Wikipedia talk:Automated taxobox system

This is the talk page for discussing improvements to the automated taxobox system as a whole – not just one page.

Put new text under old text. Click here to start a new topic.
New to Wikipedia? Welcome! Learn to edit; get help.

Archives: Index, 1, 2, 3, 4, 5, 6, 7: 41.5 days

This talk page can be used to discuss issues with the automated taxobox system that are common to the entire system, not just one of its templates. Discussions of this nature prior to 2017 can be found at Template talk:Automatic taxobox

Those familiar with the system prior to mid-2016 are advised to read Notes for "old hands".

Need to remove italics for Candidatus genus

At Template:Taxonomy/Protochlamydia I am trying to create a link of the form "Ca. Protochlamydia". However, the mighty taxobox system is insisting on italicizing the whole thing. Trying to use the template {{normal}} does not work, and a quick examination of the output HTML by inspect element shows that it's been filtered out somewhere. Trying to just type ''Ca.'' Protochlamydia or Ca. ''Protochlamydia'' gives the same result.

There are two routes (I think) to fixing this, given my limited 5-minute read of the call chain.

It's easy and tempting to just change Module:TaxonItalics, but that comes with the issue of this module not being called for anything higher than a genus – Candidatus higher taxa do exist, and even there you are supposed to italicize the Candidatus while leaving the rest alone.

The other route is to just directly do the formatting in the callers, which would require similar changes in multiple modules: Module:Autotaxobox AND Module:Automated taxobox. The caller logic would be basically changed to:

local i, j = linkText.find("^Ca(%.|ndidatus) ")
if i != nil then
  linkText = "''" .. name.sub(i, j-1) .. "'' " .. name.sub(j+1)
else
  if italics == "yes" then   -- now an elseif would work here, but i want to make it resemble the original a bit more
    linkText = TaxonItalics.italicizeTaxonName(linkText, false, abbreviated=='yes')
  end
end

(Because we are supposed to leave the rest alone, there will be no need to call italicizeTaxonName even for a rank that ordinarily requires italics.)

Artoria 2e5 🌉 11:53, 5 May 2025 (UTC)[reply]

Is this something that could perhaps be handled with a parameter for Candidatus taxa in taxonomy templates? I know I haven't worked much on implementing automatic taxoboxes for bacteria because the system can't handle Candidatus status well. Plantdrew (talk) 16:50, 5 May 2025 (UTC)[reply]

Updated/improved documentation of acceptable ranks

In several discussions lately, I have realized that unless an editor looks at the raw code of parts of the automated taxobox system, the acceptable ranks and how they are handled were not fully documented. Accordingly:

I have made some changes to Module:Autotaxobox to support new documentation.
I have created Wikipedia:Automated taxobox system/ranks which lists all the ranks accepted by {{Anglicise rank}} and notes how they are handled. The numerical values that are used to check consistent rank ordering are now shown here rather than in Wikipedia:Automated taxobox system/taxonomy templates.

I hope this new documentation page should make it easier to explain how the rank part of the system works.

I've checked the test cases very carefully, but changing Module:Autotaxobox is always fraught with danger, so please revert if you see any problems! Peter coxhead (talk) 06:55, 6 May 2025 (UTC)[reply]

Headers

Why do we have headers? The first line in a taxobox just reproduces the title of the article (unless overrriden by |name=). As far as I've been able to determine digging through TOL and Template talk:Taxobox archives, it's something that has basically just always been done without any question. I'm against redundant repetitions of the article title in a taxobox (i.e. image captions that repeat the title), and I'm not sure what the justification for headers is. Taxoboxes were the first infoboxes on Wikipedia, and other infoboxes mostly have headers, but I'm not sure why. A couple different systems are {{Infobox food}} (see e.g. Pie) where the header floats outside of the infobox), and {{infobox ship begin}}, where a header may be specified or not (USS Enterprise (CVN-65) specifies |infobox caption=, Titanic does not). Plantdrew (talk) 02:52, 7 May 2025 (UTC)[reply]

Use with genera of uncertain family placement

I was hoping to convert Neoalbatrellus caeruleoporus' manual taxobox over to speciesbox, but have realised that the genus Neoalbatrellus is unplaced to family within Russulales by Index Fungorum[1] and MycoBank[2]. How should cases such as this be handled? Ethmostigmus 🌿 (talk | contribs) 05:40, 7 May 2025 (UTC)[reply]

@Ethmostigmus:, set the parent in the taxonomy template to "Incertae sedis/Russulales" ({{Taxonomy/Incertae sedis/Russulales}} already exists, but there are a lot of fungi that are incertae sedis at ranks where an incertae sedis taxonomy template doesn't yet exist).

