Talk:Pangerang

this edit

'Use the stable URI to the entire work, so the reader can decide which bit they want. Also don’t point them at a 44 MB PDF if it can be avoided.'

To save everyone wasting a huge amount of time.

• All aboriginal articles draw on Tindale's original work, citing in each case the precise page referred to.
• For convenience a link was provided in each case to the South Australia Museum's tribe by tribe transcription of Tindale's 1974 work
• For some unknown reason, these links no longer work.
• The attempt to fix this by this link is no fix. The link simple tells one where to find his 1974 work if one cares to consult a digital copy. No improvement. To get to Tindale's text it requires two clicks. And it does not leave the reader to 'decide which bit they want.'
• It's no skin off anyone's nose to give the immediate direct link to the pdf provided by the ANU. Only one click, and it is a 'stable url'.
• Trying to improve these articles without any familiarity with Tindale's book is pointless. Many of the ostensible corrections show no evidence of actually looking at the easily accessible text by Tindale to verify the accuracy of what had been reported. For example 1982 was a case of numeric metathesis. Had one checked Tindale, that would have been obvious, but it was retained.Nishidani (talk) 05:23, 3 August 2024 (UTC)

@Nishidani: ANU provides two parts: (1) the text and (2) maps. You’re now linking to the text part only, a 44 MB file, that is behind the URL published by ANU for Tindale (1974).

However, I want to understand why you reverted the use of Wikidata using {{cite Q}} to cite Tindale (1974). What are your reasons for doing this? If you object to linking to the ANU-published URL, it is trivial to change the Wikidata item ONCE to link to the 44 MB file instead of having to do so laboriously on MANY Wikipedia pages. Indeed, the Wikidata item has HAD a link to the 44 MB file since before you reverted the {{cite Q}}, it’s just not marked as the preferred link.

Before I updated it, the citation had a dead URL. I replaced the dead one with a live one. That IS an improvement. I also replaced the {{cite whatever}} with a {{cite Q}} for Wikidata item QID 128257949, and did so on quite a few other Wikipedia pages as well, all referencing the same Wikidata item QID 128257949. That is an even better improvement, because instead of having to edit MANY Wikipedia pages should this new URL die in future as well, just ONE Wikidata item would need to be changed: QID 128257949. You went and reverted a lot, if not all, of this {{cite Q}} work, without explanation. And then changed the URL to the 44 MB file behind the URL published by ANU for Tindale (1974), which will likely annoy both ANU and readers of these Wikipedia pages. ANU because their published URL, surely at least in part designed that way to help keep data usage down, is being bypassed, and so they might just do one thing {{cite Q}} and the singular Wikidata item QID 128257949 would help with: kill the URL you now manually plonked in very many Wikipedia pages. And Wikipedia readers because clicking on the link in the citation launches a 44 MB file download. Not all providers have the good sense to offer a jump-page for large files; when they do we ought to honour them. Honouring ANU’s jump-page to Tindale (1974) is just being a good Netitizen.

Finally, I did not create the problems that existed in that Wikipedia page on 1 August 2024. I found them. And I marked them up so they would be addressed, if not by me then someone else, before then focusing on a bit of copyediting and solving a technical problem. You on the other hand first edited that page in May 2017. So before you get even more flustered about what seems to be your assessment of my familiarity with Tindale's book, may I point out that these problems included two different citations to Tindale (1974), with the second linking to the exact jump-page to which you now seem to object (even though it claimed to be for the Berkeley edition with the ANU edition’s ISBN ... it really was a confused mess). And now you point out that the broken URLs in the first one weren’t to Tindale (1974) at all but rather to "South Australia Museum's tribe by tribe transcription of Tindale's 1974 work" BUT which were still cited as Tindale (1974)!

If the Wikipedia articles sourced "South Australia Museum's tribe by tribe transcription of Tindale's 1974 work", then they ought to say that; the dead URLs are archived. So can you please, IF you know, clarify here exactly WHAT the Wikipedia articles were sourced on? Tindale (1974) or "South Australia Museum's tribe by tribe transcription of Tindale's 1974 work"?

If they were to Tindale (1974), which I must assume since that is what they claim, I suggest you go and undo all your reverts of {{cite Q}}, and I go and update QID 128257949 to make the 44 MB file the preferred URL (until something better comes along or we get tired of having to keep finding new URLs and end up settling for ANU’s jump-page to Tindale). If however they were to "South Australia Museum's tribe by tribe transcription of Tindale's 1974 work", I can create a Wikidata item for that, and we can then replace the current {{cite whatevers}} for Tindale (1974) with a {{cite Q}} for the new Wikidata item.

