Template talk:IPNI

Accessdate??

I can't see the purpose for the accessdate parameter. I can only see that it would only be useful to document the data from a particular retrieval if that were stored on a page, not the link itself. The parameter should at least be optional. If not specified the sentence should not be displayed. Bob Burkhardt (talk) 20:55, 18 February 2009 (UTC)

Sorry, I've thought about it and now I can see the accessdate parameter would make sense for a citation of a living botanist's accomplishments. But for some situations it is just a nuisance, and it should be optional. And this template should be consolidated with {{Ipni}}. Bob Burkhardt (talk) 01:34, 19 February 2009 (UTC)

Consequently merged--Michael Goodyear (talk) 16:50, 29 January 2015 (UTC)

|name=

I'm seeing this template appear in Category:Pages with citations using unsupported parameters because of |name=:

{{IPNI|query=Christoph.|name=Erling Christophersen|accessdate=March 16, 2011}}

"Search for 'Christoph.'". International Plant Names Index (IPNI). Royal Botanic Gardens, Kew; Harvard University Herbaria & Libraries; Australian National Botanic Gardens. Retrieved March 16, 2011.

—Trappist the monk (talk) 22:05, 21 July 2019 (UTC)

@Trappist the monk: Ok, I stopped the error appearing, but the query doesn't work the way it should, because IPNI recently changed their website radically, and some URLs don't work now. Sigh... Peter coxhead (talk) 22:01, 22 July 2019 (UTC)

Publishers

At https://www.ipni.org/citeus, IPNI shows there is a combination of publishers, not just the Royal Botanic Gardens, Kew. I would like to change the template so that the publisher field has the following as the default value to mirror what is shown on the IPNI website.

[[Royal Botanic Gardens, Kew]], [[Harvard University Herbaria]] {{&}} [[Harvard Library|Libraries]] and [[Australian National Botanic Gardens]]

I know it's long, but it's technically correct. Thoughts? Eewilson (talk) 06:24, 21 December 2021 (UTC)

@Eewilson: I agree that we should cite IPNI as it requests, although I'm never really sure that the publisher field of a citation is appropriate for an internet only database. Peter coxhead (talk) 09:51, 21 December 2021 (UTC)

Does listing all of the publishers aid the reader in locating the source that supports an en.wiki article's text? If yes, then include them all; if no, then do not, if sometimes, then tweak the template to support editor discretion with a default.

—Trappist the monk (talk) 13:17, 21 December 2021 (UTC)

cs1|2 wrapper templates never faithfully adhere to a source's preferred citation style. The recommended citation at IPNI cite us is missing stuff that en.wiki editors consider important (like the database record title) and, were we to faithfully adhere to their format, requires a date format that violates MOS:DATE.

—Trappist the monk (talk) 13:17, 21 December 2021 (UTC)

@Eewilson and Trappist the monk: what I should have made clear above is that my comment on citing IPNI as it requested was only meant to apply to the publisher. We certainly shouldn't take any notice of the style of the recommended citation; we have our own house styles. However, if we give a publisher it should be correct, and it's clear that Kew alone isn't. So the real question for me is whether to give a publisher or not. Peter coxhead (talk) 14:04, 21 December 2021 (UTC)

Yes, I was referring only to the publisher, not the style. IMO, unless the publisher completely duplicates some other value being used in the citation and would be redundant, then we should have the publisher. In this case, it would not. Eewilson (talk) 17:53, 21 December 2021 (UTC)

MOS:CITE for web pages does say that web citations typically include the "publisher, if known". We may just want to go with that. We know it, so use it. Eewilson (talk) 17:56, 21 December 2021 (UTC)

Agreed. Peter coxhead (talk) 07:01, 22 December 2021 (UTC)

date and access-date

Is there a reason why this template mirrors the value assigned to |access-date= in |date=? A brief look at IPNI pages seems to indicate that they are not dated; there is a 'publication' date but that, I presume, it the date of original publication of the plant name not the date of the record in the IPNI database or the date of 'that' version of the database.

Without good reason for the duplication, I think that the practice should stop.

—Trappist the monk (talk) 13:17, 21 December 2021 (UTC)

I agree. (There seems to be a quite common view that a citation should always include a date, but as IPNI is updated more-or-less continuously and doesn't have distinct versions – I've sent in corrections on a Sunday morning that have appeared online by the afternoon – it doesn't make sense.) Peter coxhead (talk) 14:09, 21 December 2021 (UTC)

We could leave the date blank. Fine by me (I corrected what appeared to be a bug). We could also set it to the current year, which would make more sense. In their suggested citation, they have IPNI as author and current year as date. For shortened footnotes, I manually set IPNI citations to have those two values in those fields. Eewilson (talk) 17:47, 21 December 2021 (UTC)

Ready to go live?

