User talk:Taxobot

You can use this page to report bugs, to make suggestions, or to discuss the bot's functioning.

"Unable to log in"
This is my first attempt since the automatic taxobox template has been updated to use the toolserver.

Edit summary: +Taxonomic info from "Existing taxobox"; User: User: Session was not initiated by valid user; edit will not be completed

I did have "Bob the Wikipedian" entered in the username field on the initial page. Bob the Wikipedian (talk • contribs) 21:50, 2 November 2010 (UTC)


 * Oddly, the second one I tried worked. Looks like a wonderful tool. Just a shame I couldn't complete it the first time. I'll keep you posted if this happens again. Bob the Wikipedian (talk • contribs) 21:57, 2 November 2010 (UTC)

It happened again. Resolved by hitting "BACK", then reloading the form page. This involves resending the article name and username data via the Resend Data button, and requires the user to re-enter all taxonomy data. Bob the Wikipedian (talk • contribs) 03:11, 3 November 2010 (UTC)

"You've linked..."
Taxonomy exists at Actinobacteridae. [Article] [Talk] [Taxonomy template] Thank you! You've linked Actinobacteridae [Article] [Talk] [Taxonomy template] to the tree of life.

If the taxon template already exists, I'd prefer to see a message stating it has already been linked, not one that says I've just linked it. Bob the Wikipedian (talk • contribs) 23:49, 2 November 2010 (UTC)

Downtime
It would seem the Taxobot has been down for a good amount of time now. I click the first "Submit" button and the page attempts to load, and after a few minutes, the page finally gives up loading, but returns no errors, just a white page. Bob the Wikipedian (talk • contribs) 00:27, 3 November 2010 (UTC)
 * Up again. Bob the Wikipedian (talk • contribs) 00:40, 3 November 2010 (UTC)
 * Down again. Is perhaps my query for Aplacophora crashing it? That's been the first one to crash both times it's gone down. Bob the Wikipedian (talk • contribs) 00:43, 3 November 2010 (UTC)
 * Aplacophora seems to be the bot-crasher. I'll leave this one uncreated for the time being until you work out what's wrong. Bob the Wikipedian (talk • contribs) 01:58, 3 November 2010 (UTC)

Source
When no taxobox can be found as a source, the "Data source" text field should not load the words "existing article". Bob the Wikipedian (talk • contribs) 03:08, 3 November 2010 (UTC)

Default Enter_pressed event
The default "enter_pressed" event should be pressing submit button, not the following link to the Bot's userpage. I typically hit enter if I've typed something in one of the fields. Bob the Wikipedian (talk • contribs) 04:32, 5 November 2010 (UTC)

Crashing
It appears the bot crashes if the target taxon has an article but no taxobox. Bob the Wikipedian (talk • contribs) 06:38, 5 November 2010 (UTC)

Cannot create incertae sedis taxa
Display in ALL children: false Taxon: Incertae sedis (Alphaproteobacteria) Display text: Incertae sedis Link to: Incertae sedis Rank: ordo Parent: Alphaproteobacteria I've verified the above: true Data source: WikiSpecies

Please match the requested format: The taxon's complete name, e.g. Uca speciosa (crab). Please match the requested format: The taxon's name, as it will be displayed in the taxobox, e.g. U. speciosa.

Bob the Wikipedian (talk • contribs) 21:19, 21 November 2010 (UTC)


 * Similarly, "Dictyoptera (cohort)" is invalid as a taxon parameter. Bob the Wikipedian (talk • contribs) 23:09, 21 November 2010 (UTC)

Minor cosmetic issue
When the taxobot displays the taxoboxes for Culicidae, the synonyms and diversity appear before the authority, and there is no opening set of braces to match up with the end of the synonyms/diversity. This is not hurting anything; it just looks quite broken. Perhaps knowing this will help locate the source of another problem. Bob the Wikipedian (talk • contribs) 22:41, 21 November 2010 (UTC)

Down
While implementing taxoboxes, the template-generator toolserver appears to be down. Bob the Wikipedian (talk • contribs) 04:50, 26 November 2010 (UTC) undefined