But let’s start with why you reverted the use of Wikidata item QID 128257949 using {{cite Q}}. Your reasoning may be sound, and if so I would be glad to hear it. Betterkeks (talk) 02:51, 4 August 2024 (UTC)

You changed

• Tindale, Norman Barnett (1974). "Pangerang (VIC)". Aboriginal Tribes of Australia: Their Terrain, Environmental Controls, Distribution, Limits, and Proper Names. Australian National University Press. ISBN 978-0-708-10741-6.

Into

• Norman B. Tindale (1974). "Pangerang (VIC)". Aboriginal Tribes of Australia: Their Terrain, Environmental Controls, Distribution, Limits, and Proper Names. Canberra: ANU Press. ISBN 0-7081-0741-9. LCCN 75516991. OCLC 3052288. OL 4795942M. Wikidata Q128257949.

Where's the improvement? We had a link to the SA Museum's files, now dead (well, probably not. They are extremely careful about not offending local indigenous communities and now place an intervening banner notifying readers of the ethics, which has to be clicked before you access their resources. The prior links leapt past that statement, and probably for this reason, no longer works with its former immediacy.)

• What does your version do to remedy that defect? It supplies us with five clickable refs that do not remedy the lack of a link to Tindale's text, but merely marshall a list of things at wikidata vaguely redolent of Tindale's book, where it is held and just one tells the reader where they may download it. If you double click the last, Tindale's text emerges finally, but it has no reference to Pangerang, as the code leads one to suggest it will.
• You are looking at the problem from the perspective of a wikidata technician. I am looking at the problem in terms of WP:V:readers should be linked at a minimum to Tindale's text, and that is why I replied by supplying one (the other viable link is this or this. My url allows the reader to get to the detail by one click to the source, and then a manual search for the specific tribe topic. Your system requires not two but three operations to ascertain the same data, and it advertises, as here, 'Pangerang' deceptively, for when you click the url, it doesn't give you access to Pangerang or any other tribe. It just gives you all of the formal data on Tindale's book. Useless.
• 99 percent of the sources in our several hundred articles are untroubled by wikidata protocols. You get an url for each datum, generally, click on it, and the source of any datum comes up from google books or the internet, as a page in a book or an article. This because these articles are mostly at just a slightly higher than stub level, and require far more editorial improvements of the content. The links serve to assist anyone willing to press on with improving the text to access the relevant literature, which I read but, for pressure of time, from which I only excerpted a small amount of the available data in each link.
• Your wikidata system looks to me, a content specialist, as not helpful to that end. It may satisfy technicians at wikidata, but it does not incentivate readers to participate in article construction. To the contrary, it looks like a laborious and distractive detour. I say this with all respect to those who engage in important technical issues. This is not a technical issue, but a compositional issue, and content must take priority over formalities.Nishidani (talk) 12:43, 4 August 2024 (UTC)

Technically the ideal solution would be to have access by googling a version of Tindale's book which would give one the precise page for any tribe, and thereby dispense with a generic ref to the complete digital text. Ideally 'Tindale+1974+Pangerang' would take you to that entry. But that is not possible under present copyright apparently.Nishidani (talk) 12:56, 4 August 2024 (UTC)

@Nishidani: The citation renders almost identically and is fit for purpose as a citation (but I’ll elaborate on that further down below), while the differences in links aren’t due to {{cite Q}} nor Wikidata. Instead, the links are different because the old URLs are both (1) wrong for the citation and (2) dead, and the new URL is for the work actually being cited: the ANU edition of Tindale (1974). If the correct citation is for "South Australia Museum's tribe by tribe transcription of Tindale's 1974 work", the citation needs to say THAT and link to a museum URL (and an archive since they are dead). In this case the citation will render differently but you’ll get "chapter" links directly to the relevant parts of "South Australia Museum's tribe by tribe transcription of Tindale's 1974 work". Conversely, if the correct citation IS vanilla Tindale (1974), the citation needs to say that (as it does now) and link to Tindale (1974). Unless you have an online version of Tindale (1974) that supports directly linking to a specific part within, the citation needs to link to what we have – the entire work or one of its two volumes (that is, main text or maps) – which Wikipedia does by hyperlinking the title of the work, in this case "Aboriginal Tribes of Australia ...". Please state clearly here in a reply to this thread WHICH it is (IF you know): Tindale (1974) or "South Australia Museum's tribe by tribe transcription of Tindale's 1974 work". Otherwise we will have to continue to assume that the correct citation is Tindale (1974) since that is what the pages currently say.