Hi, folks. The changes I made months ago to fulfill IPNI's suggested citation detail are, I believe, working properly and will be backwards compatible. Please review the test cases and sandbox, should you choose. I'd like for them to be released if that could be done. See the thread above from December 2021. – Elizabeth (Eewilson) (tag or ping me) (talk) 19:13, 10 October 2022 (UTC)

Clarification, the thread entitled "Publishers". – Elizabeth (Eewilson) (tag or ping me) (talk) 19:14, 10 October 2022 (UTC)

I'm ok with the |publisher= changes except for a couple of things:

• there is no need for {{&}}
• you are creating a comma-separated list where at least one member of that list itself includes a comma. The normal fix for that, as I understand it, is to use semicolons. The value assigned to |publsiher= should be:

  [[Royal Botanic Gardens, Kew]]; [[Harvard University Herbaria]] & [[Harvard Library|Libraries]]; and [[Australian National Botanic Gardens]]

I am not ok with setting |work=www.ipni.org. Don't make citations more cryptic than they need to be; the current form |work=[[International Plant Names Index]] (IPNI) is readable and understandable.

Wherever I have looked at IPNI, I have not seen entry authors; they may exist, but I haven't seen such entries. Defaulting the entry's author to IPNI is, to me, wrong – IPNI is the collection of 'authored' things, not the author of the 'things'. Further, as currently written, |author-link1=International Plant Names Index will link to International Plant Names Index even when |last1=EB Greene (or any other name). I see no need for either of these parameters to be preset. If an entry author is known, adding |last=<surname> |first=<given name> to the in-article template will cause {{cite web}} to render with that entry-author's name (unless excluded, Module:Template wrapper passes all named parameters to the working template).

I see no point in setting |date= for a source that is updated more-or-less continuously; |access-date= is all that is needed.

This is in both the live and sandbox versions of the template:

| _template = {{#ifeq:{{{mode|cs1}}}|cs2|citation|cite web}}

it is not needed and can be replaced with:

|_template=cite web

For those editors who need cs2 style, when they set |mode=cs2, Module:Template wrapper will pass that on to {{cite web}} which will in-turn render the citation in cs2 style.

And lastly, to rephrase what I've written previously: We are Wikipedia; we are not beholden to a source for how we cite that source. So long as we provide a citation that will allow our readers to verify what has been written in our articles, we have done all that needs doing. We are not beholden to the source but we are beholden to our readers.

—Trappist the monk (talk) 21:59, 10 October 2022 (UTC)

Good input. When you get a chance, could you refer me to something like how to write templates and something on template syntax?

I agree with "...we are not beholden to a source for how we cite that source." However, because it's a guideline to use the publisher, if known, we should use it (MOS:CITE).

We don't need the author or date, true. It won't break anything to add it to the template code, and it would make life easier for those of who desire it, for shortened footnotes or otherwise. Yes, the editor could always manually set the |last1= and |date= (or equivalents), or manually set |ref=sfnref. The template uses a field |author= and one called |authority= which mean two different things, and neither is the author of the work. This is why |last1= (or possibly |last= because I actually don't know that there ever would be more than one author) would be the parameter to use for that.

Parameter |work= is being set to www.ipni.org because ideally, the parameter would be |website= since this is a Cite web template. However, the code already sets |work=, and you can't set both, so for backward compatibility, it needs to set |work=. Using the raw website, but not with html, is what I've been guided to do in a web source during a GA, an FA, and an FL if the work name is already used somewhere else in the citation (such as in the work's author) so as not to be redundant. I think readers are clear on what a url looks like, and if the source is a website, I think showing the website (e.g., www.ipni.org) makes more sense than the title of the website. But if we don't set an author for IPNI, then yes, |work= could remain as it is rather than be the website, and if editors want to manually set it, they could.

My ideal way to use the template for a taxon would be as follows, only setting these four minimum parameters:

*{{Cite IPNI | id = 23589-2 | taxon = ''Aster kentuckiensis'' | authority = Britton | access-date = 1 October 2022 }}

The following would be set internally, all allowing an override by the editor: |last1=, |author-link1= (yes, I sure did miss that!), |date=, |language=, |work=, |title=, |url=, |publisher=.

The code I used above would return

[[International Plant Names Index|IPNI]] (n.d.). "Aster kentuckiensis Britton". www.ipni.org. [[Royal Botanic Gardens, Kew]]; [[Harvard University Herbaria]] & [[Harvard Library|Libraries]]; and [[Australian National Botanic Gardens]]. Retrieved 1 October 2022.

