Wikidata:Identifier migration/0

See Wikidata:Identifier migration for full history and User:ArthurPSmith/Identifiers/0 for statistics for properties on this page.

Properties in range P1 - P999: Wikidata:Identifier migration/0

Properties in range P1000-P1999: Wikidata:Identifier migration/1

Properties in range P2000-P2999: Wikidata:Identifier migration/2

Properties

Good to convert

Properties with serious objections to conversion

FAA airport code (P240) 99.21% unique, 100% single-valued on 2283 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
ISIL (P791)
- ~~Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures~~
- See P345.
  --- Jura 08:31, 25 February 2016 (UTC)[reply]
  - I don't understand this comment. P345 has multiple formatter URL's; P791 has only 1. What are you implying is the problem with P791? ArthurPSmith (talk) 13:12, 25 February 2016 (UTC)[reply]
  - It is really a technical detail to be sorted out by the devs. Don't worry about it. It assumes some knowledge about the identifier at hand.
    --- Jura 14:02, 27 February 2016 (UTC)[reply]
    - How is this a Wikidata property for a multi-source identifier (Q21264328)? Multichill (talk) 13:41, 28 February 2016 (UTC)[reply]
      - It's a good sample: (a) and (c), but not (b)
        --- Jura 16:56, 28 February 2016 (UTC)[reply]
- The first three letters of "ISIL ID" stand for "International Standard Identifier". It should be moved to "good to go" ASAP; not least as no good reason not to do so has been given. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:00, 29 February 2016 (UTC)[reply]
- See also Wikidata:Property_proposal/Authority_control#DE-ISIL.
  --- Jura 12:57, 2 April 2016 (UTC)[reply]
  - Have you tested this one, ArthurPSmith?
    --- Jura 14:44, 2 June 2016 (UTC)[reply]
    - Is there a URL linking issue? Please be specific on what the problem is. ArthurPSmith (talk) 16:21, 2 June 2016 (UTC)[reply]
      - I think the problem is still the one described on the property talk page. Only DE values can link to the German website.
        --- Jura 16:24, 2 June 2016 (UTC)[reply]
        Ah, ok. Looks like there's a list here. I'll take a look. ArthurPSmith (talk) 16:37, 2 June 2016 (UTC)[reply]
        To avoid many dead links, we could just remove the formatter url and add it through the gadget only.
        --- Jura 16:40, 2 June 2016 (UTC)[reply]
        @ArthurPSmith, Jura1: So why didn't this become an identifier? --Epìdosis 11:27, 2 August 2018 (UTC)[reply]
        I could not track down a workable formatter URL or combination of them. We could, as Jura suggests, just make it an identifier without a formatter URL, but that loses most of the advantage of making it an external id. ArthurPSmith (talk) 13:21, 2 August 2018 (UTC)[reply]
        @Epìdosis: We could convert it once formatter urls can take in account parameters (see some phab ticket). Also, I think it could benefit from being placed in the identifier section even with string datatype (see Wikidata:Contact_the_development_team#Placement_in_identifier_section for some other properties).
        --- Jura 17:46, 6 August 2018 (UTC)[reply]
        @ArthurPSmith, Jura1: Thank you. I agree with the placement in the identifier section and I hope the phab tickets will be solved. --Epìdosis 16:52, 12 August 2018 (UTC)[reply]
IATA airline designator (P229) - only 53.12% unique out of 2065 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- See User:Addshore/Identifiers#How_is_IATA_airline_designator_.28P229.29_an_identifier.3F.
  --- Jura 17:05, 29 February 2016 (UTC)[reply]
  - How did this one end up in the good to convert section Lydia? I understand you want to finish this, but quick and dirty is not the way to go. Multichill (talk) 16:31, 13 May 2016 (UTC)[reply]
ICAO airline designator (P230) - only 92.42% unique out of 2454 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
IATA airport code (P238) - only 82.56% unique out of 9437 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
ICAO airport code (P239) - only 78.91% unique out of 15384 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- I object against converting these three airline properties. These are more like abbreviations, not identifiers. Multichill (talk) 16:32, 13 May 2016 (UTC)[reply]
- They are generally unique within a given scope. Similar to P428 some are abbreviations.
  --- Jura 14:49, 2 June 2016 (UTC)[reply]
  - what scope? If you try to resolve via the IATA the designators that are currently in wikidata, MOST of them do not currently resolve to anything. This is very different for example from P428, where the abbreviations are persistent over a very long period of time. The IATA designators for an airline often don't even last 10 years. Plus they are - by the IATA - deliberately allowed to be ambiguous simultaneously, so even if you know the time period you can't be sure what it's designating. I don't know what you mean by "given scope", but this is the same situation as a number of other properties that we have decided NOT to convert due to ambiguity. And unilaterally moving these to "good to convert" with minimal explanation is a rather surprising action after Multichill's comments, I have moved them back here. ArthurPSmith (talk) 16:21, 2 June 2016 (UTC)[reply]

