Template talk:Speciesbox

Fetch image from Wikidata

It'll be great to fetch image from Wikidata image property (if available) if image parameter value is empty or not specified. Bot could cleanup image parameter values if they are same as in Wikidata. EugeneZelenko (talk) 23:52, 20 March 2023 (UTC)

Some language editions of Wikipedia already do this (albeit with {{Taxobox}}, rather than Speciesbox). Images on Wikidata may be misidentified, but that's also true of images on English Wikipedia. I'm not necessarily opposed to fetching images via Wikidata, but I think I am opposed to "bot could cleanup image parameter values if they are same as in Wikidata", if I'm understanding it correctly. Images on Wikidata items for species should have a reference that asserts that the image is what it is claimed to be. Often that reference is (some language edition of) Wikipedia. I don't think a bot should be removing image parameter values from Wikipedia, when those values may represent the reference on Wikidata (Wikispecies went this route with vernacular names, which I think was a mistake; Wikispecies used to list vernacular names locally, and is a reference for (some) vernacular names on Wikipedia, but Wikispecies is now fetching vernacular names from Wikidata and has deleted local values; references to Wikispecies for vernacular names are no longer supported by Wikispecies). Plantdrew (talk)

• As with any kind of data retrieval from Wikidata, this needs great caution. Wikidata has many fewer active editors than we do. Because it scrapes data from what I have found to be highly unreliable language wikipedias (VI springs to mind), it frequently has out-of-date taxonomy and wrong taxon names. I absolutely agree with Plantdrew that under no circumstances should Wikidata information be allowed to over-ride any information here. If images were retrieved in the way suggested, the articles should be put in a tracking category so that editors here can check them. Peter coxhead (talk) 11:39, 22 March 2023 (UTC)
• I agree with the above comments. I'll add that Wikidata items are on taxon names (despite what the instance statements say) and can simultaneously cover different concepts using that name.

A tracking category may be the way to go, but instead of importing the image from Wikidata. If there is no image in the taxobox AND Wikidata has an image THEN add a category, say Category:Taxoboxes without images with potential image on Wikidata. It would need people willing to monitor the category to be useful. —  Jts1882 | talk  13:07, 22 March 2023 (UTC)

That seems a good idea. The only difficulty I see is that if the image wasn't then used here for some good reason, the article would still appear in the tracking category. Peter coxhead (talk) 16:39, 22 March 2023 (UTC)

I suppose we could have a parameter to indicate the wikidata image has been checked and is unsuitable, although a clear short parameter name might be difficult to find. The whole thing becomes rather convoluted if its not just a simple use Wikidata image or not. I think people do make an effort to find an image for species, so I'd guess there are not a huge number of missing images where Wikidata has images. I admit I could be completely wrong on this, though. —  Jts1882 | talk  17:16, 22 March 2023 (UTC)

@Jts1882: I tend to agree with you. I'm not sure how feasible it would be but it would be interesting to take one of the smaller subcategories of Category:Wikipedia requested images of biota and try to determine how many images would be available on Wikidata. Peter coxhead (talk) 07:20, 23 March 2023 (UTC)

The requested images categories are a mess. Images get added, but nobody updates the talk page to remove the request for an image, so there may be a lot of false positives, or a talk page gets created without a request for an image (and no image is in the article), leading to false negatives. I've been chipping away at false positives in Category:Wikipedia requested images of plants (using Imagechecker to find them), but there are still more than 2500 plant articles that Imagechecker thinks have images that have a talk page requesting an image (in some cases, the image is a map, or the wood of a tree, or just very poor quality; I've left the request for an image in place in those situations). Category:Bivalve articles needing images should be pretty clear of false positives, as I went through it recently. Plantdrew (talk) 16:55, 23 March 2023 (UTC)

Ah, I hadn't realized how inaccurate the categories were (e.g. the Imagechecker tool says 2,518 plant article talk pages are probably incorrectly tagged). Peter coxhead (talk) 11:18, 24 March 2023 (UTC)

Fixing speciesbox

Moved from Wikipedia:Village pump (technical) § Fixing speciesbox

 – –Novem Linguae (talk) 14:03, 4 July 2023 (UTC)

Hi, it appears that using "taxon" parameter in speciesbox is causing issues with the "scientific classification" section. Specifically it is parsing clades and other nonstandard parent-level taxa and displaying a "missing taxonomy template" message in the box. My workaround is to replace the taxon parameter with separate genus and species parameters on species pages with errors. Could there be a better solution? 108.18.207.147 (talk) 13:43, 4 July 2023 (UTC)

Not able to fix this genus taxon page because automatic taxobox does not take a genus parameter: Ameiurus. 108.18.207.147 (talk) 13:53, 4 July 2023 (UTC)