But you can’t cite one and link to the other, that is misleading.

If you have an online version of Tindale (1974) that supports directly linking to a specific part within, let’s have it. None of the URLs you have provided thus far do, as far as I know, but please tell me if you know of a way. If you can convince ANU to donate Tindale into Open Library or into Wikisource, we can use that. Otherwise page numbers and other forms of indexing such as chapter and section names will have to suffice; a practice that predates Wikipedia and with which you’re clearly familiar, having claimed that "all Aboriginal articles draw on Tindale's original work, citing in each case the precise page referred to".

Be grateful that there IS an online version of Tindale (1974) at all (by ANU). This is no given and if ANU should decide to take it down, there will be NO version online except what we have archived (although the copyright holder could ask for that to be taken down too), and the Berkeley edition (see QID 128387717) on Google books. The other link you provided looks illegitimate and we probably shouldn’t be linking to it, and would surely get taken down too if ANU decides to take theirs down. ANU might, for example, decide to do this after getting annoyed by us bypassing their jump-page as already discussed! I suggest, again, that we respect their jump-page. Clicking twice is better than not being able to click at all!

Now I’ll turn to elaborating why the {{cite Q}} as rendered above is fit for purpose.

First, it provides the customary information for locating a work in the real world, and it hyperlinks to the online version of Tindale (1974) that we do have and which correspond to the citation itself: the ANU edition of Tindale (1974). Second, besides the difference in hyperlinking as already discussed above, the citation renders identically, except for the following:

1. Author name is shown EXACTLY as shown on the cover page of the cited work itself. If this cosmetic difference is what caused you to revert {{cite Q}}, this may be changed readily in any of two different ways to match the existing citation, one being including "|first=Norman Barnett |last=Tindal " in {{cite Q}}.
2. The "chapter title" isn’t hyperlinked because ANU's online Tindale (1974) doesn’t support links to a specific part, whether you use {{cite Q}} or not. But since "all Aboriginal articles draw on Tindale's original work, citing in each case the precise page referred to" and Tindale (1974) has the customary page numbers, using page numbers allows readers to locate the bit in the work being referred. I am in favour of dropping the "chapter titles" (e.g., "Pangerang (VIC)") altogether because there are no such chapters in Tindale (1974), especially if linking to a specific part in the work is not possible and traditional page numbers suffice (in this case page 207), but retained them from the existing citation since they name the "tribe" and state in part II: let me know what you prefer.
3. The book title is hyperlinked because we have a URL to the entire work provided by ANU. Linking to the main text only, bypassing the ANU jump-page, is possible (although a bad idea, but let’s not go over covered ground again), for example by decomposing QID 128257949 into volume 1 for the main part and volume 2 for the map part and then citing the new Wikidata item for volume 1 to get to the PDF of the main text bypassing ANU's jump-page.
4. The place of publication is included, as it should be.
5. The current name of the publisher is ANU Press, not Australian National University Press. If this difference is what caused you to revert {{cite Q}}, this may be changed readily by manually including “|publisher=Australian National University Press” in each {{cite Q}}. But I suggest the current name be used, for self-evident reasons.
6. The short ISBN is shown rather than the long one. If this cosmetic difference is what caused you to revert {{cite Q}}, this may be changed readily in the Wikidata item. Either is as valid as the other, but short is generally preferred because it is, well, shorter.
7. The LCCN, OCLC, OL and the Wikidata QID are shown for additional assistance in locating the cited work, similar to the way the ISBN is included, because in general as much help as possible in locating the work is appreciated. Anyone not familiar with these IDs may ignore them, or become familiar with them. If their presence caused you to revert {{cite Q}}, they could be suppressed in the new decomposed Wikidata item for the main text part for example by simply not including them there. But why not become familiar with them.

It is possible to make the {{cite Q}} render and link EXACTLY like the current citation IF there were an online resource that supported it and you REALLY wanted to.

I’ll now turn to your question as to where the improvement is.

First, the data for the {{cite Q}} comes from a singular definition in Wikidata: item QID 128257949. Defined once, used many times in many places, and if it needs to be changed it only needs to be changed once in a single place. Benefits include consistency, no usage being forgotten when a change is made, reduction of effort for new instances of use and making changes over time, and therefore better resilience to change. These benefits are immediately obvious to any first-year information, data and computer science student.