I have moved here properties that are either less than 60% unique-valued or have otherwise received serious objections about their validity as identifiers. ArthurPSmith (talk) 17:33, 13 May 2016 (UTC)[reply]

HomoloGene ID (P593) - only 51.90% unique out of 40363 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
ICPC 2 ID (P667) - only 42.06% unique out of 699 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
Strunz 8th edition (series ID, updated) (P711) - only 21.68% unique out of 4087 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
Nickel-Strunz 9th edition (updated 2009) (P712) - only 52.15% unique out of 4443 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
Nickel-Strunz '10th ed', review of (9th ed/ 2009 update) (P713) - only 51.35% unique out of 4513 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
ISCO-88 occupation class (P952) - only 53.92% unique out of 217 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
aircraft registration (P426)
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
HURDAT ID (P502)
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
ISIN (P946)
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures

already done

see https://www.wikidata.org/w/index.php?title=User:Addshore/Identifiers/0&oldid=342520410

Historic Environment Scotland ID (P709) - only 66.29% unique out of 41400 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- I took a look at this one. We have external id's, these are unique and distinct, so why only 66.29%? Judging from Wikidata:Database reports/Constraint violations/P709, the bot import was not so clean and nobody bothered to clean up yet. What a mess. I would convert this one anyway and not have the local mess here prevent it. Multichill (talk) 13:37, 28 February 2016 (UTC)[reply]
OpenStreetMap relation ID (P402) Check duplicate values --Lydia Pintscher (WMDE) (talk) 12:26, 13 May 2016 (UTC)[reply]
- I checked the duplicates, they were mostly incorrect statements. Should now be over 99% unique. - Nikki (talk) 17:30, 13 May 2016 (UTC)[reply]
ISO 15924 alpha-4 code (P506) - only 96.63% unique out of 89 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- @ArthurPSmith: I think this one can be converted. It's designed to be an identifier (like most of the ISO properties). The low percentage is just because of the low number of uses. Of the three non-uniques, one was wrong, one has been migrated to ISO 15924 numeric code (P2620) and one appears to be a valid exception. Also, now that ISO 15924 numeric code (P2620) has been created, it's just weird to have this as a string and that as an identifier. Any objections to me moving it to the section to be converted? - Nikki (talk) 23:58, 19 March 2016 (UTC)[reply]
  - No objections, it seems. :P - Nikki (talk) 17:32, 13 May 2016 (UTC)[reply]
State Water Register Code (Russia) (P884) - only 96.85% unique out of 29793 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- This has since been cleaned up. Now only has 1 duplicate. - Nikki (talk) 17:37, 13 May 2016 (UTC)[reply]

