Template talk:Automatic taxobox/Archive 8

Template loop warnings
Any idea why we have so many taxonomy templates (and articles) in Category:Template loop warnings? Thanks! Plastikspork ―Œ (talk) 05:55, 9 December 2010 (UTC)
 * A bunch of stuff is displayed when you go to a template's page, eg: template:Taxonomy/Physeterida, and I think there are loop errors in generating that display stuff, but it is not a problem on any mainspace pages as far as I can tell. ErikHaugen (talk | contribs) 06:09, 9 December 2010 (UTC)
 * The underlying problem should now be fixed, and performance improved. When retrieving a taxon's parameters, the most efficient way is to use {taxonomy/TAXON|machine code=rank (etc)}.  Martin  (Smith609 – Talk)  13:58, 9 December 2010 (UTC)
 * Looks great! Thanks! ErikHaugen (talk | contribs) 15:43, 9 December 2010 (UTC)
 * It's not fixed. Can you look at the article pages in the category? I had to change out Mammal to a fixed taxobox. Please examine the articles and you will see the problems.
 * Gremiphyca -- Template loop detected: Template:Get link
 * Mammal -- Phylum Chordata, (unranked): 	Template loop detected: Template:Get link
 * --Kleopatra (talk) 18:06, 9 December 2010 (UTC)
 * Kleopatra, what was wrong with the taxobox on Mammal? Thanks, ErikHaugen (talk | contribs) 18:17, 9 December 2010 (UTC)
 * Oh, I see you said it had a template loop on it. I think it's been fixed, but I won't change it back to automatic taxobox for now. ErikHaugen (talk | contribs) 18:19, 9 December 2010 (UTC)
 * (EC)Instead of giving an unranked taxon after the Phylum, it said, "Template loop detected:Template:Get link" in big red spill over letters. --Kleopatra (talk) 18:22, 9 December 2010 (UTC)

I'm just going to start posting problems at the community boards in the future. I'm tired of being ignored and told off for pointing out errors. This coding is not ready for prime-time, and it's detrimental to wikipedia to have so many high traffic articles repeatedly appear with ugly red code in the taxobox.

It's a great idea, but it needs coded better than what is going on. --Kleopatra (talk) 18:31, 9 December 2010 (UTC)
 * "I'm tired of being ignored and told off for pointing out errors." What are you referring to? Thanks, ErikHaugen (talk | contribs) 18:39, 9 December 2010 (UTC)
 * Um...your bug reports certainly won't be addressed anywhere but here, Kleopatra. Please be patient with us, as these manually-filed bug reports are often the only indication we even have of errors. Also-- I don't see anything anyone said to you that sounded rude or hateful, so I'm confused as to why you're offended.
 * I've noticed that a few purely aesthetic templates are still generating the loop error (check that category posted at the top of this section), though the pages that use them don't appear to have any issues. At this time I'm intensely busy finishing up coursework for this semester, so I don't have time right now to investigate this error that (I think) only appears behind the scenes as of now. Bob the Wikipedian (talk • contribs) 05:21, 10 December 2010 (UTC)
 * At the moment, these errors occur when a taxon uses the template "Same as taxon", and one of its parent taxa also uses this template. Unless I can do something fancy with {#iferror}, avoiding this would require some esoteric coding that'd make things more complicated for editors (so I'd like to avoid it).  I wonder whether these cases (Template:Taxonomy/Cetacea/Mammalia is the example I'm familiar with) could in fact be resolved by modifying the taxonomy to avoid the dummy taxa?  Martin  (Smith609 – Talk)  23:05, 14 December 2010 (UTC)
 * I'm certainly open to suggestions regarding that dummy template. I'm assuming we don't want Artiodactyla in whale taxoboxes, and I think the ancestor chain was too long for whale species. This fixed both those problems. Can we just live with this loop error? ErikHaugen (talk | contribs) 01:34, 15 December 2010 (UTC)
 * Probably. But as a programmer, I have this urge to eliminate errors... Bob the Wikipedian (talk • contribs) 06:01, 16 December 2010 (UTC)
 * Yeah, me too, but working with these templates is worse than XSLT! ErikHaugen (talk | contribs) 17:58, 16 December 2010 (UTC)

Cetacea

 * I suppose we could move Cetacea/Mammalia back to Cetacea (Mammalia)? I'd rather have the loop error, but either way. ErikHaugen (talk | contribs) 18:00, 16 December 2010 (UTC)
 * Don't allow this error to limit us, Erik. We shall overcome. Just not right now... lol. Bob the Wikipedian (talk • contribs) 20:56, 16 December 2010 (UTC)
 * I think that the reason that Template:Taxonomy/Cetacea/Mammalia exists is that the order Cetacea is the child of the (paraphyletic) order Artiodactyla. The solution is that the two orders have since been merged into the Cetartiodactyla – see for example . Anybody want to fix it? Martin  (Smith609 – Talk)  06:30, 17 December 2010 (UTC)
 * Are you suggesting that deer/cow taxoboxes all say "order Cetartiodactyla" instead of "order Artiodactyla" and that whale taxoboxes say "order Cetartiodactyla" instead of "order Cetacea"? I left it like this because a.) that seemed like a pretty bold change and b.) in my limited reading it seems people still use "Artiodactyla" a lot and include whales in there, and I thought "Cetartiodactyla" was sort of just another name for the same thing. IOW, it didn't seem like "Cetartiodactyla" had really replaced "Artiodactyla" despite nobody really disputing Artiodactyla being a parent of Cetacea – Artiodactyla is just considered to now include Cetacea. I could very easily be totally wrong – if I am let's do it. But, the other problem, as I outlined at Template talk:Taxonomy/Cetacea/Mammalia, is that the ancestor chain is too long and "Kingdom: Animalia" wasn't showing up. ErikHaugen (talk | contribs) 23:19, 17 December 2010 (UTC)


 * See Wikipedia talk:WikiProject Tree of Life/Archive23. If I'm not mistaken, that WikiProject still treats Mammals of the World as gospel, so it'd be a no-go. Bob the Wikipedian (talk • contribs) 03:27, 18 December 2010 (UTC)
 * If that's what they are advocating, then that's what our back-end templates should reflect; i.e. cetaecea should not be a child of Artiodactyla. Martin  (Smith609 – Talk)  22:02, 1 January 2011 (UTC)
 * Well, do you think the way it is now is ok? ErikHaugen (talk | contribs) 09:01, 2 January 2011 (UTC)
 * Someone from Wikiproject Mammals would be better qualified to answer that than I am. The list displayed at Template:Taxonomy/Cetacea is quite confusing at the moment. Martin  (Smith609 – Talk)  14:15, 2 January 2011 (UTC)
 * I'm assuming folks at Mammals mostly care what the taxoboxes look like, and those pretty much look the same as they always have. The list there needs to look like that to get pages like Cetruminantia to have the correct automatic taxooboxes, right? ErikHaugen (talk | contribs) 15:15, 2 January 2011 (UTC)
 * But Cetruminantia is very much a part of Cetartiodactyla, and (since it contains the Cetaecea) surely does not belong within Artiodactyla, as the taxobox suggests! Martin  (Smith609 – Talk)  15:47, 2 January 2011 (UTC)
 * Paging all Wiki-Mammalogists to provide some conversational feedback here... Bob the Wikipedian (talk • contribs) 00:40, 3 January 2011 (UTC)

Reasons why I wish we could abandon taxonomy and stick with phylogeny instead.... Seriously, though, this is just one of those curses we won't have a clear answer for until either some taxonomist has the balls to "officially" fix the taxonomy in a major publication (which many seem to be shying away from now since its more meaningful to talk about phylogeny), or until we finally switch over to PhyloCode. In the meantime, couldn't we just treat the these articles separately, with separate (conflicting) taxonomies, and then use the text of the article to briefly explain the contradiction and point to the alternative? Anyway, if you think this is bad now, just wait 'til we talk about Taxonomy/Lemuriformes, Taxonomy/Lorisiformes, and Taxonomy/Chiromyiformes. In fact, when you're ready to tackle those, just let me know. –  VisionHolder « talk » 01:50, 3 January 2011 (UTC)
 * I think you can consider Cetartiodactyla and Artiodactyla to be the same thing? Then Cetruminantia is part of it. All is well. I just don't think everyone will be too thrilled if we change to Cetartiodactyla in the taxoboxes, and it's not clear that such a change is ever going to happen for real or is needed? Check this out - Darren Naish knows where his towel is, I think, and he is essentially using the two interchangeably, putting Cetacea under (or over in his picture) Artiodactyla. ErikHaugen (talk | contribs) 05:03, 3 January 2011 (UTC)
 * That makes sense to me. Since Artiodactyla is what most of the project seems to use right now, I'd say let's stick with that until someone is bold. We're already upsetting enough people with this template; we don't need to pick more fights by forcing people to adopt Cetartiodactyla. Bob the Wikipedian (talk • contribs) 04:04, 4 January 2011 (UTC)

resolved

Ready to release?
Hi,

I've made some stability changes to the template and it now seems to work fully and intrinsically support taxon-dummying. I can't think of anything else that I'd need to edit, or any functionalities that I'd like to add, with the exception of automatically-generated subdivision lists and fossil ranges. Because these two features are "plugins" that don't require the core code to be edited, I'd like to suggest that no further edits are necessary. If this is the case then we could start rolling out the template a little further.