OR

[[International Plant Names Index]] (IPNI) (n.d.). "Aster kentuckiensis Britton". www.ipni.org. [[Royal Botanic Gardens, Kew]]; [[Harvard University Herbaria]] & [[Harvard Library|Libraries]]; and [[Australian National Botanic Gardens]]. Retrieved 1 October 2022.

To get a value for the year, the editor would just have to set |date= (or |year=), and yes, they should have to do that anyway if they want it.

If setting |ref=sfnref is preferable for the template to do, then we could do that. We could do that anyway. If that won't successfully carry out from the template to the article (I've had problems and don't know why), then it could just be documented that to use the template with shortened footnotes, set |ref=sfnref or set the |last1= and |date= (or equivalents). Using |ref=sfnref would then by default give this if www.ipni.org is not used for |work=:

"Aster kentuckiensis Britton". [[International Plant Names Index]] (IPNI). [[Royal Botanic Gardens, Kew]]; [[Harvard University Herbaria]] & [[Harvard Library|Libraries]]; and [[Australian National Botanic Gardens]]. Retrieved 1 October 2022.

I suspect this is the default option you would prefer, correct?

To get an author and year in the citation, then the editor would have to set it. That's the way it is now. Would I like it to be easier for shortened footnotes? Sure. But I can live with it.

So at a minimum, change |publisher=, set |ref=sfnref, and change this line to |_template=cite web. Correct? I think setting |language=en is benign, and if that's not already in the template, it could be added. That leaves open the question of last1, date, author-link1, work.

Although IPNI is updated quite frequently, only the year is necessary if an editor sets |date= because no month and day is known, nor would we desire it. We don't want to default to the current year, and you are right, it's in the sandbox! I remember taking that out of the live code. Thought I took it out of the sandbox, too. Either I didn't or it got copied back over.

– Elizabeth (Eewilson) (tag or ping me) (talk) 00:01, 11 October 2022 (UTC)

Wikipedia:Templates?

I never said that |publisher= shouldn't be used, only that the preset value needed tweaks.

Sure adding IPNI as an 'author' won't break anything but it is semantically incorrect.

{{IPNI}} is used in about 390 articles. None of those use |lastn= to name an IPNI database-record author which would be the only reason to to support |lastn= (or any of its aliases except |author= which {{IPNI}} usurps for its own purposes).

So far as I can see, records displayed at IPNI do not show a date that reflects when the record was created/updated. |date= is the publication date of the source. If IPNI don't publish a date on the records that they display, we can't know when that record was published. The template should not render a date that can't be known. The database may be updated more-or-less continuously but that does not mean that every record has been updated. Linnaeus' record likely has gone years since it was last updated; rare, obscure, stable taxa are rarely updated. Don't state a date unless you know it.

Since the IPNI database records are neither 'authored' nor 'dated', there is no reason for {{IPNI}} to support |lastn= and |date=. I'll discuss |ref= later.

I do not understand what you wrote about |work= is being set to www.ipni.org. And, perhaps that means that you don't understand the relationship between |work= and |website=. |work= and |website= are aliases of each other and also aliases of |journal=, |magazine=, |newspaper=, |periodical=. All of those parameters have the same purpose: to hold the name of the source's enclosing work (where the source is that thing named in |title=). If you are citing an article in The New York Times, |newspaper=The New York Times. So, since we are citing a database record in the International Plant Names Index: |website=International Plant Names Index (or |work=International Plant Names Index).

We are having this discussion because {{IPNI/sandbox}} reflects your ideal way to use the template for a taxon. I disagree with your ideal. Except for copying the value from |access-date= into |date=, I think that the live {{IPNI}} – with the changes I suggested in my first post to this discussion – gives a correct rendering. So, yes, change |publisher= and |_template=cite web.

I do not think that you should preset |ref=. Apparently, that parameter has not been used in article space (see this search). It might be better to suggest a standardized form of |ref={{sfnref|IPNI: <unique name>}} in the template documentation. The {{sfn}} and {{harv}} family of short-form templates look for matching long-form templates. When long-form templates are named {{cite <something>}}, the short-form template looks for author/date parameters in the long-form template. Failing that, short-form templates look for |ref={{sfnref|<something>}}. When neither are found, the short-form template emits a red error message. If you are going to support short-form to long-form linking, the simplest way to avoid the short-form error messages would be to rename {{IPNI}} in article space to {{cite IPNI}} and then make {{cite IPNI}} the canonical template name.