GOST 7.75–97 code (P278) - has one duplication among 88 uses - ? ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]

callsign of airline (P432) - only 96.19% unique out of 2334 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
IMA Number, broad sense (P484) - only 82.44% unique out of 3821 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
MeSH descriptor ID (P486) - only 89.06% unique out of 4663 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
A review of the line at Wikidata:Database reports/Constraint violations/P486#"Unique value violations" for D000230 seems to indicate these identifiers are associated with a particular class of disease (for which we have an item at Q356033). The external database also has the notion of "entry terms", which appear to be subclasses of the one disease. I believe that this could be treated as a unique identifier after cleanup for the entry terms, but maybe we should approach the relevant wikiproject. --Izno (talk) 14:36, 15 March 2016 (UTC)[reply]
This way of doing things is being maintained also by User:ProteinBoxBot it appears. There are other issues in this case regarding PBB's work. That is a problem IMO for conversion. --Izno (talk) 14:39, 15 March 2016 (UTC)[reply]
OMIM ID (P492) - only 90.88% unique out of 5649 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
ICD-9 ID (P493) - only 68.58% unique out of 6635 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
ICD-10 ID (P494) - only 61.78% unique out of 8670 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
ISO standard (P503) - only 94.50% unique out of 109 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
DiseasesDB (P557) - only 90.49% unique out of 2818 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
ICD-O (P563) - only 91.67% unique out of 348 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
Ensembl gene ID (P594) - only 96.27% unique out of 46805 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
MedlinePlus ID (P604) - only 91.96% unique out of 1456 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
E number (P628) - only 98.83% unique out of 511 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
cultural properties of Belarus reference number (P632) - only 98.03% unique out of 760 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
RefSeq protein ID (P637) - only 98.77% unique out of 123240 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
PDB structure ID (P638) - only 77.90% unique out of 42779 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
RefSeq RNA ID (P639) - only 98.97% unique out of 315931 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
RTECS number (P657) - only 97.96% unique out of 1716 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
GeneReviews ID (P668) - only 96.18% unique out of 157 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
MeSH tree code (P672) - only 88.89% unique out of 531 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
eMedicine ID (P673) - only 92.02% unique out of 2356 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
BHL page ID (P687) - only 97.76% unique out of 357 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
UN number (P695) - only 68.76% unique out of 461 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
Ensembl transcript ID (P704) - only 96.03% unique out of 243100 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
Minor Planet Center observatory code (P717) - only 98.19% unique out of 441 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- This has been improved to 98.87% unique over the last few days. It seems to be intended to be unique in almost all cases; I would be comfortable putting this in "Good to Convert" if somebody else agrees. ArthurPSmith (talk) 19:56, 29 February 2016 (UTC)[reply]
BGS Lexicon of Named Rock Units ID (P732) - one duplicate out of 39 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
Czech cultural heritage ID (P762) - 3 duplicates out of 258 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
ROME Occupation Code (P867) - only 76.45% unique out of 327 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
SSRN article ID (P893) - never used ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