One optional change that could be made would be to replace " " in the templates with "  " (or "  "); these options would make back-back-end templates a little sleeker (making small, perhaps 1-2%, reductions in the overall template postincludesize, or some such statistic; and making the code easier to read; the words "machine code" could be dispensed with) at the cost of clarity to the user ("don't edit this line (((machine code|)))" would be replaced with "don't edit this line (((|)))" or  "don't edit this line (((1|)))"). This would be quite a bit of work now so I thought I'd gather opinion before enacting it.

Whilst we discuss that point, is anyone aware of any improvements or problems that should be fixed before we roll out this template more widely? Martin  (Smith609 – Talk)  23:47, 19 December 2010 (UTC)

The only problem I see at the moment with releasing it is that the automated taxon generator has been down ever since the bot was assigned its first automatic edit run. Bob the Wikipedian (talk • contribs) 06:04, 21 December 2010 (UTC)
 * Hmm, this needs substantial edits to bring into line with the modified syntax. I'm finding it quite easy to get along without -- do you think it's worth me investing the time to fix it?  Martin  (Smith609 – Talk)  08:36, 21 December 2010 (UTC)
 * It certainly made entering taxa much faster and more streamlined, but I can do without it. After all, I did before. Bob the Wikipedian (talk • contribs) 20:39, 21 December 2010 (
 * Okay, then let's consider the template "ready to go", and I can work on fixing the tool when I have the opportunity.
 * What is the best way to roll out the template? Subject to approval of WP:BRFA/Taxobot 3, a bot can upgrade taxoboxes where automating makes no difference to the output; but should we actively go out of our way to update other taxa?  Presumably we should discuss this elsewhere too.  Martin  (Smith609 – Talk)  13:01, 23 December 2010 (UTC)
 * I've not fixed the old version, but I've added a new function allowing a top-town approach that might make things quicker.  Martin  (Smith609 – Talk)  18:23, 2 January 2011 (UTC)
 * I think it's only logical that those who helped put this together continue working to automate the tree. I've successfully automated the 800+ most-linked-to taxa in WikiSpecies, with the exception of protists; Kleopatra has requested we leave those alone so she can do some research and straighten out that mess. I intend to continue automating taxa over the next few months. If we ever get all of the taxonomies automated, we'll have a major advantage over the EoL and several other databases (although I don't expect databases like PaleoDB or FishBase to be outsourced-- we don't keep track of sightings for most taxa, with only a small handful of exceptions).
 * As far as rolling out goes, I suggest we make an announcement at all relevant WikiProjects and encourage them to use the updated templates, show them the benefits of doing so, and make sure we keep an eye on the talk pages for all the involved templates-- as I'm sure we'll get lots more questions than we've already had once this goes mainstream. Bob the Wikipedian (talk • contribs) 21:20, 23 December 2010 (UTC)
 * The bryophytes also should not be automated. Neither Wikispecies nor Wikipedia has been fully updated with regard to mosses, and there is an imminent reclassification of the liverworts coming (major revision phylogenies came out circa 2000).  This isn't a minor issue, since there are 12000 species of moss.  You really ought to publicize this on the various taxonomic project pages and get a response from their tests first.  I only found out about this project earlier this month and haven't yet had the time to figure out how some of the problems will be solved (such as the problem of animal and plant genera with the same name). --EncycloPetey (talk) 06:53, 27 December 2010 (UTC)
 * Wouldn't that be a good reason to use automatic taxoboxes? Then it will be updated and it will be easier to implement the imminent reclassification? What is the problem with animal and plant genera of the same name? We have some of those, and I don't think it's been a problem. ErikHaugen (talk | contribs) 07:03, 27 December 2010 (UTC)
 * When the names, ranks, and number of taxa in a group face imminent major restructuring, there is nothing about implementing those taxoboxes now that will make anything easier. I say that based on all the work I've done (and still have to do) cleaning up the classifications for these same groups on Wikispecies.  Automating the taxoboxes will make consistency easier to maintain, but will not make it easier at all to modify a classification in the face of a major change.  In such a situation, it would be easier to implement as part of the change, rather than doing it all twice. I also helped on Wikipedia to effect the dissolution of some of the large polyphyletic Cronquist families of flowering plants.  Each genus had to be separately altered, and that wouldn't change with automation.  Only the species-level changes would be easier in that situation, and most plant taxa have yet to have their species articles written.  Think of it this way: What if the 27 orders of Neognathae we list were restructured into 35 orders in four classes, with some families elevated to ordinal rank, some  submerged into existing ones, some families swapping genera like baseball cards, and the names changing because of the Code's strictures about types and priority. That's what's recently happened to the 10,000 species of mosses and it's going to take me a long while to update all that on Wikispecies.  And they're not even the worst example of impending major rearrangement in botanical taxa.  Look at just the basic hierarchy on the red algae page.
 * All these changes seem like they would be a lot easier to do with automatic taxoboxes than with the current system. That is kind of the whole point – right now, we express these relationships in many places, but this system allows us to express each relationship once. So if a relationship changes, you only have to do change it in one place. You said "doing it all twice" which makes me wonder if you are just talking about the automated migration to automatic taxoboxes? ErikHaugen (talk | contribs) 17:18, 28 December 2010 (UTC)


 * The problem with plant and animal genera of the same name is: how will the automation know where to glean the information to display? There is no indication at all in the template documentation as to how this should work or what editors must do, and the few examples of this phenomenon I've examined have not had automated taxoboxes inserted.  If you can point to some example articles where this has been tried, I'd like to see how it's done.


 * I'd also like to know how fossil taxa like Protosalvinia would be handled, as it's not at all obvious. This genus has no certain classification, but is generally placed in the kingdom Plantae as a land plant.  Every other taxon of rank between kingdom and genus is uncertain as a result.  Or what happens when the order is known, but we don't have a division (phylum) or class in which to place the taxon?  The fossil group Bennettitales is an example of this (Ignore the current taxobox; it's rubbish.  You can easily see this by following the link to the listed class and division, which go to an unranked clade page whose taxobox is missing many of its children).  --EncycloPetey (talk) 07:45, 27 December 2010 (UTC)

By release do you mean you'll stop making edits to it, or that there are just a few edits left? These templates are in a lot of articles, and the automatic taxobox should not be released for general use throughout wikipedia until it is thoroughly checked and all editing is done. I would like experienced wiki coders to look it over and comment first. I would also like to know that no more editing will be done for some time. --Kleopatra (talk) 23:11, 23 December 2010 (UTC)
 * By release, I mean that as far as I can tell there is no further need to edit the template. I would be delighted if experienced wiki coders were to take a look over everything – any ideas who we could ask, or whether there's somewhere that we could put up a notice?
 * Ultimately the only way to check the template further is to try it out on more articles. We have it on almost a thousand, and there don't seem to be any problems with these, so this seems like evidence that nothing else will need changing.  Naturally if modifications are desired as we go, I'll be glad to make them if requested.
 * Incidentally, if I do upset you, feel free to let me know directly, so that I can apologise and clear up any misunderstanding. Martin  (Smith609 – Talk)  23:33, 23 December 2010 (UTC)
 * "Also, "Geologic" is in a horrible Americanism to my ears (as no doubt "Geological" is horrid to Americans)." How could that "upset" anyone?

You still are coming up with broken links in taxoboxes. I would like this to be checked and some time without broken links to pass before going full live with these. --Kleopatra (talk) 05:35, 27 December 2010 (UTC)
 * Good point. There's some problem with autotaxoboxes of "unranked" taxa. BTW what is the difference between "unranked" and "cladus"? ErikHaugen (talk | contribs) 07:03, 27 December 2010 (UTC)


 * This discussion has gone so many directions at once!!! I'll try to answer all the questions here...


 * Regarding Protosalvinia, does this work? This is how we're handling all incertae sedis taxa in all kingdoms. The only visual difference here is that "division" now says "phylum" due to the way the incertae sedis code has been set up, but this can be modified as needed. Protosalvinia will be grouped with all other taxa of indeterminate classification beyond Plantae. The whole thing works rather like most taxon databases I've seen across the Internet, only better, because we're handling incertae sedis and masking complex clades where appropriate, such as hiding Reptilia from the birds and mammals.


 * Regarding same-name taxa, we've already considered that. Turns out we've even got same-name taxa to describe the same creature-- and that got confusing for a moment. You'll find your answers at the naming conventions page where Erik has kindly documented our strategies for handling complex situations like these.


 * Regarding the withdrawal of comments, it makes discussions more confusing.


 * Regarding testing, Martin's correct. We've actually created far more templates (hundreds more) than what have been integrated into articles already. Only recently have we begun actually placing them into articles, but only after fully inspecting each one individually. I personally have tested over 1,000 test cases, and every single one of these tests is working correctly. Martin, Erik, and Marty have been testing quite a bit as well-- I'm not sure how extensive their numbers are, but with over 1,000 successful tests, we're confident at this point that the code has no bugs that would hinder normal or advanced usage. If someone has a special case that does produce a bug, such as possibly a triple-level uncertainty (which I don't think exists, but I've been wrong before), Martin, Erik, and I are standing by with this page on our watchlists to respond as soon as possible.


 * Release-- mostly when we say "release" we mean informing the general public about the template and encouraging its use. It's clearly being used on one or two hundred articles already, but since we're past the test phase, it makes sense to stop testing without implementing the taxobox. We've got a bot out there named Taxobot that will help implement the ones that have been tested but not applied (and it has instructions not to mess with anything that hasn't been user-tested). It's not in service yet, but its test edits were successful and I imagine Martin will be releasing the bot once we publicly declare the template "released".