I checked your recent edit history and couldn't find a diff where the speciesbox was broken. Can you please post a WP:DIFF? –Novem Linguae (talk) 14:02, 4 July 2023 (UTC)

Hi, I was looking at bullhead catfish, Ameiurus nebulosus, Ameiurus melas, etc. I fixed the speciesbox for Ameiurus natalis, but the error itself will not show now in prior page versions. The genus page, Ameiurus, currently gives me an error message due to the presence of the clade Clupeocephala, but I should not change the parameter on that page due to differences between Automatic taxobox and speciesbox. 108.18.207.147 (talk) 14:16, 4 July 2023 (UTC)

Diff: https://en.wikipedia.org/w/index.php?title=Yellow_bullhead&diff=prev&oldid=1163368391 108.18.207.147 (talk) 14:17, 4 July 2023 (UTC)

Thanks for the diff. The |taxon= version and the |genus= |species= version of the speciesbox look identical to me. Can you elaborate a bit further on what you think is broken? –Novem Linguae (talk) 14:20, 4 July 2023 (UTC)

I think something is wrong with rank "clade" because it is not in the typical hierarchy (Kingdom-Phylum-Class-Order-Family-Genus-Species), but I am not exactly sure. 108.18.207.147 (talk) 14:28, 4 July 2023 (UTC)

Looks like taxon parameter had an unusual space character. I delete the space and replaced it with a plain one. The module splits the taxon name on the space to get the genus and species. So it was look for a taxonomy template with the name "Ameiurus#nebulosus" where is some sort of space. —  Jts1882 | talk  14:54, 4 July 2023 (UTC)

No that was not the problem. My edits don't show up in the edit history so were treated as null edits and that was what fixed them. I also fixed White bullhead with a null edit. I assume it some to do with the parent in the taxonomy template for {{taxonomy/Otocephala}} being changed to Clupeocphala yesterday and a caching problem. A null edit or any other edit (e.g. changing taxon to genus + species) should fix the issue. —  Jts1882 | talk  15:05, 4 July 2023 (UTC)

It appears that this edit of {{taxonomy/Otocephala}} at 15:57 on 3 July 2023, changing the |parent= from "Otocephala" to "Clupeocephala", created the error. The template {{taxonomy/Clupeocephala}} did not exist at the time and that was why there was a fix error message on the Clupeocephala line of the taxonomy hierarchy in the taxobox. The template was created three minutes later at 16.00 on 3 July (see history). Bizarrely the first edit propagated in those three minutes so it showed up as an error in the live taxoboxes and persisted even though the required template had been created. I suppose there was no way for the template creation to be associated with those pages that needed it because it had never been used there before. It needed an edit on the page after the template had been created to trigger the propagation of template/module changes. —  Jts1882 | talk  07:46, 5 July 2023 (UTC)

Banding codes for birds?

I wanted to suggest a potential addition to the speciesbox for bird species, the banding code. These are standardized 4 or 6-letter abbreviations of bird names used for brevity in checklists and database entry (4-letter based on the common name, e.g. MODO for MOurning DOve, 6-letter based on the binomial name, e.g. ZENMAC for Zenaida macroura) based on the common name. For many they are self evident, but for others there is ambiguity (should Great Blue Heron be GBHE? GRBH?) or conflict (Canada Goose, for example, is CANG, rather than CAGO, as CAGO conflicts with Cackling Goose). I think it would be useful for birders to be able to quickly reference these codes in the speciesbox. Gvbox (talk) 22:58, 25 July 2023 (UTC)

With two standards (BBL & IBP) and limited geographic applicability (North & Central America), I'm not sure these abbreviations should be incorporated into speciesboxes. If somebody is interested in these, in what direction would they be likely to search? 1) From the full name of the bird to find the abbreviation, or 2) from the abbreviation to find the full name of the bird? I would guess it's 1, but if it's 2, creating redirects for the abbreviations might be appropriate. iNaturalist lists abbreviations as common names in the "language" "Aou 4 Letter Codes"; does AOU follow BBL or IBP? WT:BIRDS might be a better place to discuss presenting the abbreviations in articles. Plantdrew (talk) 01:53, 26 July 2023 (UTC)

The {{taxonbar}} might be a better location if the codes were available on Wikidata. I can't find the AMBBP/BBL or IBP codes on Wikidata, so they would need to be added. The EURING number (P3459) is already available and that seems a similar thing. It might also be possible to allow it as a manual parameter in the taxonbar, although these seem to be discouraged. —  Jts1882 | talk  09:58, 26 July 2023 (UTC)

Error tracking when genus parameter (or first word in taxon parameter) isn't a genus?

At Peracle reticulata, the family Peraclidae had been put in |genus=. A taxobox displayed, but with no genus and the binomial displaying as Peraclidae reticulata. It seems like this is something that should go in an error tracking category. Plantdrew (talk) 20:28, 16 November 2023 (UTC)