SOC Code (2010) (P919) - only 87.50% unique out of 80 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- Should be converted, uniqueness is not an issue. The ##-#### format will need some custom URL handling to remove the hyphen when creating links (submitted here. Also, there is a new version of this coming in 2018, and we will need a new property for it (SOC Code 2018). That version is due out Summer 2017. Might be relevant depending on how long this project takes. -- Netoholic (talk) 11:39, 16 April 2016 (UTC)[reply]
ARICNS (P999) - never used ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
CAS Registry Number (P231)
Each CAS number corresponds to an unique chemical. But each chemical can have several CAS number. For me, we can consider it as an identifier, but we will have some problems with the single constraint. Snipre (talk) 20:17, 16 February 2016 (UTC)[reply]
- In practice this seems to be good: 99.44% unique on 19504 statements, 99.43% single-valued. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
- Actually, each CAS is not unique for a chemical... it's more complicated... there is not a 1-to-1 relation between Wikidata entries and a CAS number. For example, the CAS for formalehyde and formaline are identical (deliberately, I verified that with CAS). That said, I agree the CAS number is an "identifier" for a chemical entity, but not necessarily a chemical compound... Egon Willighagen (talk) 14:17, 5 March 2016 (UTC)[reply]
EC number (P232)
Not unique: can be used to define a mixture of stereoisomers and for pure stereoisomers. Snipre (talk) 20:10, 16 February 2016 (UTC)[reply]
- Again in practice this seems to be pretty good: 99.53% unique, 99.94% single-valued on 3631 statements ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
InChI (P234)
Not an identifier but a linear representation of a chemical. Is unique. Snipre (talk) 20:15, 16 February 2016 (UTC)[reply]
- 99.96% unique and 99.99% single-valued on 14360 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
- Not unique (particularly the Standard InChI is not) but is an identifier (or meant as such). It's not a linear notation because cannot always be converted back into a chemical graph (though often possible). Egon Willighagen (talk) 14:26, 5 March 2016 (UTC)[reply]
  - BTW, I would say the InChI is more unique than some of the identifiers already approved. In fact, ChemSpider bases it's concept of uniqueness on the InChI! If ChemSpider is unique enough, then InChI is too, almost by definition! Egon Willighagen (talk) 14:51, 5 March 2016 (UTC)[reply]
InChIKey (P235)
Not unique. Snipre (talk) 20:15, 16 February 2016 (UTC)[reply]
- 99.90% unique and 96.17% single-valued on 15277 statements. Not sure why the single-valued issue is so significant here, that doesn't seem to be what Snipre was suggesting... ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
- Same as comment as above for InChI (P234). Egon Willighagen (talk) 14:26, 5 March 2016 (UTC)[reply]
ISSN (P236) - note 99.51% unique but only 66.72% single-valued on 45807 statements. New ISSN's are issued whenever a journal name changes in the slightest way, and additionally for differing formats (print, CD, online) when it's actually the exact same journal as far as content goes. I think this should qualify as it is a good persistent identifier, but there is a lot of duplication so the same entity has multiple ISSN's over time and even more than one at the same time. To me this is a good example of why the uniqueness constraint is important, but single-valued not necessarily so relevant here. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
OCLC control number (P243) 99.08% unique, 96.13% single-valued on 25578 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
ATC code (P267) 99.82% unique, 79.19% single-valued on 2206 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
CALIS ID (P270) 100% unique, 96.93% single-valued on 163 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
ISO 3166-1 alpha-2 code (P297) 99.21% unique, 100% single-valued on 254 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
ISO 3166-1 alpha-3 code (P298) 100% unique, 100% single-valued on 253 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
ISO 3166-1 numeric code (P299) 99.60% unique, 100% single-valued on 253 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
ISO 3166-2 code (P300) 100% unique, 99.94% single-valued on 4916 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
IETF language tag (P305) 99.89% unique, 99.74% single-valued on 2733 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]