 * Please remember when applying an automatic taxobox to preview the edit and resolve all missing taxa before saving the page. The preview button is our best friend!


 * The ichnobox, on the other hand, has not been tested extensively and still has at least one minor issue I haven't gotten around to working on.


 * Regarding "unranked" and "clade" -- no difference, as best I can tell. I would have set up the database with "cladus", but it appears zoologists on Wikipedia prefer "unranked" in the taxoboxes, so I used that instead for zoological taxa. Botanists seem to prefer "cladus", so that's what got used in those templates.
 * The only advantage of unranked_superordo (for example) is that the subdivision_rank of "order" can be generated from this parameter. Martin  (Smith609 – Talk)  14:19, 28 December 2010 (UTC)


 * Cycloneuralia had a problem, yes, but it wasn't an automatic taxobox problem. It looks like someone applied an undeveloped and untested template to the taxobox and that template broke when it got modified later. I've fixed the problem by removing the strange alien template (designed to display subdivisions) from the autotaxobox and replacing it with the traditional subdivision method. Bob the Wikipedian (talk • contribs) 19:58, 27 December 2010 (UTC)


 * Broken--Microschedia. --Kleopatra (talk) 15:31, 28 December 2010 (UTC)
 * Fixed, sort of? As part of something I'm doing I'm planning to generate lists of orphaned taxa, or taxa who point to parents that don't exist, like this, but it won't be done anytime soon. ErikHaugen (talk | contribs) 17:28, 28 December 2010 (UTC)
 * Thanks for the bug report, Kleopatra. I'd call that particular instance fixed (thanks, Erik), but you're right-- we don't have any method (that I know of) for tracking down orphans. I could be mistaken (caching might be an issue), but could we add a hidden orphan category to the error message? Would it work properly, or does the page have to be reparsed in order for it to be added to that category? Bob the Wikipedian (talk • contribs) 04:31, 29 December 2010 (UTC)
 * The bugs I find are all listed in categories associated with broken templates, they're usually related to recent edits to the code, although not always the direct edits. Also, most of them could have been caught with preview, as you point out above. Hence, when you're ready to stop editing the code is the time to begin rolling out the trial. And use the preview button. --Kleopatra (talk) 04:38, 29 December 2010 (UTC)
 * Suddenly it all makes sense how you keep coming up with them; thanks for that. I'll add it to my pages to monitor daily. I just had a look after you mentioned that at and noticed Calluna and Emblingia are in it. I took a glance at the two, but don't see what's wrong with them. Any ganders? Bob the Wikipedian (talk • contribs) 05:18, 29 December 2010 (UTC)
 * I think this has to do with both being monospecific genera with automatic taxoboxes at the genus and the species level. The species level taxoboxes are used in the articles, but their links to their articles are links to redirects. Emblingia also has a family level article. For plant articles, in monospecific genera, there is only an article at the genus or family level. There is no Calluna vulgaris, and there won't be one. Ditto Emblingia calceoliflora, it should be at the family level, though, not the genus. The automatic taxoboxes are done correctly, meaning the way they should work, in that they are tied from the article to the automatic species taxon, but they also include the genera authorities, and the family authority for Emblingia. This is the one way these two articles stand out among plant articles that may cause them to be in the category. --Kleopatra (talk) 05:51, 29 December 2010 (UTC)
 * p.s. The error you reported here wasn't preventable with the preview button. Rather, this is an instance where a template was moved and the person moving it didn't check to see first if it was in use. Folks, I can't stress enough the importance of seeing if a template is in use before moving it. Bob the Wikipedian (talk • contribs) 05:20, 29 December 2010 (UTC)
 * What template was moved? --Kleopatra (talk) 05:51, 29 December 2010 (UTC)
 * I made a null edit to Calluna and now neither are showing up, so I imagine that someone has fixed these. Thanks! Martin  (Smith609 – Talk)  10:11, 29 December 2010 (UTC)
 * It appears to have to do with the links to redirects in the species' templates in this case. I removed the link from the species in the Calluna vulgaris template (and Rkitko did the same with Emblingia calceoliflora template) that links to the species page which is a redirect to the genus page. When the templates were created, the Emblingiaceae template linked the family to the genus in the template link code. However, both species templates linked to the species' pages, both of which are redirects to the genus. I assumed, because the family was not showing up on the cleanup templates page, that its code was correct. This is a deeper coding error within the template, that linking to a redirect creates a cleanup. This should not happen, and it could possibly create other types of errors in the future as additions are made to the code. However, the templates, as they are created, should not link to redirects, and, in the write-ups for directions, this should be pointed out explicitly to users, as it is in no way obvious. --Kleopatra (talk) 16:08, 29 December 2010 (UTC)

I'm not sure what I just said will be clear to others. For plants, monospecific genera articles are under the genus name, and the species' pages are redirects to the genus. I think policy is that the a monospecific family article is also at the family, not the genus, but I'm not sure on this.
 * We had two monospecific genera, Calluna vulgaris and Emblingia calceoliflora.
 * Each species article is a redirect to the genus article.
 * Each genus article is an article.
 * In the case of Emblingia only, the family article is also a redirect to the genus.
 * All taxa, species and genera, correctly had templates made for them.
 * The two species templates were listed in the cleanup category.
 * Each species template had a link to the species article. But the species is a redirect to the genus.
 * Each genus template had a link to the genus.
 * The Emblingiaceae template was created with a link to the genus.
 * When the two species templates had their links to the species page redirects changed to links to the genera pages, they no longer appeared in the clean-up category, although a null edit had to be made to one to update its categorization.
 * This issue is due to the underlying code used to create automatic taxoboxes and is beyond my skills to debug, but it should be fixed.
 * The templates should not contain redirects in their links, however. This should be spelled out to users as it is a subtle demand for editors that is not readily seen. --Kleopatra (talk) 16:30, 29 December 2010 (UTC)
 * Just to clarify, the error was produced as a result of a redirect occupying the species's display link parameter?
 * Also, in answer to your question about what template was moved, causing the Microschedia error: On December 19th, Template:Taxonomy/Incertae sedis (Bilateria) was moved, or rather, deleted and recreated, at Template:Taxonomy/Incertae sedis/Bilateria in accordance with an emergency change in the naming convention for incertae sedis templates. This report tells me two things-- firstly, we need to be more careful when moving/removing templates, and secondly, that only Kleopatra has bothered to check the cleanup category. I'm adding it to my pages I'm checking daily in the hopes that errors like this won't continue for a whole week like that one did.
 * I'll echo Kleopatra's statement that redirects in the display link shouldn't be causing that error-- or if they do, the error could be far more specific, like "Redirected display link". The most useless error in the world is "something went wrong" (Thanks, Microsoft, I figured that much out). Descriptive errors are much more useful, and if we choose to place them into descriptive categories, those categories can then be placed within one another. Bob the Wikipedian (talk • contribs) 16:47, 29 December 2010 (UTC)

resolved

Vandalism
One of the higher level taxonomy templates was just vandalized, breaking nearly every taxobox for a short period of time. Perhaps some of them should be semiprotected or something? ErikHaugen (talk | contribs) 17:40, 22 December 2010 (UTC)
 * I permanently protected several of them last week...which one was it? Bob the Wikipedian (talk • contribs) 22:08, 22 December 2010 (UTC)
 * Found it. Taxonomy/Eukaryota has been semi-protected now. I don't think we need higher than semi-protection on any of the actual taxon templates, with the exception of the Life and Ichnos ones, perhaps. Bob the Wikipedian (talk • contribs) 22:20, 22 December 2010 (UTC)
 * I would say that each and every template affecting more than a 500 pages or so should be permanent protected till the time flagged revisions can be activated on all of those. -- Kim van der Linde at venus 02:38, 26 December 2010 (UTC)
 * Until...that implies nonpermanent protection. If we want simulated permanent protection until flagged revisions can be activated, then what needs to be done is the templates need fully protected with an expiry of some deadline we decide upon. But even then-- semi protection should be good enough, in my opinion-- we have to keep in mind the admins of Wikipedia aren't always the ones who know the tree the best. Do we have any idea yet of how long the trial will last? Assuming that the database will eventually match the depth and breadth of WikiSpecies (I don't expect it to, but it's possible), we're talking around 500 taxa with over 500 links pointing to them. More realistically, probably around 300 or 400 in the near future. Once we decide on a time limit, I'm sure we can get one of the template protection bots to start making rounds and protecting them with whatever expiry we decide upon. Bob the Wikipedian (talk • contribs) 04:06, 26 December 2010 (UTC)
 * As far as the number of taxa containing 500 or more pages, I would run that estimate much higher. Several insect orders have more than 100,000 members, and that means that all parent taxa to those orders (at all ranks) will need to be protected as well.  The same is true of a large number of flowering plant families (roses, buttercups, palms, grasses, sedges, mustards, composites, etc.) that have more than 1,000 species each.  We really should think about protection here in terms of the potential pages that could be affected, rather than counting current pages that would be affected, or else the protection monitoring will be an ever-lasting issue as more articles are added. --EncycloPetey (talk) 20:21, 27 December 2010 (UTC)

Speed and scalability
Can I ask if this system is going to be as fast as the convential taxobox and will it scale okay when it's in lots of articles? Regards, SunCreator (talk) 20:04, 27 December 2010 (UTC)


 * I'm not sure what you mean by scaling-- it takes up the same amount of space and looks virtually the same. The only visible difference is the text edit link that appears in the right side of the "Scientific classification" bar, where the entire classification of the organism is accessible, including all hidden taxa and even little-known clades that most people couldn't care less about. Load time is hurt significantly enough that several taxoboxes ( 6 to 10 4) could prevent a page from loading properly, but it's rare that there are even two on one page. Setting up an automatic taxobox takes a little bit more know-how, but modifications to the working taxonomic tree are quick and simple. For example, let's say overnight someone decides taxon A and its sister taxon B overlap, and that the entire scientific community agrees without any question (a rare sort of event, but ...only the immediate daughters of these two would need to be edited to link to taxon C, rather than the myriad of articles that would otherwise need edited. It's a little more technical, but it makes cleanup of erroneous taxonomies quick and simple. It also requires less research for the person setting up a new taxobox, in most cases.


