulf (talk: local | en.wp | en.wikt) 12:36, 16 March 2016 (UTC)[reply]
        Accession numbers can and do change; we have start- and end- date properties for use as qualifiers in such cases. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:48, 16 March 2016 (UTC)[reply]
I think that it is important that inventory number (P217) is only used for a collection's inventory reference, and nothing else. It's an ID which unambiguously pinpoints an item in a collection which could be very large, and contain very similar items. If we start storing any "other marks" in inventory number (P217) that just creates work for anyone using the data, they will have to code some method of separating values stored in inventory number (P217). It also increases the possibility of mistakes and mix-ups. That's just making things more complicated than they need to be, with no benefit that I can see. Danrok (talk) 18:32, 14 March 2016 (UTC)[reply]
- I agree entirely. The fleet number of a vehicle is "a collection's inventory reference" and it does "unambiguously pinpoint an item in a collection which could be very large, and contain very similar items". That is precisely what it is for. We already have "method of separating values stored in 217": qualify them with the name of the collection or issuing body; and dates if relevant. However, this is not the place to debate such matters. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:45, 14 March 2016 (UTC)[reply]
  - It is clear from the description, constraints and proposal for inventory number (P217) that it is only intended to be used for GLAM collections. A vehicle in a museum may have one of these, an operational vehicle does not as it is not in a GLAM collection (or in most cases in any collection at all). A registration number issued by, e.g. the DVLA is a conceptually very, very different thing in both form and purpose to an accession number issued by e.g. the National Motor Museum. Vehicles may have multiple concurrent fleet and registration numbers, but only one accession or inventory number is current at any one time. Thryduulf (talk: local | en.wp | en.wikt) 12:36, 16 March 2016 (UTC)[reply]
    - You are, sadly and inexplicably, making things up. There is nothing in the description which mentions, let alone limits this to, GLAMs. Anything with fleet number is part of a collection; that is what the "fleet" is. As I said early on, constraints which are wrong can be - and should be - changed. A registration number issued by the DVLA is a conceptually very, very different thing in both form and purpose to a fleet number; which is why, I repeat, I wrote "No prejudice against a separate property, if that is not deemed appropriate, for registration numbers". And once again, this is not the place to debate such matters; and the use of P217 is now being discussed on Project chat. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:03, 16 March 2016 (UTC)[reply]
Support Useful property. To me, inventory number (P217) is conceptually something completely different. --Srittau (talk) 14:18, 16 March 2016 (UTC)[reply]
Support (though I would simplify the label to "fleet number" and clarify that it should be owner/operator-based, not an external registration number.) It seems to me Andy's claim rests upon an assumption about the meaning of the word "collection". What does "collection" in the context of inventory number (P217) signify? Well, it's clear from the discussion page and usage that it refers specifically to the property collection (P195), which has corresponding item public collection (Q2982955), described as a collection of "works", and a subclass of art collection (Q7328910). The set of locomotives owned or operated by a particular railroad, while technically a "collection" in the sense of something that contains multiple items, does not to me by any stretch qualify as either a public collection (Q2982955) or art collection (Q7328910). I also am quite concerned that Andy acted to close the discussion rather than simply add an "oppose" comment. If there is a significant concern about the creation of duplicate properties let's maybe start an RFC with concrete proposals for how to deal with it. I don't think this qualifies as a duplicate in any case. ArthurPSmith (talk) 17:57, 17 March 2016 (UTC)[reply]
- @ArthurPSmith: I'm OK with this being made "fleet number" only, but in that case we will also need a separate property for registration number. Thryduulf (talk: local | en.wp | en.wikt) 19:36, 27 March 2016 (UTC)[reply]
Comment The linked discussion on Project chat (now archived at Wikidata:Project chat/Archive/2016/03#P217 - should it be broadened?) was not explicitly closed, but does not seem to my (biased) eyes as concluding in favour of broadening inventory number (P217). Accordingly I think the objections to this proposal with that reasoning can be dismissed when judging whether this property has gained consensus or not (currently I'm seeing a likely consensus for either combined or separate properties, but no consensus between the two options). Thryduulf (talk: local | en.wp | en.wikt) 19:36, 27 March 2016 (UTC)[reply]
- There is no consensus for your proposal, in either form (there may be for a "registration number" property, as I have already indicated) - since, as I demonstrate irrefutably above, it is flawed and unnecessary for a fleet number (which is equally irrefutably an "identifier for a physical object... in a collection"), with P217 being perfectly adequate for that purpose, and the objections to using that being based on a false premise otherwise. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:06, 27 March 2016 (UTC)[reply]
  - I think we should let an independent person determine the consensus regarding this, but when (almost) nobody has agreed with you regarding P217 (and people have explained why they disagree in depth) saying you have demonstrated it "irrefutably" leaves me speechless. Thryduulf (talk: local | en.wp | en.wikt) 20:09, 28 March 2016 (UTC)[reply]

@Thryduulf, Joshbaumgartner, Danrok, ArthurPSmith: Done: Majority of people agree that this is a useful property and not the same as inventory number (P217). --Srittau (talk) 23:31, 3 May 2016 (UTC)[reply]