 * Has the wizard for creating templates been taken down entirely? I don't see it when I go to http://toolserver.org/~verisimilus/Bot/taxobot/taxobot.html Bob the Wikipedian (talk • contribs) 00:46, 7 January 2011 (UTC)


 * Yep, it doesn't work at the moment; and I'm not sure that it's the best way of rolling out taxoboxes. This is a discussion that needs to be had at some point, but it seems that the best way to roll out taxoboxes across a group is to start at the top and work down; otherwise inconsistencies creep in that are much more difficult to resolve.  I've had good success with this approach with a number of the minor phyla, that either use auto-boxes or now have the back-end templates in place ready to go.  Martin  (Smith609 – Talk)  14:49, 7 January 2011 (UTC)


 * Hmmm. Doesn't really suit my needs, but I'm glad it's useful to you. I might play around with it sometime, but bottom-up seems to be much more reliable for what I'm doing, which is adding one child taxon at a time to the already-existing parents. The top-down approach certainly would require more specific referential material to create new templates. Bob the Wikipedian (talk • contribs) 21:46, 7 January 2011 (UTC)
 * If you're adding one child at a time to parents that exist already, then does the bot (which often runs quite slowly on the Toolserver) really save you much time over the manual process? Does the importing of taxonomy from existing taxoboxes help much, or not?  Martin  (Smith609 – Talk)  22:31, 8 January 2011 (UTC)
 * Actually, that's the biggest advantage of it. Without that tool, I have to navigate to the article and leave the original taxobox there while automating it. But don't change it back just for my cause. It stopped working around a month ago, so I've been making do. Bob the Wikipedian (talk • contribs) 19:34, 9 January 2011 (UTC)

Actinobacteridae
What is the reasoning behind this automatic taxobox? The bacteria articles on wikipedia include mixed and incomplete taxonomies from various sources that often disagree with the current nomenclature, in addition to the fact that the taxoboxes don't always match the wikipedia article. This article is a case in example, and I disagree with the structure of this automatic taxobox, and I don't have days to read up on the current taxonomy of Actinobacteridae to correct it, but suborders are more commonly used, I wouldn't use the phylum, and the orders are incomplete, as are the suborders.

Doesn't it say somewhere than "no information is better than the wrong informaation?" I think cementing the wrong information and making it harder to quickly edit is worse than the wrong information also.

--Kleopatra (talk) 21:13, 28 November 2010 (UTC)


 * That taxonomy was formed using already-existing taxoboxes on Wikipedia; any error in that taxobox would lie in prior editors' edits, not ours. If you have a correction to make, please correct it. Bob the Wikipedian (talk • contribs) 23:36, 28 November 2010 (UTC)
 * It would take me a month of research in high level eubacterial phylogenetics to correct this. --Kleopatra (talk) 00:10, 29 November 2010 (UTC)


 * By the way...how's the protist classification coming along? I've deliberately left them alone so you can work out the bugs. Bob the Wikipedian (talk • contribs) 23:37, 28 November 2010 (UTC)
 * I will look, correct classes and orders in some, add references, then post a list. --Kleopatra (talk) 00:10, 29 November 2010 (UTC)

Multiple authorities listed
The bot shouldn't add a new authority if:
 * The authority is already set {example)
 * binomial_authority is present.

Martin  (Smith609 – Talk)  04:56, 12 January 2011 (UTC)
 * Here's another example: . ErikHaugen (talk | contribs) 05:11, 12 January 2011 (UTC)


 * I question whether Taxobot should be finding authorities at all. Finding a realible source for authorities can only really be done manually, and there will often be bad alternatives available somewhere or other. It should absolutely not be overruling authorities already provided by an editor, and replacing it with a whinge category. Authorities should always be sourced (ideally in the text, where it should also be discussed), and Taxobot is not doing this. The whole task needs to be looked at again. What is this for, for instance? Ugh. --Stemonitis (talk) 08:21, 12 January 2011 (UTC)


 * Apparently, Taxobot is also incapable of distinguishing genera and species. This must stop. --Stemonitis (talk) 08:24, 12 January 2011 (UTC)