 * I tested the post-cache load time of Tyrannosaurus with both versions of the taxobox (this ensured that the majority of the time being calculated corresponded to text parsing and template parsing) and came up with the following figures:
 * Tyrannosaurus article with Taxobox: 23.0 seconds
 * Tyrannosaurus article with Automatic taxobox: 29.5 seconds


 * And here is the load time for the taxoboxes without the article-- that article is huge, so I figured it must be taking up a lot of the load time:
 * Tyrannosaurus taxobox with no article: 5.0 seconds
 * Tyrannosaurus automatic taxobox with no article: 12.3 seconds


 * So, it appears that an automatic taxobox at genus level with a complex taxonomy adds between 6 and 8 seconds of server-end processing time to a page. It's significant enough to be noticed, but not so significant that it will adversely impact the page. Bob the Wikipedian (talk • contribs) 03:37, 28 December 2010 (UTC)


 * For someone who in a webmaster capacity finds 0.4 second page load speed slow. I find times of 6.5 or 7.3 second page load delay for essentially no improvement quite unacceptable. There are editors what will include/remove reference templates and spend ages writing references by hand because of the slight delay in page loading - see article pain as an example. So I guess they aren't going to be to happy either.
 * Scalability is the ability to handle growing amounts of work without delays, lockups, capacity limits and other downsides. In the taxobox we are perhaps going to get items nested 40 levels at a guess. It seems there are roughly 30 million species on earth, clearly they don't all have wiki articles with a taxobox at present, still are we going to encounter limits with an automatic taxobox? Regards, SunCreator (talk) 06:11, 3 January 2011 (UTC)


 * The only limits I foresee would be Wikimedia's fundraising abilities. So long as the servers continue to be upgraded as needed, I don't think this will cause a huge issue.
 * As far as nesting goes, we're overriding the 40 nests limit using some dummy nests, allowing something like 60 or 80 nested taxa, I believe. We've not hit the new limit yet...the old limit was 12, but we've improved the template drastically since then and we are now supporting well over 40 (Tyrannosaurus rex is successfully displaying 46 nested taxa using this new tricky sort of linked list traversal that seems to override the 40 limit).
 * As far as database size goes, the only thing affecting how quickly a parent taxon is found is the same factor that controls how quickly a client can navigate to a specific link on Wikipedia. The links are calculated, rather than searched for, making it highly efficient at locating the required data.
 * I haven't examined the code to calculate the Big O time yet, which I assumed would be O(n) or understandably at worst O(n)2. However, using a stopwatch, I found an average of about 1 second's load time difference between a top-level taxon and a 46th-level taxon. I tested around 50 or so standalone automatic taxoboxes tonight (with no fossil range/image/subdivision/etc), and all of them ranged from 6 to 7 seconds' load time; standard taxoboxes (using the same contraints) around 2 seconds (not sure what the difference is tonight, perhaps the servers are less bogged down?).
 * As for "essentially no improvement"; have you ever tried maintaining taxoboxes? They get outdated and eventually begin to conflict with one another terribly. This is an effort to keep all taxonomies up-to-date with minimal effort. Bob the Wikipedian (talk • contribs) 07:41, 3 January 2011 (UTC)
 * Well thanks for replying. I appreciate the technical brilliance of what your doing but I don't see the improvement for the reader. I can see some benefit to the editor, but nothing major to outweigh the slow pages. My main problem is speed. If the wiki servers can power it fast then fine. If it all works out nicely think about . Regards, SunCreator (talk) 01:52, 5 January 2011 (UTC)
 * I'll certainly keep that cladogram idea in mind. Thanks for the constructive criticism; this template most likely has a few corners that can be sped up here and there if we get the time to dissect it and inspect it closely. Bob the Wikipedian (talk • contribs) 06:34, 5 January 2011 (UTC)

Taxon parameter, moved pages
Can we change the advice at the automatic taxobox documentation to suggest that we always use ? There's such a learning curve with the automatic taxobox that I foresee innocent and uninitiated newb editors moving a page, maybe from a species name to a common name, or from a monotypic genus to the species, often with good intentions. Without the taxon parameter, the taxobox would break. Doing something routine like moving a page should not break the taxobox. I imagine most editors could handle it, but it's an added headache easily solved by insisting each automatic taxobox have the taxon parameter.

Also, I'm concerned about monotypic genera being moved. Is the way that Calluna is setup correct? I can see some well-meaning editor moving a monotypic genus article to the species name, unaware of WP:FLORA. On Calluna, I set ; Template:Taxonomy/Calluna vulgaris is now   and Template:Taxonomy/Calluna is just. If Calluna were moved to Calluna vulgaris, would both Calluna and C. vulgaris still be in bold in the taxobox? I imagine this would be hard to notice if the page was moved and even harder for an editor who knows nothing about this template to fix. Rkitko (talk) 14:16, 29 December 2010 (UTC)


 * It is set up correctly, yes. Bearing in mind the error Kleopatra pointed out, I don't think the error produced will be significant enough to impact anyone's browsing experience; a red flag will pop up on our category page, and we'll take care of it, but that's the only negative thing that I can see. However, you're right about the taxon parameter-- it would be appropriate for us to begin mandating it as usage policy. There are currently only a few hundred working taxoboxes out there that would need revised in reflection of this (bot task?), and once that's done, the code can be modified to break if a taxon parameter is left off. Bob the Wikipedian (talk • contribs) 17:00, 29 December 2010 (UTC)
 * I agree the templates were set-up correctly for these taxa.
 * Yes, usage of the taxon= parameter should be mandated to make a connection to the underlying code apparent for at least some editors, if for no other reason. I had not considered its potential as a fail-safe, but yeah, sure, good also.
 * Someone should move one of the two plant articles and see what happens. --Kleopatra (talk) 17:11, 29 December 2010 (UTC)
 * I agree that it's good to specify the "taxon=" parameter in many/most cases; however I think that the option not to should be retained; in some cases (obscure families that will never get a common name, for example) the taxon is (and will always be) redundant, and adds to editor burden when creating pages (something that I'm keen to avoid, partly for selfish reasons). Martin  (Smith609 – Talk)  21:37, 1 January 2011 (UTC)


 * Thanks re: the taxon template. I moved Calluna to Common heather briefly. The automatic taxobox was still fully functional, but Calluna was no longer in bold. Like I suspected, for a move to be successful without altering the style of the taxobox, Template:Taxonomy/Calluna would have had to be changed from  to   and that's just too complicated for most editors moving articles. Is there a way to preserve the chosen style (bold because it is the article on both the genus and species) without having to make multiple edits to back-end taxonomy templates when a page is moved? Rkitko (talk) 17:36, 29 December 2010 (UTC)


 * I suspected as much...drat. Perhaps there's a way to detect a redirect, but how, I have no idea-- unless we are able to trim the  off of a transclusion and then   it...would that do it? Bob the Wikipedian (talk • contribs) 01:02, 30 December 2010 (UTC)
 * Working on this at the moment...I expect I'll have at least some kind of result in an hour or so, depending on how quickly I find the line to change. Bob the Wikipedian (talk • contribs) 01:23, 30 December 2010 (UTC)
 * I've determined that Taxonomy appears to be the one needing modified, but I'm not sure how to go about extracting the redirect target page title. I've posted a cry for help at the reference desk--- probably the wrong place to ask, but I figure it's worth a try. Anyone know how to do this? Bob the Wikipedian (talk • contribs) 04:41, 30 December 2010 (UTC)


 * I do not think this is possible, but we can use the technique of m:Template:Link with self link simulation to avoid a link to Calluna, even if that page is moved. Alternatively, if this taxobox does not need a link to Calluna at all (such as in the case that this taxobox is not needed on any other page), an additional parameter of Template:Automatic taxobox can be given the value Calluna.--Patrick (talk) 10:47, 1 January 2011 (UTC)
 * Thanks, but I don't really see how it could be used...perhaps I'm missing the obvious, though. That made me think about this, though-- compare the display link's transcluded page text with the current page's display text. Unfortunately, I think it might cause a loop error or at least double or triple the load time-- in which case I wouldn't consider it any further. Bob the Wikipedian (talk • contribs) 04:30, 3 January 2011 (UTC)