Ensembl protein ID (P705) Check duplicate values --Lydia Pintscher (WMDE) (talk) 12:26, 13 May 2016 (UTC)[reply]
UBIGEO code (P844) Check duplicate values --Lydia Pintscher (WMDE) (talk) 12:26, 13 May 2016 (UTC)[reply]
IMSLP ID (P839) - This is a wiki page name. When used with a formatter URL, it produces an external link, but is that really an identifier? There's nothing to stop pages being renamed and the old page name being used for something else. - Nikki (talk) 20:00, 17 February 2016 (UTC)[reply]

ISNI

ISNI (P213) 99.99% unique, 98.37% single-valued on 279153 statements. ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
- Apparently, the stated criteria hadn't been applied when evaluating these.
  --- Jura 13:28, 7 February 2016 (UTC)[reply]
  - @Jura1: can you give a detailed reason why you moved the above properties to the disputed section? --Pasleim (talk) 14:52, 8 February 2016 (UTC)[reply]
    - Absent a reply, in over two weeks, I propose that we move these back to "good to go". In particular, ISNI and BnF should be there. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 06:18, 26 February 2016 (UTC)[reply]
      - I agree - all of these meet at least the over 99% unique constraint and are widely used, I think these (the items remaining on Jura's "not applied" list here) should all be in "good to convert". ArthurPSmith (talk) 14:42, 26 February 2016 (UTC)[reply]
        I see it has 4552 items with multiple links. Not sure if that's an indication any problems. Multichill (talk) 16:49, 28 February 2016 (UTC)[reply]
        What has 4552 items..?" Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 09:55, 29 February 2016 (UTC)[reply]
        Andy Mabbett I believe Multichill was talking about ISNI - I've been discussing this elsewhere, ISNI definitely has some issues with duplicate identifiers for the same person or organization. However, it is an international standard identifier (ISO 27729) and as such I think among those on the above list that most clearly should qualify as external identifiers... ArthurPSmith (talk) 14:57, 29 February 2016 (UTC)[reply]
        Since ISNI is not about persons &c. but their "public identities" the number of single value constraint violations is partly by design and does not matter for the other part. Being recognized here as an external identifier does not imply that the target follows Wikidata's criteria for notability and entity shaping. -- Gymel (talk) 05:46, 13 April 2016 (UTC)[reply]
        A problem specific with ISNI is that the standard formatter URLs currently don't work. We would need to make sure the code works before converting them. We don't want to end up with conversions that apparently can't be undone as for IMDb.
        --- Jura 06:26, 13 April 2016 (UTC)[reply]
- ISNI must be converted. Nemo 16:39, 13 May 2016 (UTC)[reply]
- I would like a headcount from the above users about conversion since it is not obvious to me whether there is consensus to convert. @ArthurPSmith, Jura1, Pasleim, Pigsonthewing, Multichill, Gymel: Please just either {{Support}} or {{Oppose}} conversion below. --Izno (talk) 13:17, 27 May 2016 (UTC)[reply]
  - Strong support ArthurPSmith (talk) 14:49, 27 May 2016 (UTC)[reply]
  - This is not a vote; and voting is not how we determine consensus. (See en:WP:VOTE for background.) The few objections here are either not substantiated, or are irrelevant, or both. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:05, 27 May 2016 (UTC)[reply]
    - Oh come on, we've had discussion spanning months here, Izno was just trying to get a feel for where everybody stands right now. Multichill's comment above seems ambiguous for example. ArthurPSmith (talk) 15:19, 27 May 2016 (UTC)[reply]
      - We have, lamentably indeed had "discussion spanning months here". And in all that time, we have had no relevant, substantiated reason not to do the conversion. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:05, 27 May 2016 (UTC)[reply]
    - If it is true that the few objections are neither substantiated nor relevant, then the others I requested feedback from will surely agree to support conversion. --Izno (talk) 16:00, 27 May 2016 (UTC)[reply]
  - Sure it is an external identifier for entities sufficiently aligned with Wikidata's view upon the world, had there ever been real dispute over it? So if the discrepancy between the way we store them here (in the official presentation form with interspersed spaces) and the URI/URL form (plain 15+1 digit string) is not an obstacle for conversion - just do it. -- Gymel (talk) 15:21, 27 May 2016 (UTC)[reply]
  - It is for me hard to understand why ISNI (P213) does not yet have the external id datatype. Lymantria (talk) 15:52, 27 May 2016 (UTC)[reply]
    If you want to understand why, you might want to follow the history of edits here - in this one on Feb 4 I moved ISNI among several others to "good to convert". That was objected to as the above discussion suggests, without any clear explanation. Later on the issue of the format/formatter URL problem has been raised but we have a number of external ID's that don't even have formatter URL's so I'm not sure why that's been blocking it so far. ArthurPSmith (talk) 16:18, 27 May 2016 (UTC)[reply]
    - Support I think I already identified this 6 months ago one. Just do it correctly. -- Jura 03:56, 28 May 2016 (UTC)[reply]

with a new URL formatter there is really no reason this should not be converted now. ArthurPSmith (talk) 14:30, 2 June 2016 (UTC)[reply]

Support who can convert? @Lydia Pintscher (WMDE):? -- JakobVoss (talk) 21:45, 5 January 2018 (UTC)[reply]