 * Another (serious) problem. Between User:Taxobot and User:Smith609, the article on Mucor has been compromised. The authority used to be listed, correctly, as "Fresen." After those two edits, the reader is given no information, and editors are given the choice of two options, both wrong. Fungi are covered by the ICBN, which has different rules from the ICZN. Dates are not included, and author names are reduced to standardised abbreviations where appropriate. Whatever source Taxobot is using is clearly not reliable. I have therefore blocked Taxobot, since its edits cannot be trusted. --Stemonitis (talk) 08:41, 12 January 2011 (UTC)


 * OK, I've reviewed Taxobot's last 50 edits. Apart from one or two articles, all the ones where Taxobot is not the last editor are because Taxobot has been reverted. All in all, that makes it about 25 obviously bad edits, and around 25 where it's not obviously wrong (but the information could still be inaccurate). I think we can all agree that a 50% error rate is excessive. Someone will have to go through the remainder of Taxobot's recent run and clear up all the dross. It may also be worthwhile checking the 50% that weren't obviously bad, in case further errors have slipped through. --Stemonitis (talk) 09:51, 12 January 2011 (UTC)
 * Which ones have you checked? I will be glad to check. --Kleopatra (talk) 15:54, 12 January 2011 (UTC)


 * I've checked all of them now, I think, but only for obvious errors like repeated information. I have my concerns about the data source used, so all the newly authorities should ultimately be checked manually by a knowledgable editor, but that's a very large amount of work. I don't think there's any straightforward work left. --Stemonitis (talk) 16:09, 12 January 2011 (UTC)
 * I'm not good with animals, except for some of the lower phyla, but I'll check what I can. Can we post a list and cross them out or something? I am concerned about a bot adding authorities, but with the reference included it might be okay. --Kleopatra (talk) 04:22, 13 January 2011 (UTC)


 * Really sorry about the inconvenience; I'd disabled the test for existing parameters during testing, and hadn't realised that I'd not enabled it again. As before, the bot will now leave well alone any taxoboxes that already contain authority-related parameters.  Given that this has been fixed, the "bad edits" listed above will not be repeated.  Where there are multiple authorities (e.g. when the bot cannot distinguish between genus and species), the bot doesn't display any to the reader until a human has selected the right one.
 * Before I start it up again, I'll make the bot display the authority source as a reference in the authority parameter.  Martin  (Smith609 – Talk)  23:03, 12 January 2011 (UTC)


 * Thanks for that, Martin. I am, however, still concerned about the data source: it seems to contain a significant proportion of bad data. I am assuming from the hidden links that Taxobot has been including that the source is the Global Names Index, which I would consider a very unreliable source for authority information. ITIS would be more reliable for authorities (it contains a few mistakes, and is woefully incomplete, but is generally accurate where it has authority information), and a well-curated source like WoRMS would be great where available. GNI is only in beta, and appears to contain names scraped without human input from anywhere and everywhere. (If I see a GNI reference in an article, I do my best to replace it with something reliable.) I would consider it the least appropriate source for this task. Would it be possible to have Taxobot look elsewhere? --Stemonitis (talk) 07:03, 13 January 2011 (UTC)
 * As I understand it, the GNI already incorporates information from ITIS and WoRMS (see http://gni.globalnames.org/data_sources). As you've seen, the bot only adds an authority where it is unambiguous: that is, where any record in WoRMS agrees with other sources.  Otherwise it lists the possibilities in a hidden comment.  I would understand this to be the best way to ensure that only reliable information is presented to WP's readers: if there is another data source that disagrees with WoRMS, then no data will be included without human confirmation, making it harder for errors to creep into WP.  Do you think that there is any way of improving on this?  Better still, is there a low-labour way of estimating the accuracy of the data that makes it into WP? Martin  (Smith609 – Talk)  18:19, 13 January 2011 (UTC)