Hold up. We've got two issues here: I agree with both of you, and let me clarify: Yes, it's undesirable to have nonbolded redirects. That part of the code needs modified. But yes, we should keep an eye out for redirects-- let's keep the error category there. The reason? Supposing a page is moved, a double-redirect is created that can't be spotted or corrected by a bot. If we keep an eye on the cleanup category (which, I'm learning, is more of a maintenance category than an error category-- there's not actually anything malfunctioning, just a chance for improvement), then we can update the links as they change. We need to look at this cleanup category as not necessarily an error list so much as a "to do" list suggesting pages to be improved. In order to clear up the confusion, I suggest we create two subcategories in the cleanup category-- one for actual errors and another for opportunities for improvement. So, two questions-- Bob the Wikipedian (talk • contribs) 04:39, 4 January 2011 (UTC)
 * Is there any disadvantage in just setting up a bot to watch for redirects in link parameters? The links should never be redirects to redirect pages anyway, ideally, so would still need fixing by hand even if the template could prettify them.  Martin  (Smith609 – Talk)  21:42, 1 January 2011 (UTC)
 * I'd support it. Bob the Wikipedian (talk • contribs) 03:30, 2 January 2011 (UTC)
 * What do you mean "The links should never be redirects"? There are plenty of situations I can think of where the taxon name is not the article name, and so a redirect must be used to avoid error.  For example Kelp is the article about the Laminariales.  The former is the common name and standard article name; the latter is the taxonomic name for the order.  This is also true for high-rank algal groups like Brown algae, where the common name is far more stable than the taxonomic one.  --EncycloPetey (talk) 01:12, 3 January 2011 (UTC)
 * Sorry to be unclear (I've now rephrased above). I meant that the target of the link should not itself be a redirect page: for instance if I link to [target|example], then the page at "target" should not redirect somewhere else.  Martin  (Smith609 – Talk)  02:23, 3 January 2011 (UTC)
 * But that's still not correct. The whole reason that redirects exist is so that piping doesn't have to be done all the time.  What you are saying now is that piped links must be used if the taxonomic name is a redirect rather than an article.  While there are situations where a piped link should (or must) be used, especially in cases of ambiguous names, there's no reason to forbid linking to redirects. --EncycloPetey (talk) 04:45, 3 January 2011 (UTC)
 * It's acceptable for piped links to point to redirects, but better for them to point to the right place (to avoid double redirects in the future, and to make it clearer to the reader where they're being directed to). The taxobox thus expects the links to point to the article when deciding whether to boldify higher taxa. Martin  (Smith609 – Talk)  04:56, 3 January 2011 (UTC)
 * I'm still not following your reasoning. Firstly, your latest reply seems to go in a completey different direction from before.  The issue before was that link targets shouldn't be redirects, but that's not true, as I pointed out.  Secondly, the logic of bolding or not can't be solved this way.  There exist quite a number of redirects out there (that probably shouldn't exist) that redirect a species link to a genus page only because the species article doesn't exist yet, so someone made a redirect instead.  If the taxobox is making the decisions, then it will be prone to making errors. --EncycloPetey (talk) 06:28, 3 January 2011 (UTC)
 * 1) Whether redirects should be permitted.
 * 2) Whether redirects should be bolded.
 * 1) Are there any categorizing pieces of code that might not fit specifically into one of these two cleanup subgroups?
 * 2) Are there any possible bugs that would result if we bolded the redirects as expected?

Omitting debatable infraorder
Is there a way to either omit a rank, such as "infraorder" for groups such as lemurs (including the Aye-aye) and lorises, or offer the option to say "See text"? In short, I have problems with Taxonomy/Chiromyiformes, Taxonomy/Lemuriformes, and Taxonomy/Lorisiformes, and you can read more about the issue at Lemur evolutionary history. Basically, there's conflicting taxonomies, but more importantly, few researchers are even addressing this suprageneric classification in favor of focusing on phylogeny. (Some of the people involved in the debate, such as Anne Yoder, considers taxonomy a waste of time. Regardless, she still disagrees with Groves' taxonomy and separation of the Aye-aye from the rest of the lemurs.)

I don't want to bias any of my articles (including some important FAs) towards one taxonomy in the taxobox. So what are our options? –  VisionHolder « talk » 02:52, 3 January 2011 (UTC)


 * In the case of the Aye-Aye, the infraorder is not displayed in the taxobox, so there's no need to omit it. Same with the Loris.  Can you give a concrete example of what you'd like to see rendered, so that I can wrap my mind around the problem?  Could a template like Template:Taxonomy/Incertae sedis/Mollusca (used to produce the "class" in Marocella) will do the job?  I'm sure that we'll be able to devise something!  Martin  (Smith609 – Talk)  03:24, 3 January 2011 (UTC)


 * If it could display exactly as it does on Lemur for that article (saying "See text"), I would be happy. I'm not really sure what I'm going to do with the Lorisiformes article at this point, so I have no idea.  (It might get moved to Lorisoidea, which currently redirects to it.  No matter what I do, I'm going to be taking sides...)  As for all the lemur articles, including the Aye-aye, I would prefer that they appear like they do for Ring-tailed Lemur.  Does that help? –  VisionHolder  « talk » 04:17, 3 January 2011 (UTC)


 * Template:Taxonomy/Lemuriformes/see text made it work for Lemur. Ring-tailed Lemur works with an automatic taxobox; Aye-aye does too. Martin  (Smith609 – Talk)  04:52, 3 January 2011 (UTC)


 * In examining the situation briefly, it seems to me that both systems could be used without conflict so long as only one system is applied to any given tribe/genus/species. Bob the Wikipedian (talk • contribs) 04:59, 3 January 2011 (UTC)


 * Do you mean let the lemurs use the Godinot taxonomy (1 infraorder) and the lorises use the Groves taxonomy (3 infraorders, only one of which would be visible to them)? If so, I guess we can do that until the loris research crowd starts attaching Groves, too. –  VisionHolder  « talk » 05:34, 3 January 2011 (UTC)


 * Exactly. It doesn't seem to me they would overlap within their own schools, so to speak. If lemurs use one taxonomy and lorises use another, why bother with lemurs and lorises fighting over which one is correct, when neither affects the other? Of course, I'm all for science moving forward and accepting new research if it is legitimate, but this is one instance where it seems that both might be able to survive together, at least on Wikipedia. If all the lorises are currently following Groves and all the lemurs following Godinot, take it and run. Fast. Bob the Wikipedian (talk • contribs) 05:51, 3 January 2011 (UTC)


 * I wouldn't say the lemur crowd agrees. (If you want the unpublished details, I'll share them.)  In fact, I prefer what is currently visible on the Lemur article -- a "See text" placeholder, either linked or unlinked.  –  VisionHolder  « talk » 06:13, 3 January 2011 (UTC)

Documentation
Just a note; the documentation's post-expand size is counted twice on Template:Automatic taxobox, because it is transcluded; therefore a maximum of two example taxoboxes can be supported. Martin  (Smith609 – Talk)  23:00, 3 January 2011 (UTC)
 * Odd, I just got four to work in my sandbox no problem. Bob the Wikipedian (talk • contribs) 05:12, 4 January 2011 (UTC)

Italics for "section"
I was surprised to see sections italicized in Ascophora. Is this the correct behaviour? Martin  (Smith609 – Talk)  20:13, 4 January 2011 (UTC)


 * No. Zoological sections should not be italicised, although botanical sections (a rank below the genus) should. --Stemonitis (talk) 20:44, 4 January 2011 (UTC)


 * Ah, I should have used "zoosectio", rather than "sectio". Thanks for putting me on the right track.  Martin  (Smith609 – Talk)  23:28, 4 January 2011 (UTC)

resolved

display_children_authorities
I'm currently creating a stub at Lokotunjailurus. I thought I'd try out the display_children, which worked beautifully, but there's one small problem-- I can't display the authorities for each species. Such a shame, especially since one of them is a new species and the other is a very hard-to-find-on-the-Internet type species. The easiest way, probably, would be for it to function much like the anticipated automatic fossil range-- checking all known children for an authority and displaying it if available. Until then, I'll be using the subdivisions for the lower taxa. Bob the Wikipedian (talk • contribs) 22:47, 4 January 2011 (UTC)
 * Yes, subdivision affords more flexibility. I tend to use bold species list in these situations. Martin  (Smith609 – Talk)  23:30, 4 January 2011 (UTC)
 * Eh, don't worry about this parameter, I just realized it would be useless anyway, since the references are really only required when they're not on the target links-- i.e. only for redlinks. So even if the parameter were developed, it wouldn't solve any problems. Bob the Wikipedian (talk • contribs) 06:41, 5 January 2011 (UTC)

resolved

μ
Why is it that every template I create instantly gets put into "μ — templates missing taxonomy, i.e. featuring links to create taxonomy"?

