Wikidata:Property proposal/Beta Code

‎Beta Code edit

Originally proposed at Wikidata:Property proposal/Lexemes

Done: Beta Code (P12434) (Talk and documentation)

Description	representation of Ancient Greek as ASCII characters
Represents	Beta Code (Q752325)
Data type	String
Domain	form
Allowed values	[ "-}]+
Example 1	Ἠέλιος/)he/lios (L1095729) → )he/lios
Example 2	εἲρειν (L1020946) → ei)\rein
Example 3	γλύφω (L7961) → glu/fw
See also	ALA-LC romanization (P8991)

Motivation edit

Beta Code is a form of representing Ancient Greek (Q35497) letters as ASCII used by some research institutions, e.g. Perseus. You can try it in this web converter.-- Kristbaum (talk) 14:43, 25 April 2023 (UTC)[reply]

Discussion edit

Maybe it would be useful for forms too, but in most dictionaries it's only done once for the lemma. Kristbaum (talk) 14:51, 25 April 2023 (UTC)[reply]
Comment @Kristbaum: do we really need a property for that? I think it can simply be a form, see what I tried on L:L1095729#F1. Cheers, VIGNERON (talk) 11:24, 1 May 2023 (UTC)[reply]
- @VIGNERON: Cool idea, but wouldn't that conflict with the idea of a spelling variant? It's not a different spelling just a different representation. Is there maybe another example of a similar format to model this after? (If not I'm fine with your suggestion) --Kristbaum (talk) 12:14, 1 May 2023 (UTC)[reply]
  - @Kristbaum: good question, the model is flexible enough to accept it but is it okay to do it? I'm not sure. Maybe a discussion on WD:LD would help to see clearer. Meanwhile, I don't have exact equivalent but I know that representations are used for very different things (from basic spelling variation L:L1473#F1 to script variation ਲੌਟਣ/لَوٹݨ (L1096159), or even syllables in Latin: L:L10907#F1 - the last probably being wrong), but I didn't crossed often such case. PS: see also the modeling on L:L1000845#F1. Cheers, VIGNERON (talk) 12:51, 1 May 2023 (UTC)[reply]
    - @VIGNERON: Great examples, thank you! But wouldn't a property for Beta Code fit in with the numerous Romanization properties? E.g ALA-LC romanization (P8991) or ALA-LC romanization for Ukrainian (P9453)? Or do you think it would be possible to model them as spelling variants too? – The preceding unsigned comment was added by Kristbaum (talk • contribs) at 10:57, 1 May 2023 (UTC).[reply]
Support Ionenlaser (talk) 13:01, 6 May 2023 (UTC)[reply]
Comment; I would consider a property like this to be more suitable than adding a variant representation on the form. On the example of ਲੌਟਣ/لَوٹݨ (L1096159) above, I would use ISO 15919 transliteration (P5825) if I wanted to add a Romanized transcription, as this information is for specialized purposes and not part of the general written use of the language, and because it is useful to add qualifiers to statements about transcriptions. I would actually suggest that this be used on forms rather than the lexeme itself—print dictionaries are not necessarily concerned with recording every form of a word and make necessary compromises for space. As we do not have these limitations on Wikidata, I think it makes sense to attach these transcriptions to forms (and this would be in line with how existing properties are used). --عُثمان (talk) 18:07, 8 May 2023 (UTC)[reply]
- @عُثمان: Attaching this as a statement only to forms seems to be a sensible way to use it --Kristbaum (talk) 22:58, 16 May 2023 (UTC)[reply]
Support, an important property for lexemes.--Arbnos (talk) 17:17, 17 January 2024 (UTC)[reply]
@Kristbaum, Ionenlaser, Arbnos: Done. I set this property to lexemes instead of forms, because the examples only link to Lexemes. Please, change the examples if it is an error. --Tinker Bell ★ ♥ 23:24, 9 February 2024 (UTC)[reply]