 * The problem is not that GNI doesn't include things, it's that it includes too much, including typos, outdated taxa, outright errors, and so on. It is not limited to high-quality data. Until an editor has manually linked an article to an external database (e.g. with TaxonIds), I don't think you can assume that a taxon with a similar name in GNI is the same one. Taxobot hasn't yet encountered the fact that taxa under different codes can have the same name, for instance (want to guess the authority for Oenanthe or Iris?), and there are all sorts of other quirks that no-one will think of until they crop up in practice. The only way to be sure is to limit the data we include to that which has been seen by a human and accepted as good. That could be an editor here, a contributor to WoRMS (and note that not all WoRMS pages have been curated yet; I should have mentioned that before), or someone on another site, but I think it's a step that cannot be excluded. The answer to your direct question then is no, there is no easy way to do this. That's why I doubt whether a bot can perform the task at all; a lot of human editors find it very difficult, and all make mistakes (myself included). To be honest, I don't see any advantage in having a bot flag an article that it can't provide an authority for by putting it in a cleanup category. An editor still has to find some external grounds for choosing an authority to add, so the bot hasn't helped at all. There might be an advantage in flagging articles without authorities, but that's quite a different task, and must encompass thousands of articles. --Stemonitis (talk) 18:42, 13 January 2011 (UTC)
 * It would be interesting to have some data here.
 * What is the error rate in GNI - and given the bot's conservative strategy, how many errors would the bot introduce? (I'm not aware of a single error having been reported yet.)
 * What is the error rate in Wikipedia? Given that few of Wikipedia's editors are experts in classification, I would be very surprised if there there were fewer errors in Wikipedia at large than in GNI.
 * The bot has one key advantage over human errors - it consistently labels the added authorities with the data source. (I added this function after your previous comment.) Most of the manually-added authority information in WP is unsourced, and thus unverifiable - I for one am guilty of adding authority information obtained via Google from external sites without conducting a thorough review of their correctness.  Given that manually-added authorities are likely to be more error prone and unreferenced, surely it is better for a bot to record where it sourced the information, making it easier for readers and editors alike to assess the validity of the authority information that appears in a taxobox?  Martin  (Smith609 – Talk)  15:51, 27 January 2011 (UTC)


 * I wouldn't like to set a precedent in which GNI is considered a reliable source for taxon authorities. It includes botanical style citations for animals (because that used to be common practice in zoology, too); all the examples I've checked recently include mutiple variants, some of which are trivial (comma before date or not, authors initials included or omitted), but some are not; and there are other problems, too. Do you intend to program an engine which can reliably detect that "Linneaus" [sic], "Linnaeus", "Linné", "Linne", "Linn." and "L." refer to the same person, but "Linn.f." does not, or that "Milne Edwards" and "H. Milne-Edwards" may or may not be the same person (and countless similar examples)? Judging from a small number of crustaceans I've just searched for, unless you can combine slightly different orthographies of authorities, there will be almost no data left to include. All these problems derive from the architecture of the Global Names Index, which is (deliberately) to include any form of a name seen in a wide range of literature, and that is exactly the problem here. The source to be used for a task like this is one which includes a single, reliable, confirmed authority, not a collection of anything that has been printed at least once.
 * In principle, I agree, the practice of providing a good citation for every authority on Wikipedia is a welcome one. I would recommend it to any editor, and it is admirable that Taxobot should implement it. My main problem is with the data source. (I wonder if there would be support for a bot task which adds references to existing manually-added authority data.)
 * I don't think it's entirely fair to compare the error rates for bots and humans directly. Humans are much more accountable and edit at a much lower rate. If a human starts to introduce errors, they can usually be engaged in discussion before anything too drastic happens. With a bot, it can have got through thousands of edits before anyone spots a problem. Our standards for bots have to be much higher than human editors. Particularly with taxonomic stub articles, which may not be seen by a human for weeks or months after the bot's edit, the error rate has to be reduced to the absolute minimum. It may be that no-one has reported any errors, but maybe no-one with the necessary expertise has looked. --Stemonitis (talk) 16:49, 27 January 2011 (UTC)
 * Well, how about we let the bot do the small number that don't include different orthographies? This accounts for many of the minor taxa that I routinely create pages for, so it should allow the bot to have some utility, even if it does limit its scope.  Martin  (Smith609 – Talk)  00:15, 13 February 2011 (UTC)