There are no create taxonomy links on the Template's page, and I've purged my cache on the category page, but the template still appears in the category. Bob the Wikipedian (talk • contribs) 19:48, 6 January 2011 (UTC)


 * I've been observing the category as you've been editing, and it seems that pages flash up as you edit them, then disappear shortly afterwards. A quirk of the Categorization algorithms?  A similar thing happened to me whilst I was creating taxonomies with Taxobot; sometimes as many as 300 pages would be needlessly in the category. No damage done, but it does make it hard to spot genuine errors.  Martin  (Smith609 – Talk)  22:35, 6 January 2011 (UTC)
 * Indeed. I'll update the category there to state that templates may hang there for a few minutes. Bob the Wikipedian (talk • contribs) 23:46, 6 January 2011 (UTC)

resolved

What is the authority
For this template Template:Taxonomy/Eukaryota? And why is it so different from what's in the article? Is the intention to change the article to this template? If so, please propose the change on the article talk page with the appropriate citations. --Kleopatra (talk) 05:28, 7 January 2011 (UTC)
 * Most likely the source was Animal, where Neomura is not displayed. I've modified the parent per your suggestion; thanks for pointing out the oversight. Please feel free in the future to add intermediate taxa where you see fit. Bob the Wikipedian (talk • contribs) 06:51, 7 January 2011 (UTC)
 * I reverted this edit. Although I agree with probably more that Cavalier-Smith has to say about the relationships among advanced bacteria, archaea, and eukaryotes, than most of my colleagues, wikipedia's article on neomura, and the current scientific acceptance and controversy of the classification does not support wikipedia organizing all life in such a manner. Please discuss this taxonomy in appropriate community areas if you disagree, before changing the template in this manner again. Thanks. --Kleopatra (talk) 07:30, 7 January 2011 (UTC)
 * Also-- the taxonomy template database is too incomplete to begin adding the calculated child lists to articles yet...you've asked protists not be entered, which would account for some of the missing child taxa, but remember there are millions of taxa, and we've only entered a thousand or so. Bob the Wikipedian (talk • contribs) 06:59, 7 January 2011 (UTC)
 * "Protista" is not and hasn't been currently supported as a clade for quite some time, so its particular absence from anything should only impact its own and other improperly written taxoboxes. --Kleopatra (talk) 16:17, 7 January 2011 (UTC)
 * Animal and Eukaryote are not synonyms, so I have no idea what you mean by the source was animal? How would animal be the source for the Eukaryote taxonomy? I would like to know the source for the Eukaryote automatic template. It is not the same as the two different taxonomies currently offered in the Eukaryote taxobox, so how does it arise? From what reference? --Kleopatra (talk) 07:10, 7 January 2011 (UTC)
 * Martin created that template; I don't know what reference he used; I'm just guessing he was adding the parents of a descendent taxon, saw Eukaryota at the top, and stopped. Bob the Wikipedian (talk • contribs) 07:54, 7 January 2011 (UTC)

Wait...which part of the template did you have a problem with, if it wasn't Neomura? Bob the Wikipedian (talk • contribs) 07:56, 7 January 2011 (UTC)


 * 1) What is the authority. We need some kind of common language here. These are taxonomy templates, so, when I ask who the authority is, I assumed you knew I meant who is the authority for the taxonomy. We don't make up taxonomies on wikipedia (well, we do, but we shouldn't), we take our taxonomies from the peer-reviewed literature and secondary and tertiary sources. When I ask who is the authority, I want a citation to the literature that lists this authority.
 * 2) For this template Template:Taxonomy/Eukaryota? which lists Bikonta, Centrohelida, Plantae incertae sedis, Plantae, Telonemia, and Unikonta as subgroubs.
 * 3) And why is it so different from what's in the article? Why are the template subgroups different from the Eukaryota article subgroubs? The article lists these two subgroupings in its taxobox: Kingdoms: Animalia, Fungi, Amoebozoa, Plantae, Chromalveolata, Rhizaria, Excavata or Unikonta and Bikonta
 * 4) Is the intention to change the article to this template? Is there a difference between the two that will change the taxonomy as it is in the current taxobox? Or am I reading this template wrong?
 * 5) If so, please propose the change on the article talk page with the appropriate citations. If the intention is to change the taxonomy in the current taxobox, please discuss on the article talk page and cite your reasoning by referencing specific literature. In the case of taxonomies, at least secondary or tertiary, no primary literature. --Kleopatra (talk) 16:17, 7 January 2011 (UTC)
 * I don't have an answer for question #1; Martin put that one together. I don't usually add a reference for the parent if it came straight from the Wikipedia article, though if I used WikiSpecies or another reference I'll usually mention it (though I often forget).
 * In answer to questions 2 and 3, the subdivisions listed on the template page have nothing to do with the template itself (the only thing coded onto that template is the parent, rank, and display link, not the daughters); rather, these represent other templates which have been linked to it, each likely using a different system. This is one flaw with the calculated display_children parameter-- not all the children linked are using the same system. This is the result of using taxonomies which have been placed into articles and modified as agreed upon by various WikiProjects. It is not our goal as template implementors to normalize the taxonomies such as you're proposing, merely to link them all so that they can be normalized more easily.
 * In answer to question 4, I have no intention of placing nonmatching subtaxon lists into articles. I'm only doing that in the rare event that the calculated daughters are the same as those listed already.
 * In answer to your final question, we're merely automating what already exists, using the articles on Wikipedia as a source and the WikiSpecies as a secondary reference, with Wikipedia's existing taxonomies taking precedence over WikiSpecies's. We have no intent of changing what's already there-- editors have battled over it before and it doesn't seem right to just barge in and say, we're changing it to this. In a few instances, we've found it necessary to normalize a taxonomy due to its odd nature, and we never proceed with that until we've notified the members of the relevant projects and gotten their opinion first. Recent examples-- Sauropsida vs. Reptilia, Theropoda vs. Amniota as a parent for Avialae, Cetartiodactyla and Cetruminantia. Bob the Wikipedian (talk • contribs) 21:08, 7 January 2011 (UTC)
 * Then why does this template exist, what does it do, if it doesn't display these taxa in an article?
 * WikiSpecies taxonomies should not be used on wikipedia, they don't have consensus here; they use major outdated nomenclature, up to 50 years old, and, they used to be the home base of the Cavalier-Smith rah rah club on wikipedia, producing original and completely unsupported taxonomies. These should not be spread anywhere. --Kleopatra (talk) 03:19, 8 January 2011 (UTC)

Kleopatra has a valid point. I would summarize it thus: If someone changes a taxonomy in a (old school) taxobox in a manner that I dispute, my first act is to ask for the change to be sourced. If someone changes the taxonomy in one of these embedded templates, thereby changing numerous automatic taxoboxes in a manner that I dispute, how shall I proceed? Hesperian 05:43, 8 January 2011 (UTC)


 * Kleopatra-- Taxonomy/Eukaryota does several things-- It holds, as you are likely aware, the data necessary to link it to Taxonomy/Life (or whatever parent is appropriate-- in this case, no parent is). It holds the information necessary to tell the automatic taxobox that Eukaryota is a domain. It tells the automatic taxobox to display Eukaryota using the link Eukaryota.
 * Additionally, the template serves as a parent to any other templates that are instructed to link to it. This information is embedded in the parent parameter of each daughter-- entirely separate from the code on the Eukaryota template. The navigational table shown on the right side of the template's page is just there to help navigate from one template to another-- very little of that table's data is affected by modifying the template.
 * Hesperian-- I agree. WikiSpecies does use different systems quite often. However, it includes many noncontroversial intermediate taxa left out on Wikipedia, which usually don't hurt to add in. Taxa displayed in the "old-school" (to borrow your phrase) taxoboxes serve as my primary guide, though, when working with this. If the WikiSpecies taxonomy conflicts (often the case), Wikipedia's taxonomic system wins.
 * Sourcing unclear references in these templates is a good policy which I try to maintain but often forget. I'll do my best in the future to cite the source if it's not straight from an old taxobox. Bob the Wikipedian (talk • contribs) 08:28, 8 January 2011 (UTC)
 * This is exactly how WikiSpecies gets it wrong: their primary taxonomy is an old one, then they add in new higher order taxa created for newer taxonomies. So, if you're adding missing taxa from WikiSpecies to Wikipedia articles, you are creating totally original taxonomies. The taxa are missing from wikipedia articles, because the taxonomy used in the taxobox is referenced to a newer taxonomy, and the missing taxa have been deprecated, often to a different Linean level or often subclade, sometimes superclade.
 * I would like to have time to find a specific example, but, what wikispecies does is, if an old taxonomy has a class with 5 orders, and the newer taxonomy has a new name for the class with one order subsumed one in another and has only 4 orders, wikispecies will sometimes use the newer taxon name and its primary reference, but with the 5 orders from the older taxonomy. The old order may be used in older taxonomies, or it may have been deprecated to family status. By adding "intermediate taxa left out on Wikipedia" you are invalidating the taxonomy and its references used in the article and creating brand new original taxonomies sourcable only to you and wikispecies. Please don't do this, and please remove any such additions you have made to wikipedia that are not directly tied to a specific reference.
 * A taxonomy is not a collection of all "noncontroversial intermediate taxa." There are plenty of noncontroversial intermediate taxa running around that don't belong in specific taxonomies. And, wikipedia's collective ability to identify controversies in taxonomies means that finding the controversies takes a lot of research outside of wikipedia. --Kleopatra (talk) 16:24, 8 January 2011 (UTC)

Hesperian asked how to correct this problem when he/she sees it. I agree that it is quite a problem when an automatic taxobox changes multiple articles incorrectly. It is not easily seen how to correct the templates. The explanation given is not comprehensible, and it is an explanation, rather than instructions. --Kleopatra (talk) 16:31, 8 January 2011 (UTC)