The easiest would be if you open a ticket on phabricator.wikimedia.org. Then I'll get it to the developers. --Lydia Pintscher (WMDE) (talk) 15:06, 7 January 2018 (UTC)[reply]

Not going to convert

ticker symbol (P249) - non unique, not managed by a single authority (symbol depends on market) ArthurPSmith (talk) 16:18, 4 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
LEM ID (P920) Check duplicate values
RefSeq (P656) - qualifier for RefSeq protein ID (P637) and RefSeq RNA ID (P639) Laboramus (talk) 23:47, 6 March 2016 (UTC)[reply]
Commons category (P373) - multiple items may share the same commons category, I don't think this is suitable as an identifier ArthurPSmith (talk) 21:36, 4 February 2016 (UTC)[reply]
- note only 82.42% unique out of 1369383 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
local dialing code (P473) - only 18.36% unique out of 41214 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
country calling code (P474) - only 76.07% unique out of 280 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
Unicode character (P487) - only 98.22% unique out of 2366 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
catalog code (P528) - only 91.94% unique out of 104627 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
  - Correct, same issues as with inventory number. Multichill (talk) 13:37, 28 February 2016 (UTC)[reply]
  - See User:Addshore/Identifiers#Question_to_sort_out_.281.29.
    --- Jura 17:05, 29 February 2016 (UTC)[reply]
runway (P529) - only 16.93% unique out of 443 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
website username or ID (P554) - never used (actually, only used as a qualifier) ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
NATO reporting name (P561) - only 93.88% unique out of 441 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
yard number (P617) - 2 duplicates out of 99 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
military designation (P798) - only 95.84% unique out of 985 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
Air Ministry specification ID (P799) - 2 duplicates out of 15 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
arXiv classification (P820) - never used, or only as a qualifier ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
United States Navy aircraft designation (P847) - 1 duplicate out of 38 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
pennant number (P879) - only 91.77% unique out of 948 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
P940 (P940) - only 27.54% unique out of 167 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
burial plot reference (P965) - only 68.09% unique out of 47 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
P969 (P969) - only 95.89% unique out of 291490 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
OS grid reference (P613) - As I understand it, this refers to the square on a map where something is located. Unless something is big enough to fill the whole square, this is not unique. - Nikki (talk) 20:00, 17 February 2016 (UTC)[reply]
- Yes and no - it depends on the level of precision given in the grid reference. The most common is a six figure reference (e.g. TQ 301 796) which gives a 100m square but 10 figure references are possible (e.g. TQ 30143 79652) to give a 1m grid square (in this case for Statue of Winston Churchill (Q7604505)). It doesn't make a good external identifier though. Thryduulf (talk) 00:48, 11 July 2016 (UTC)[reply]
Commons gallery (P935) - This is also a wiki page name, as with P839 above. - Nikki (talk) 20:00, 17 February 2016 (UTC)[reply]
author citation (zoology) (P835) Check duplicate values Does not seem to be a reference to a specific external database or authority. Laboramus (talk) 07:01, 15 March 2016 (UTC)[reply]
provisional designation (P490) - no common authority DB. Laboramus (talk) 07:19, 15 March 2016 (UTC)[reply]
postal code (P281) - not unique, not managed by a single authority (each country has its own system of postal codes) ArthurPSmith (talk) 18:43, 4 February 2016 (UTC)[reply]
- note only 77.65% unique among 663638 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
- This needs P17 for formatter URLs to work.
  --- Jura 06:26, 13 April 2016 (UTC)[reply]
station code (P296) - not unique, many different sources (every railway system can use a different code) ArthurPSmith (talk) 18:43, 4 February 2016 (UTC)[reply]
- note only 95.65% unique out of 2391 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