Taxobot 7
Hey would you consider getting a bot flag for Taxobot 7? I patrol Special:NewPages for templates, and things like this can get a little annoying to scroll through. If Taxobot 7 had a bot flag I would be able to hide these from view, but I can't because the bot doesn't have a flag. Sorry to bug you, --- c y m r u . l a s s  (talk me, stalk me) 16:54, 12 January 2011 (UTC)
 * Yes, please flag this bot. Martin  (Smith609 – Talk)  23:15, 12 January 2011 (UTC)
 * Well, I'm not a bureaucrat so I can't flag bots, nor do I actually know where to ask for a bot to be flagged, but I would back you in a request for a bot flag. --- c y m r u . l a s s  (talk me, stalk me) 23:49, 12 January 2011 (UTC)
 * Any update on the bot flag for Taxobot 7? --- c y m r u . l a s s  (talk me, stalk me) 07:44, 28 February 2011 (UTC)
 * Wikipedia_talk:Bots/Requests_for_approval Martin  (Smith609 – Talk)  18:27, 28 February 2011 (UTC)
 * All right, thanks for letting me know!! :) --- c y m r u . l a s s  (talk me, stalk me) 02:28, 1 March 2011 (UTC)
 * Sorry to bug you about this, but could you try to get an autopatrolled flag for Taxobot7? That will likely happen sooner, and it'll clear up the problem of it clogging Special:NewPages for templates. — Preceding signed comment added by Cymru.lass (talk • contribs) 08:29, 13 March 2011 (UTC)


 * http://en.wikipedia.org/w/index.php?title=Wikipedia%3ABots%2FRequests_for_approval%2FApproved&action=historysubmit&diff=418985275&oldid=418283818

Taxobot 7 creating nonsense
This is nonsense. Why is the bot creating this? Moreover, why is it recreating previously (recently) deleted material? — This, that, and the other (talk) 11:04, 21 February 2011 (UTC)
 * I have a question-- what triggers the bot creating this? As far as I'm aware, Taxobot only does what it's explicitly told to do, though the means by which it receives its instructions aren't quite clear to me. The taxon template Taxonomy/Dinosauroid-Spherulitic Basic Shell Type had been deprecated over a month before the bot recreated the children template, so this has me puzzled as well. Bob the WikipediaN (talk • contribs) 23:02, 21 February 2011 (UTC)
 * The bot creates a Template:Child taxa/XYZ subpage for each Template:Taxonomy/XYZ page. I'm not aware of this causing any problems, or of any advantage of deleting templates in Template:Child taxa/ space.  Martin  (Smith609 – Talk)  17:57, 22 February 2011 (UTC)
 * The problem is deletionists vs. a bot. It would be appropriate for the bot to recognize when a taxon is not used (inclusion in the Category:Deprecated taxon templates is a good indicator of this) and delete the corresponding child template (not to be recreated again). Otherwise, the bot will be edit warring forever with the deletionists. Bob the WikipediaN (talk • contribs) 02:17, 23 February 2011 (UTC)
 * Unless the deletionists don't delete the templates? (The simplest solution is usually the best.) Martin  (Smith609 – Talk)  15:58, 23 February 2011 (UTC)
 * .... Well, I'm going to go ahead and delete the deprecated taxon template itself, as it's actually just a wording deprecation rather than a different name (I'll let you decide how to handle the child template). It should be simple enough for people to find the correct one in the future. Bob the WikipediaN (talk • contribs) 16:51, 23 February 2011 (UTC)

Nomination for deletion of Template:Child taxa/Life
Template:Child taxa/Life has been nominated for deletion. You are invited to comment on the discussion at the template's entry on the Templates for discussion page. — This, that, and the other (talk) 05:43, 4 March 2011 (UTC)

Nomination for deletion of Template:Taxonomy/Albuginales
Template:Taxonomy/Albuginales has been nominated for deletion. You are invited to comment on the discussion at the template's entry on the Templates for discussion page. Sfan00 IMG (talk) 14:00, 2 May 2011 (UTC)

Edit summaries for Taxobot #
Could you please take a look at Bot owners' noticeboard. I'm pretty sure the edit summaries from the taxobots could be vastly improved with just a few minutes'. Headbomb {talk / contribs / physics / books} 20:42, 21 August 2011 (UTC)