 * I'll have a look at the documentation and see if I can't doctor it up a bit-- you're right, proper documentation should allow anyone with the level of knowledge required to operate a taxobox to operate this. Regarding WikiSpecies, I'd like to think that the regular editors there are just as honest as the regular editors here and would not post original research there. I'm going to go work on that documentation now. Bob the Wikipedian (talk • contribs) 22:06, 8 January 2011 (UTC)
 * I have no idea what their policies are, but their taxonomies are mixed and unsourced to the actual references linked on the taxon pages. They cannot be used in wikipedia. You can go to their sources and see the issue. --Kleopatra (talk) 22:40, 8 January 2011 (UTC)
 * YIKES....that is all technical jargon. This might be a bigger chore than I planned for. I'll let you know when I've finished. Bob the Wikipedian (talk • contribs) 22:08, 8 January 2011 (UTC)
 * Good luck. --Kleopatra (talk) 22:40, 8 January 2011 (UTC)
 * Thanks, I'm gonna need it...would you mind taking a look at Template:Automatic taxobox/doc/Step-by-step and letting me know if I'm on the right track here? It's a work in progress, entirely unfinished, but if it doesn't look useful at all then I'll abandon it and try another format. Bob the Wikipedian (talk • contribs) 05:05, 9 January 2011 (UTC)

Automating a null edit
I attempted to create a null edit button for the Taxonomy/ pages, but quickly found there's a difference between a null edit and a purge (the difference being that Category:Automatic taxobox cleanup isn't updated). Is there/could we create a template similar to purge button that would accomplish this much-needed null edit?? Bob the WikipediaN (talk • contribs) 03:06, 28 January 2011 (UTC)
 * I've created Template:Null edit and linked to it from Template:Taxonomy key; a link will now display underneath the info table on each Taxonomy/ subpage. Martin  (Smith609 – Talk)  23:43, 7 February 2011 (UTC)
 * Wonderful! It works! Tiny and hard to find, but it definitely works! Bob the WikipediaN (talk • contribs) 03:13, 8 February 2011 (UTC)
 * Feel free to move it as you see fit. Martin  (Smith609 – Talk)  13:47, 10 February 2011 (UTC)

resolved
 * Since it's not crucial (Wikipedia's cache does get updated eventually), I'll leave it. It's probably best small and unobstructive. Bob the WikipediaN (talk • contribs) 22:45, 11 February 2011 (UTC)

One to make?
80 transclusions of Template:Taxonomy/Irenesauripus. 134.253.26.9 (talk) 22:08, 7 February 2011 (UTC)
 * The "what links here" is giving me zero hits. Are you sure this isn't resolved? Bob the WikipediaN (talk • contribs) 02:54, 8 February 2011 (UTC)
 * Yes, it looks like someone fixed it. Recently, it had over 90 transclusions per this report. Plastikspork ―Œ (talk)  03:14, 8 February 2011 (UTC)
 * Sorry, forgot to mark as fixed. The deleted page was listed as a child of "Unclassified dinosaur trackways", so I updated the child list. Martin  (Smith609 – Talk)  03:17, 8 February 2011 (UTC) resolved

Update sister taxa downage
"Update sister taxa" has been down now for probably about a week. Is this intentional? Bob the WikipediaN (talk • contribs) 03:24, 8 February 2011 (UTC) resolved

Italics in Apatosaurus title
For some reason, the taxobox title at Apatosaurus isn't being italicized. Adding the name = parameter doesn't seem to work and I can't find any coding differences to the Tyrannosaurus box which has the italics. Anybody know what's going on? MMartyniuk (talk) 23:02, 11 February 2011 (UTC)
 * The bug appears to have deeper roots than this week. I'm looking into it. Bob the WikipediaN (talk • contribs) 23:55, 11 February 2011 (UTC)
 * This is the fix. Martin  (Smith609 – Talk)  01:39, 12 February 2011 (UTC)
 * Sneaky....I wondered if perhaps that was the culprit; looks like you fixed that before I was able to test that theory. My calculations show the name parameter shouldn't even be needed now that that's fixed. Thanks, Martin! Bob the WikipediaN (talk • contribs) 02:26, 12 February 2011 (UTC)
 * Thanks guys! MMartyniuk (talk) 03:03, 12 February 2011 (UTC)

resolved

Breaks usage of navbox template
Something with a recent change in this template has caused the navbox template to malfunction, see Mus (Mus). At the bottom of that article, the navbox template Template:Murinae appears as redlinked "Navbox". -64.85.216.158 (talk) 17:20, 9 January 2011 (UTC)
 * The page got an overly large post-expand include size, no doubt because of the automatic taxobox. I've returned it to the normal taxobox. Ucucha 17:24, 9 January 2011 (UTC)


 * Report for the Automatic taxobox at Mus (Mus):

NewPP limit report Preprocessor node count: 48777/1000000 Post-expand include size: 252809/2048000 bytes Template argument size: 410639/2048000 bytes
 * Report for Template:Murinae:

NewPP limit report Preprocessor node count: 18267/1000000 Post-expand include size: 1487602/2048000 bytes Template argument size: 1253809/2048000 bytes
 * So I'm trying to reduce the size of Template:Murinae. Martin  (Smith609 – Talk)  17:54, 9 January 2011 (UTC)

Hornworts
OK, here's a challenging problem I haven't seen addressed before now. First off, the hornworts taxobox will need to be updated, so please don't point out the discrepancy between the current outdated taxobox and the situation I'm going to describe. Also, please don't start moving articles around simply because a discussion has started. I am seeking feedback and a consensus on how to deal with the issue before taking any action, and this situation or something like it could arise again.

For the Hornworts (division Anthocerotophyta), there are two classes. One class is the Leiosporocerotopsida, with the single species Leiosporoceros dussii. The remaining 100-200 species of hornworts are in the other class Anthocerotopsida. Clearly the former taxon will, by usual standards, be a page about the species, will be located at the genus name Leiosporoceros, and will cover all taxonomic ranks from class to species. The problem is how to treat the other class. The standard solution when there are two classes is to have a separate page about each class. However, this would result in essentially a duplicate article to the one about hornworts, since the circumscription of the division and larger class differ by the inclusion or exclusion of a single oddball species. The more logical solution would be to redirect the class to the division page. This also makes historical sense, since in times past the bryophytes were treated as a single division and the hornworts were ranked at the level of class under that name. I don't like the idea of a nearly duplicate article on practically the same subject, but our current taxobox setup seems to force the issue.

Opinions? Suggestions? --EncycloPetey (talk) 20:26, 15 January 2011 (UTC)


 * When creating the taxonomy templates, I'd set the target for the link parameter to the article on L. dussii, but set the display text for the link to the taxon name. That way, the taxon will appear bolded in the L. dussii article. Also, you'll probably need to use display_parent if there are any intermediate taxa needing displayed. Does that answer your question? Bob the Wikipedian (talk • contribs) 21:18, 15 January 2011 (UTC)
 * No, it doesn't address my question at all. --EncycloPetey (talk) 16:50, 16 January 2011 (UTC)


 * I think I agree with you, EncycloPetey, that redirecting Anthocerotopsida to hornwort is appropriate here. Any point on which the two differ can be mentioned in the text – e.g. "Hornwort spores are 30 μm to 80 μm in diameter (except in Leiosporoceros, which has smaller spores)". I assume they'll be rare enough that the text won't be swamped by parenthetical exceptions. --Stemonitis (talk) 07:06, 16 January 2011 (UTC)


 * I'm afraid I don't understand the underlying question, then. Bob the Wikipedian (talk • contribs) 22:38, 16 January 2011 (UTC)
 * There are two considerations: one practical and one technical that come out of my suggestion. On the one hand, it means that (of the two classes) one is redirected to a much lower rank (genus) since there is a single genus in the class, and the other is redirected to a higher rank (division/phlyum) since the other class includes almost all the species and is nearly identical in composition to the higher rank.  The technical problem that comes out of this is how to implement the automated taxobox for the page of the genus Leiosporoceros.  That page's taxobox must includes authorities and citations for every rank (class, order, family, genus, and species) and these authorities differ at almost every level.  Further, is there any sensible reason to bother with automating the taxobox in a situation like this, where all the information occurs on a single page anyway? --EncycloPetey (talk) 16:58, 17 January 2011 (UTC)


 * I think I agree with Stemonitis that it must be, for the reasons EncycloPetey mentions, that the class is a redirect to the division. But how will that mess up the automatic templates, since the other class will also be a redirect, although to a species as is standard? --Kleopatra (talk) 23:16, 16 January 2011 (UTC)
 * Some time ago, we discussed (but did not really resolve) a kind-of-similar issue involving the genus Cryptoprocta, which has a well-known living species (Cryptoprocta ferox, the fossa) and a quite poorly known extinct species (Cryptoprocta spelea). Ucucha 01:12, 17 January 2011 (UTC)


 * Looks like the only thing keeping us from duplicating it exactly as it appears is the absence of a greatgreatgrandparent_authority. I'll do some experimenting in my sandbox and see if this is feasible to add support for. Thanks for pointing out this rare quirk. Bob the Wikipedian (talk • contribs) 05:00, 18 January 2011 (UTC)


 * A related potential issue: How would this inclusion of greatgreatgrandparent_authority be affected if an additional rank were inserted into the hierarchy at a later time, and how would editors spot the problem without being nomenclatorial experts? --EncycloPetey (talk) 05:08, 18 January 2011 (UTC)


 * I've added the necessary parameter to display the greatgreatgrandparent to automatic taxobox, taxobox/core, and taxobox/taxonomy. Updating the documentation shortly. As for your "potential issue", I actually though of that as well while carrying out the edit. I've not got an answer for that. That problem might be a little more difficult to prevent than the issues caused by pagemoves. Since display_parent isn't being used in this case, the only way it would affect this taxobox would be for a taxon to be added at the subgenus level, however, so I don't think this heirarchy will be as heavily affected as some others might be. Bob the Wikipedian (talk • contribs) 05:34, 18 January 2011 (UTC)


 * In this instance, you're probably better off sticking with the manual taxobox, rather than trying to automate it; we could shoehorn the information into an automatic taxobox, and we could modify the automatic taxobox to accommodate it, but there's no need. The standard taxobox already accomplishes everything we might want it to. Automatic taxoboxes are potentially useful for densely populated taxa, but of little assistance in cases like this. --Stemonitis (talk) 07:02, 18 January 2011 (UTC)

resolved

Upgrading to the next version of str find
Per the recent modifications to strfind short, it has been advised this faster and more capable template take the place of str find where possible. Because PAGENAMEBASE is a structural template at the foundation of the code of the automatic taxobox, this edit could potentially cause havoc if not evaluated first for soundness.