—Trappist the monk (talk) 14:57, 11 October 2022 (UTC)

Thanks for the link. Let me make some changes in the sandbox, then I'll ping you back. – Elizabeth (Eewilson) (tag or ping me) (talk) 22:10, 11 October 2022 (UTC)

I believe I removed anything extraneous. The changes are as follows:

• |_template=cite web
• date is no longer set to anything and has been removed
• date removed from |_exclude=
• |publisher=[[Royal Botanic Gardens, Kew]]; [[Harvard University Herbaria]] & [[Harvard Library|Libraries]]; and [[Australian National Botanic Gardens]]
• |language=en

Could you please review the differences, the code, and the test cases? Thank you. – Elizabeth (Eewilson) (tag or ping me) (talk) 22:32, 11 October 2022 (UTC)

Minor quibbles:

• it is not necessary to _exclude |access-date= and |access-date=; these parameters are native to {{cite web}} and Module:Template wrapper will pass them on to {{cite web}}. Consequently:
  • remove accessdate, access-date, from |_exclude= (4×)
  • remove | access-date = {{{accessdate|{{{access-date|}}}}}} (4×)
• rename | work = [[International Plant Names Index]] (IPNI) to | website = [[International Plant Names Index]] (IPNI) (4×)
• replace | _template = {{#ifeq:{{{mode|cs1}}}|cs2|citation|cite web}} with | _template = cite web (3×)

I tend to dislike the default style of {{error}} so for error messages, perhaps:

{{error-small|No {{mono|1={{pipe}}id=}} given and no id found in Wikidata.}} →

At a later date, error handling might want to be revisited...

—Trappist the monk (talk) 23:04, 11 October 2022 (UTC)

Okay, cool. More to follow. – Elizabeth (Eewilson) (tag or ping me) (talk) 23:10, 11 October 2022 (UTC)

I think I've changed everything now. Please verify.

Setting website (which is logical, imo) will result in the following red error within any articles that have manually set work in order to avoid that same error with the existing template.

{{cite web}}: More than one of |work= and |website= specified (help)

It is likely that I am the only Wikieditor who has done such (in order to format the references to my own liking), is there a category the articles with that or with errors is plopped into? – Elizabeth (Eewilson) (tag or ping me) (talk) 23:48, 11 October 2022 (UTC)

If one is to believe this search, there aren't any {{IPNI}} templates where |work= has been set manually. Can you point to an article where an editor has manually set |work=? The category is Category:CS1 errors: redundant parameter which at the moment has 1 articles; the category is usually kept close empty by gnomes.

—Trappist the monk (talk) 00:39, 12 October 2022 (UTC)

I use Cite IPNI, and your search doesn't seem to be pulling up those. Try this search. – Elizabeth (Eewilson) (tag or ping me) (talk) 00:56, 12 October 2022 (UTC)

That gets me what I need and should cover them. Yes, they are all mine. <eyeroll> – Elizabeth (Eewilson) (tag or ping me) (talk) 00:57, 12 October 2022 (UTC)

I cleaned them up. Thought there were more. Anything else? I don't think we changed anything that needs to be added to the documentation unless it would be an informational message for shortened footnotes. – Elizabeth (Eewilson) (tag or ping me) (talk) 01:10, 12 October 2022 (UTC)

Oh, also, I left out the Oxford comma before the Australian publisher in the publisher parameter. I tend to use it, but I'm not sure it matters within the template. – Elizabeth (Eewilson) (tag or ping me) (talk) 01:16, 12 October 2022 (UTC)

I left out the Oxford comma before the Australian publisher It is not obvious to me that you did that unless you mean that you left it out of the manual templates that you cleaned up. I think that the serial semicolon is necessary for the avoidance of ambiguity. Each semicolon 'inserts', if you will, a brief pause in the reading of the |publisher= text especially since '&' and 'and' are both read as 'and':

Royal Botanic Gardens, Kew<pause> Harvard University Herbaria and Libraries<pause> and Australian National Botanic Gardens

If anything is to be removed from that, perhaps it should be 'and':

[[Royal Botanic Gardens, Kew]]; [[Harvard University Herbaria]] & [[Harvard Library|Libraries]]; [[Australian National Botanic Gardens]]

Royal Botanic Gardens, Kew; Harvard University Herbaria & Libraries; Australian National Botanic Gardens

Once this issue is put to bed, I think that you are good to go.

—Trappist the monk (talk) 14:33, 12 October 2022 (UTC)

Done. I'll wait for you to say GO to move. – Elizabeth (Eewilson) (tag or ping me) (talk) 16:56, 12 October 2022 (UTC)

You need not wait on me. If you are ready to update, then update.

—Trappist the monk (talk) 17:14, 12 October 2022 (UTC)

Thank you very much for all of your help and guidance! – Elizabeth (Eewilson) (tag or ping me) (talk) 17:17, 12 October 2022 (UTC)