Another way to handle this would be to just set the parent to "Russulales". There hasn't really been much discussion I am aware of about when to create an incertae sedis taxonomy template and when to just skip a rank. I think when a source explicitly says something is incertae sedis (as Index Fungorum often does), we should use incertae sedis taxonomy templates, and perhaps also if a source explicitly mentions that different data (e.g. mitochondiral DNA vs. nuclear DNA) produces conflicting results. Skipping ranks may be more appropriate for fossil taxa where the data just doesn't exist. 15:48, 7 May 2025 (UTC) Plantdrew (talk) 15:48, 7 May 2025 (UTC)[reply]

I favour setting the parent to the order as I think it easier to keep track and having an incertae sedis taxon doesn't provide any useful information. But I'd agree with using them if the source does it that way. Being pragmatic, I'd use them if the template existed for consistency, but won't create new templates. — Jts1882 | talk 16:56, 7 May 2025 (UTC)[reply]

Thanks very much @Plantdrew and @Jts1882 - I didn't realise that incertae sedis taxonomy templates even existed. I'll go ahead use that unplaced Russulales family template as the parent for Neoalbatrellus. I wasn't sure if "skipping" a rank when creating taxonomy templates was acceptable or not, so good to get confirmation on that. I take it there is no consensus on whether it is preferable in cases like these to use an incertae sedis template or simply set the parent to the nearest known higher taxon, and it's just a matter of editor preference? Ethmostigmus 🌿 (talk | contribs) 04:27, 9 May 2025 (UTC)[reply]

@Ethmostigmus:, right, there is no consensus. I've now searched talk page archives for WikiProjects TOL, Palaeontology, Fungi, Lepidoptera and Plants and the only discussion I found that was really relevant to using incertae sedis templates versus nearest known higher taxon was Wikipedia_talk:WikiProject_Fungi/Lichen_task_force/Archive_1#Taxonomy_templates, where Esculenta admitted to doing both approaches (doing known higher taxon when lazy). Palaeontology and Fungi would be the most relevant projects with this issue, but Lepidoptera has a handful of articles such as Incertae sedis (Arctiini) (which I don't think should exist), and Plants is the most active project. Plantdrew (talk) 16:57, 9 May 2025 (UTC)[reply]

I have to say that I disagree with Jts1882. Art. 3.1 of the ICNafp, which covers fungi, says "Thus, each species is assignable to a genus, each genus to a family, etc." So we know that the genus Neoalbatrellus definitely belongs to a family, we just don't yet know which. Setting family to incertae sedis conveys this accurately. Peter coxhead (talk) 13:56, 13 May 2025 (UTC)[reply]

Automated taxobox numbers

I was looking at the recent updates to the automated taxobox system documentation by Peter coxhead at Wikipedia:Automated taxobox system/map and was surprised to see the manual taxobox transclusion count at 35,000, which is well under 10% of all taxoboxes ({{taxobox/core}} gets 475,000 calls).

Checking Plantdrew's six monthly updates (for 30_December_2024) there were over 43,000, although there might be some double counting in insects. As search for manual taxoboxes for insects gave 25,581 on January 1st and this is now down to 19,622. Someone is putting in the work. — Jts1882 | talk 12:06, 13 May 2025 (UTC)[reply]

All these counts need treating with some caution. In particular, the transclusion counts which I've used in the table at Wikipedia:Automated taxobox system/map now use Module:Transclusion count which in turn uses Wikipedia:Database reports/Templates transcluded on the most pages. These counts include all namespaces. Right now, searching for "hastemplate:taxobox" only in article space gives a page count of 32,527, whereas the transclusion count for Template:Taxobox gives a figure of 34,999. The extra 2,472 transclusions are outside article space. Peter coxhead (talk) 13:40, 13 May 2025 (UTC)[reply]

User:The Knowledge Pirate is putting in most of the work in recent months. User:Videsh Ramsahai has made some progress with prokaryotes. I've only done a few hundred automatic taxoboxes so far this year.

There isn't really any double counting in my Total row. The count for Auto in Total is a sum of the transclusion counts for Automatic taxobox+Speciesbox+Subspecies box+Infraspecies box+Virusbox, not a sum of the counts for the WikiProject rows (maybe I should call it something other than Total?). There could be some double counting if an article has multiple taxobox templates, but that is very rare.

We are getting to a point where a significant number of the remaining manual taxoboxes are outside of article space (mostly Drafts and User pages). The monthly Parameter report only counts templates in article space and article talk pages (the talk pages were added a couple of months ago). Plantdrew (talk) 16:56, 13 May 2025 (UTC)[reply]

(edit conflict)

They are generally close enough, even though there are a few thousand in user space.

I was thinking of the best proxies for total taxoboxes and total automated taxoboxes. These seem to be {{taxobox/core}} (obviously) and {{Edit taxonomy}} (obvious in hindsight), respectively. Then {{Anglicise rank}} will give to total for the system (taxoboxes and taxonomy templates). Curiously {{Taxonomy}} is used by all taxoboxes, but the transclusion count excludes {{automatic taxobox}} in the count. This must be because that template calls {{taxobox/core}} from a Lua module, although I'm puzzled why as other templates called from Lua seem to be counted. — Jts1882 | talk 17:03, 13 May 2025 (UTC)[reply]

Change to system working

I've now completed a substantial overhaul of the internal working of part of the automated taxobox system. In particular, configuring ranks, both the rank names that are accepted and the ordering of the subset of ranks that can be checked for consistent rank order, is now done in a new module, Module:Taxobox ranks. (As a side-effect, {{Anglicise rank}} is now by-passed in some cases, so its transclusion count in future may not reflect the total number of taxoboxes and taxonomy templates.) As far as I can tell, everything is working correctly.

A major motivation for the change is that it has made generating the two tables at Wikipedia:Automated taxobox system/ranks fully automatic, whereas previously the ranks had to be separately listed in the first column of the first table, thus allowing ranks to be missed. The wikilink Wikipedia:Automated taxobox system/ranks#Rank table may be useful in discussions of accepted/acceptable ranks. Peter coxhead (talk) 08:03, 15 May 2025 (UTC)[reply]