I'm requesting anyone who has enough know-how to evaluate this proposed modification respond to the RfC listed at Template talk:PAGENAMEBASE. The template is being used exclusively for the automatic taxobox. Bob the WikipediaN (talk • contribs) 02:49, 5 February 2011 (UTC) Retraction made at 03:24, 6 February 2011 (UTC) by Bob the WikipediaN  (talk • contribs) 03:24, 6 February 2011 (UTC) with explanation: Actually, it's used in all the taxoboxes, so I'm posting this notice on those talk pages as well. done

Bird
Now shows class Amphibia above class Aves - can someone look at it please. Shyamal (talk) 09:26, 10 February 2011 (UTC)
 * Looks like it was this edit which did it (now reverted). This would appear to be another disadvantage of the automated system – that a well-meant edit may have unexpected consequences. I don't understand enough about vertebrate cladistics to comment on what Petter Bøckmann was trying to achieve. --Stemonitis (talk) 09:45, 10 February 2011 (UTC)


 * Mea culpa. I was trying to tidy up the taxoboxes of the various labyrinthodont groups. The problem is that order/subclass Reptiliomorpha is anchored directly to superclass Tetrapoda, rather than to the more conventional Labyrinthodontia/Amphibia.


 * We need to have a look at the structure of the whole tetrapod tree. The problem is that it now shows a mix of Linnaean and phyllogenetic taxonomy. Now, they can partly mix, but it requires a bit of discretion. The basic problem is that someone has anchored amniota in the reptile-like amphibians, undoubtly to make phylogeny of e.g. birds make sense. This is of course phyllogentically correct, but it does upset the whole lower tetrapod taxonomy. Aminota is a non-Linnaean unit, linking it to Reptiliomorpha, which is a Linneaean unit makes all the problems. I suggest anchoring amniota directly to tetrapoda in stead. Tetrapoda is a non-Linnaean unit as well, so it will solve the bird-amphibian problem. Anchoring amniota to tetrapoda will stil be phyllogenetically correct, without upsetting the Linnaean system. Petter Bøckman (talk) 11:39, 10 February 2011 (UTC)


 * Well, if there's no opposition, I'll edit Amniota to anchor to Tetrapoda. That will solve the problem. Petter Bøckman (talk) 16:50, 10 February 2011 (UTC)
 * I think the solution here is the dummy taxonomy templates like Template:Taxonomy/Avialae/skip, which is used to keep birds out of Theropoda, etc. Doing this here would preserve the integrity of the tree (although it makes it a little less obvious how to read it) and would allow us to put "Reptiliomorpha" in Amniota's taxobox, all without putting "Amphibia" in Aves' taxobox. ErikHaugen (talk | contribs) 17:35, 10 February 2011 (UTC)


 * I'll agree with Erik here. That's being done for Mammaliaformes and for Avialae (although the naming convention for Mammaliaformes is a bit different-- "/Amniota" instead of "/skip". "/skip" makes more since, sense Amniota could technically change (though it's unlikely). Bob the WikipediaN (talk • contribs) 06:22, 11 February 2011 (UTC)


 * There's really no end of potential problems here. The tetrapoda tree is really either anchored in Ichthyostegalia (an order) or in Tetrapodomorpha (fish-grade relatives) depending on whose naming convention one follows, which again is a subset of crossopterygii, again a part of Osteichthyes etc. I edited Amniota to anchor directly to Tetrapoda, but the "skip" alternative sounds more elegant. I have had a look at it, but must admit it is a tad too technical for me. Would some of you who know this stuff please make the relevant "skip" templates? Just remember that Reptiliomorpha now anchor to Amphibia and should remain so. Petter Bøckman (talk) 07:55, 11 February 2011 (UTC)


 * I've created taxonomy/Reptiliomorpha/skip and linked Amniota to that, which will cut out anything and everything that falls between Tetrapoda and Reptiliomorpha. Peter, you should now be able to add whatever parent you like to taxonomy/Reptiliomorpha without affecting any of the amniotes. Let me know if you still need assistance! Bob the WikipediaN (talk • contribs) 08:57, 11 February 2011 (UTC)


 * It now works splendidly at the "lower" end, thank you! As for the birds and other amniotes, I really feel that the taxobox should link amniota to tetrapoda instead of shunting it via the reptile-like amphibians, preferably with a "skip" solution as it allows for the full phyllogenetic tree for those who want to see it. The origin of Amniota is otherwise made very clear in the introduction on the Amniota page. As of now, the bird and reptile pages show a mixture of Linnaean and phyllogenetic classification, which is really neither here nor there. I would think that birds/reptiles/mammals linking to amniota, then to tetrapoda would make for a cleaner taxobox, but I guess I'll run afoul with the cladisticans here. Also, I wonder why emphasis is put on the somewhat technical Amniota and Tetrapoda, while the very well known Vertebrata is left out. From a laymans navigational aid POW, I would have preferred birds/reptiles/mammals linking to Vertebrata rather than the more obscure Amniota and Tetrapoda. Petter Bøckman (talk) 09:35, 11 February 2011 (UTC)
 * Sounds like an RfC to me. Care to host it? I'm probably starting to worry folks with the number I've opened recently. Bob the WikipediaN (talk • contribs) 22:41, 11 February 2011 (UTC)
 * "As of now, the bird and reptile pages show a mixture of Linnaean and phyllogenetic classification, which is really neither here nor there." Except that many of the "strictly Linnean" groups you're talking about have been converted and given phylogenetic definitons. Though obviously the phylogenetic definition and the Linnaan concept will often conflict (see Class Aves vs. crown clade Aves). We should use whichever conception is most prevalent in the field, regardless of how this will trickle down the tree. Obviously, what labyrinthodont workers are using doesn't matter much for the Bird article since many of those probably won't be visible (or will be paraphyletic with skips, etc.). Is it the case that a majority of amphibian paleontologists are using a strictly Linneaen system?MMartyniuk (talk) 23:09, 11 February 2011 (UTC)
 * Actually it's a mixed bag, with Laurin (obviously) sticking to phylogenetic nomenclature, Carroll to classic (I'd say almost Romerian) Linnaean and Clack and Anderson and falling somewhere in between, though closer to Linnaean than to PhyloCode standard. Benton of course is doing his own thing. Much of the work in the group is older though (Romer, Säve-Söderberg), and obviously Linnaean. The basic textbooks dealing with the group are going to be Bentons, Romers, Carrolls or Colberts, or be strongly influenced by them, so any non-professional coming to Wikipedia is likely to be looking for the classic groups. My rationale to unbashfully stick to "fuzzy tree" Linnaean nomenclature in the taxo-boxes is that the labyrinthodont tree actually is fuzzy. I have tried to cover the various aspects of systematics in Labyrinthodontia (have a look!), though the taxo-boxes remain staunchly Linnaean. Petter Bøckman (talk) 23:40, 11 February 2011 (UTC)


 * I'll need to brood on the arguments first. This is going to go to the heart of the Linnaean/cladistic dispute, which can be heated at times.Petter Bøckman (talk) 23:00, 11 February 2011 (UTC)
 * Well-- I'd argue that Reptilia is no longer used and that Aves are now classified as a subtaxon of Sauropsida. But I'll respect the request of the relevant WikiProjects and leave Sauropsida out. <span style="font-family:linux libertine o, times; font-variant:small-caps">Bob the WikipediaN (talk • contribs) 23:30, 11 February 2011 (UTC)

resolved

Stem groups
Are stem groups to be treated differently? I was under the impression "stem group" was a contextual term that doesn't really refer to a taxon rank.

I'll point to the following glaring areas I've noticed within the past few minutes--
 * 3 instances of Trilobita at taxonomy/Trilobitomorpha
 * 3 instances of Arthropoda at taxonomy/Ecdysozoa
 * 2 instances of Scalidophora at taxonomy/Ecdysozoa

Thanks! <span style="font-family:linux libertine o, times; font-variant:small-caps">Bob the WikipediaN (talk • contribs) 23:00, 11 February 2011 (UTC)
 * There shouldn't be any difference besides how the definition if formulated. I suspect somebody is adding each conflicting definition of the same taxon. They should settle on one until PhyloCode standardizes them officially. MMartyniuk (talk) 23:04, 11 February 2011 (UTC)


 * The stem group taxa list the host taxon with the rank "stem-group". They list as children in the back-end templates, for completeness, but do not show up in the child lists in mainspace. Martin  (Smith609 – Talk)  01:35, 12 February 2011 (UTC)
 * I've added a disambiguation in the child lists, too, in this edit. Should make things more user-friendly. Martin  (Smith609 – Talk)  20:46, 12 February 2011 (UTC)

resolved