Wikidata:Property proposal/L-number

L-number

Originally proposed at Wikidata:Property proposal/Natural science

   Done: L-number (P7291) (Talk and documentation)

Description	Quasi-scientific classification system for catfish from the Loricariidae family.
Represents	L-number (Q1088932)
Data type	External identifier
Domain	taxon (Q16521), Loricariidae (Q663380)
Allowed values	L(\d{3})([a-z]*)(([ ][A-Za-z]+)*)
Allowed units	none
Example 1	Baryancistrus xanthellus (Q4865994) → L018
Example 2	Baryancistrus xanthellus (Q4865994) → L081
Example 3	Baryancistrus xanthellus (Q4865994) → L085
Example 4	Baryancistrus xanthellus (Q4865994) → L177
Example 5	Hypostomus faveolus (Q5527917) → L037a
Example 6	Panaque armbrusteri (Q6409729) → L027 Tapajos
Example 7	(anomalous, non-unique) Baryancistrus demantoides (Q1758685) → L200
Example 8	(anomalous, non-unique) Hemiancistrus subviridis (Q5561051) → L200
Source	w:en:L-number
Planned use	Articles about species of Loricariid catfish.
Number of IDs in source	less than 1000
Expected completeness	always incomplete (Q21873886) - In practice we can always catch up. Adding them to WD faster than DATZ defines new ones shouldn't be too hard. But once DATZ defines a new one, we're incomplete again, so in principle; No.
See also	taxon name (P225), taxon synonym (P1420)

Motivation

The L-number has been an established identifier in tropical freshwater fishkeeping for many years. Nice to have in infoboxes. Also, not all notable Loricariid species have proper scientific names yet, so it can be useful as an interim identifier. I chose string as a datatype. If the authoritative online source was useful as a link target, the datatype could also be identifier, but I don't think the DATZ list is very useful for that. - Soulkeeper (talk) 12:49, 20 August 2019 (UTC)

Discussion

•  Comment if it's a stable and unique identifier, it should have external-id as datatype, independently of a possible formatter url. --- Jura 16:50, 20 August 2019 (UTC)

  That does make sense. I have changed it, and also added a regex for allowed values, and some more examples. The identifier is unique and stable, with some reservations:

  • Many species have multiple L-numbers. I don't think that is a problem.
  • In the very beginning (the late 1980s), when a fish with an L-number received a proper scientific name, they did recycle some L-numbers. They soon realized that was a terrible idea, and stopped doing it. The L-numbers used today describe the "newer" species, but in some older magazines or books from the late 1980s or early 1990s, an L-number on the lower end of the scale just might refer to a different species, which later got scientifically named.
  • There is one newer case that I am aware of, where the identifier is not unique; When the catfishes identified as L200 were scientifically categorized, the ichtyologists were perhaps as surprised as the hobbyists to find that these fishes turned out to be two different species that looked very much alike and were found roughly in the same location; Hemiancistrus subviridis (Q5561051) and Baryancistrus demantoides (Q1758685). This is an anomaly though. See also Baryancistrus demantoides and Hemiancistrus subviridis, two new uniquely colored species of catfishes from Venezuela (Siluriformes: Loricariidae) (Q54670861)

  - Soulkeeper (talk) 18:24, 20 August 2019 (UTC)

•  Support David (talk) 07:04, 21 August 2019 (UTC)

@ديفيد عادل وهبة خليل 2, Soulkeeper, Jura1:  Done: L-number (P7291). − Pintoch (talk) 18:43, 8 September 2019 (UTC)