- Same as P281.
  --- Jura 06:26, 13 April 2016 (UTC)[reply]
UN packaging group (P876) - 1 duplicate out of 3 uses ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- The set of valid claims is {I, II, I/II, III, II/III}. clearly not unique. --Izno (talk) 18:15, 15 March 2016 (UTC)[reply]
given name (P735) - not a property with string datatype. --- Jura 11:39, 3 February 2016 (UTC)[reply]
UN class (P874) Check duplicate values
UN classification code (P875) Check duplicate values
licence plate code (P395) - non unique and covers a multitude of different encodings (countries, localities within countries) and even within the same country the same string can be shared by multiple locations as determined by local authorities ArthurPSmith (talk) 21:36, 4 February 2016 (UTC)[reply]
- note only 9.99% (less than 10% !) unique out of 9998 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
house number (P670) - not unique (only in combination of street/city/country). Michiel1972 (talk) 09:29, 5 February 2016 (UTC)[reply]
- Agree, but mainly as it would be an identifier attached to another identifier --- Jura 10:45, 7 February 2016 (UTC)[reply]
- Note only used once, other than as a qualifier ArthurPSmith (talk) 15:06, 15 February 2016 (UTC)[reply]
- Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
United States Army and Air Force aircraft designation (P897) Check duplicate values --Lydia Pintscher (WMDE) (talk) 12:26, 13 May 2016 (UTC)[reply]
inventory number (P217) - non unique, not managed by a single authority ArthurPSmith (talk) 16:11, 4 February 2016 (UTC)[reply]
- note only 95.55% unique out of 132947 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
  - Disputed by Multichill at Wikidata:Project_chat/Archive/2016/02#upcoming_deployments.2Ffeatures
    - Correct. Inventory numbers are generally only unique and distinct within a collection, globally that's most certainly not the case. Take for example inventory number 1. Multichill (talk) 13:37, 28 February 2016 (UTC)[reply]
      - There is a problem with your query: it omitted the mandatory qualifier. See User:Addshore/Identifiers#Question_to_sort_out_.281.29.
        --- Jura 17:05, 29 February 2016 (UTC)[reply]
Wikimedia language code (P424) - only 36.62% unique out of 983 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
EC enzyme number (P591) - only 20.44% unique out of 6918 uses ArthurPSmith (talk) 14:34, 15 February 2016 (UTC)[reply]
- Object, see en:Enzyme Commission number. "Strictly speaking, EC numbers do not specify enzymes, but enzyme-catalyzed reactions. If different enzymes (for instance from different organisms) catalyze the same reaction, then they receive the same EC number." this is not an identifier. Multichill (talk) 16:35, 13 May 2016 (UTC)[reply]
IUCN taxon ID (P627) - this is used mainly in references with IUCN conservation status (P141). Laboramus (talk) 23:47, 6 March 2016 (UTC)[reply]
- How does that make it a non-identifier?
  --- Jura 09:01, 9 March 2016 (UTC)[reply]
  - It doesn't make it non-identifier per se, it makes it unclear whether it is an identifier or not (since I haven't found any definite rules list). It also makes it hard to figure out uniqueness, since the table by ArthurPSmith only works with statements, but not qualifiers. So I moved it here so it would be clear that it needs discussion and special checking and would not have to be re-triaged by other people. If somebody checks it more thoroughly and finds out it's ok for identifier, then move it to the "good" list with appropriate comment, I have no objection to that. Laboramus (talk) 19:30, 11 March 2016 (UTC)[reply]
    - Of all people, I think you should be able to figure out a query to do that ;). Adam's table is just a query he made as an alternate to applying the criteria he subscribed to.
      --- Jura 11:55, 14 March 2016 (UTC)[reply]
      - See however https://www.wikidata.org/wiki/Wikidata:Project_chat/Archive/2015/10#About_Property:P627 - looks like it's not the stable ID and not recommended to be used as an ID. Laboramus (talk) 07:42, 15 March 2016 (UTC)[reply]
        That doesn't look good. Let's skip them. Property description needs updating.
        --- Jura 17:09, 15 March 2016 (UTC)[reply]
- It's an identifier. It's external. It should be converted. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:20, 13 January 2017 (UTC)[reply]
  - So what kind of information does it uniquely identify? --Succu (talk) 22:39, 13 January 2017 (UTC)[reply]