Taxobot 7 odd edit
diff - only one and there are no other edits like it, but thought I should bring it up if it's indicative of a deeper problem. Cheers, Rkitko (talk) 15:13, 10 October 2011 (UTC)
 * Thank you. I've blocked the bot until I can address this: something similar's happening with User:Citation bot.  Apparently it's to do with the upgrade to MediaWiki1.18... Martin  (Smith609 – Talk)  15:31, 10 October 2011 (UTC)

Speedy deletion nomination of Template:Child taxa/‚Ð¾Ñ‚Ð°ÐºÑÐ¾Ð½Ð¾Ð¼Ð¸Ñ˜Ð°/Ð’Ð¸ÑÑ‚Ð¸Ð½ÑÐºÐ¸ Ð²Ð¸Ð»Ð¸Ñ‡Ð½Ð¸ â€˜Ñ€Ð±ÐµÑ‚Ð½Ð¸Ñ†Ð¸


A tag has been placed on Template:Child taxa/‚Ð¾Ñ‚Ð°ÐºÑÐ¾Ð½Ð¾Ð¼Ð¸Ñ˜Ð°/Ð’Ð¸ÑÑ‚Ð¸Ð½ÑÐºÐ¸ Ð²Ð¸Ð»Ð¸Ñ‡Ð½Ð¸ â€˜Ñ€Ð±ÐµÑ‚Ð½Ð¸Ñ†Ð¸, requesting that it be speedily deleted from Wikipedia. This has been done under section G1 of the criteria for speedy deletion, because the page appears to have no meaningful content or history, and the text is unsalvageably incoherent. If the page you created was a test, please use the sandbox for any other experiments you would like to do. Feel free to leave a message on my talk page if you have any questions about this.

If you think that this notice was placed here in error, contest the deletion by clicking on the button labelled "Click here to contest this speedy deletion". Doing so will take you to the talk page where you will find a pre-formatted place for you to explain why you believe the page should not be deleted. You can also visit the the page's talk page directly to give your reasons, but be aware that once tagged for speedy deletion, if the page meets the criterion, it may be deleted without delay. Please do not remove the speedy deletion tag yourself, but don't hesitate to add information to the page that would render it more in conformance with Wikipedia's policies and guidelines. — Ryulong (竜龙 ) 19:06, 10 October 2011 (UTC)
 * Page was recreated. I've redeleted it and shut off the bot so Martin can fix this issue. Bob the WikipediaN (talk • contribs) 18:09, 12 January 2012 (UTC)
 * Looks to me that a user is requesting that the bot creates that page. Perhaps a better solution would be to salt that page to prevent its re-creation, so that the bot can carry on with its routine maintenance?  Martin  (Smith609 – Talk)  11:02, 13 January 2012 (UTC)
 * Good idea; I'll see about doing that. Bob the WikipediaN (talk • contribs) 18:08, 13 January 2012 (UTC)

Proposal of importing templates
Now, Automatic taxobox is used in 7 wikipedias of different languages. I think, in the same way that there are bots that copy and paste interwikis a good idea would be a bot that copy and paste templates of this project in other languages. This tool could create a synergy that would help to widespread the "Automatic taxobox" to important wikipedias like Spanish or Portuguese, that use similar formats to template taxobox with the names in Latin. At the same time it would reinforce this project here. This template, for example, doesn't exist here and can be imported just copying and pasting. 83.49.231.120 (talk) 12:14, 24 February 2012 (UTC)
 * This would be great, but it might require some care to avoid introducing inconsistencies. Also, sometimes bizarre features are added to the automatic taxobox system, and it would be bad to copy templates that take advantage of those features to other projects where the automatic taxobox doesn't support those features. It might make more sense to build tooling so that someone can copy missing ones faster. Or even a way to generate missing taxonomies for a particular project. ErikHaugen (talk &#124; contribs) 20:53, 24 February 2012 (UTC)
 * I don't think it would import inconsistencies if the main templates are previously created here. And they are. An other idea is a bot that import images from templates of the same taxon. If in an another wikipedia they set a file in image and the file exists in Commons this parameter could be imported. I also suppose that it could be possible a bot that push up this parameter. If Exemplus ridiculus has an image then Exemplus and Exemplidae could have the same image and import this if the parameter is empty. 193.152.188.114 (talk) 01:17, 2 March 2012 (UTC)