Second, more information is provided for locating the work, namely place of location, LCCN, OCLC, OL and the Wikidata QID. Should additional information be added to the Wikidata item in future, it could be rendered by {{cite Q}} as in the case of DOI without having to update many Wikipedia pages manually, but at least be accessible through the Wikidata item as in the case of Trove work ID and AustLit ID to name two examples (see QID 128386644).

Your apparent frustration at not being able to click on a hyperlink to get to the exact spot in Tindale you want isn’t due to a limitation of {{cite Q}} and Wikidata. You can’t do that right now because none of the online instances of Tindale support it. Perhaps if the museum jump-page had been respected, it would not have been taken down. If you ever DO find one, {{cite Q}} will allow you to use it.

And now I’ll turn to when we would use {{cite Q}} and Wikidata for citations.

I completely agree that entering citations into Wikidata currently is clumsy, but eventually tools like EndNote are bound to support it (if they don’t already, I haven’t checked recently). I certainly don’t add all citations into Wikidata. Currently I only do for often-used citations, to help make these more consistent and more resilient to change, and especially the often-used ones that may get used even more in future elsewhere, make them easier TO use more in future. It also helps to untangle the sometimes confusing mess of editions and reprints when they are modelled properly in Wikidata: see The Passing of the Aborigines (Q42194072) as an example … as cringeworthy as Bates is, if it is cited, it ought to be (and CONTINUE to be) cited CORRECTLY, and if known and available editions are pre-defined for use, it may just help to get the most appropriate edition used and cited. A globally shared and maintained repository of citations for use not just in Wikipedia but all research ... but I digress.

I thought Tindale was one such work, which is why I added it into Wikidata. I am not asking YOU to start using Wikidata to define other citations. But using the Wikidata item for Tindale (1974) here seems sensible. At least to me. Let me know what, if anything, you want to do. Betterkeks (talk) 11:49, 5 August 2024 (UTC)

My apologies if I don't reply in the same technical terms. Apart from incompetence, I'm a bit slow and distracted today after I fell out of bed while dreaming of competing in the Paris Olympics and bashed my head on the floor, aside from wrenching my transverse abdominal muscles:)

• Your apparent frustration at not being able to click on a hyperlink to get to the exact spot in Tindale you want isn’t due to a limitation of and Wikidata.

Well, I'm not quite 'frustrated'. A direct link existed, innocuously, and then was deadened. Disappointing.

As far as our differences, they are a matter of preference. You prefer wikidata as the point of reference orientation, I prefer a direct link to Tindale's text, for the reasons given. I want people to familiarize themselves with that work, and improve our, I think 650+ articles.

I think earlier you objected to the problems caused potentially to the ANU by a direct url link to the text: many people would start jamming their servers etc. Well, this won't happen for two reasons. The average visits per diem of any one aboriginal page of this type are 3-5 persons. Most, I believe, just reference wiki for a quick glance at an unfamiliar tribal name they come across, inform themselves and move on (b) If anyone wants to explore further, they will not click on all 650+articles each time. They'll download once from any one page, and thus dispense with the need to download again, and again, and again.

Changes to wiki are done by consensus, and we have diametrically opposed opinions, one vs one. I've belled NSH001 and Zero (@Zero0000), who are able technicians. That is not ringing in support since, though we are on amicable collaborative terms, they show Fibonacci intolerance for errors or unreasonable choices I, or anyone else, may make.

Ultimately my preferred choice is (a) a direct link to Tinsdale's book because my bias is WP:V, and I want readers to immediately be in a position to draw up the source and crosscheck it, rather than being directed to a, to me, meaningless page which lists publication data already provided (publisher, date, isbn and much else) where you can find a copy if you like to browse it, by three clicks and (c) direct links are more efficient (c) quick access makes it more probable that some readers might pursue the topic and improve the page. There is also an aesthetic issue (that is of course personal).

Of course whatever decision is taken, it means trawling through several hundred pages mechanically to alter the received version. Chalking up that number of edits might be tempting for someone like myself, as a cheap lurk to reach the 100,000 wikiedit score, but I can't bring myself to do that. It would mean focusing on a minor formal fix for weeks rather than reading sources for the areas I edit. RegardsNishidani (talk) 13:13, 5 August 2024 (UTC)

To be clear, I have offered a solution that links DIRECTLY to the main Tindale text (that is, this link). While I advise against it, it is nevertheless perfectly possible with {{cite Q}}. Betterkeks (talk) 13:33, 5 August 2024 (UTC)

Yes, but that is precisely the direct link I adopted in reverting you earlier over a score or more of articles (which you then objected to). I appreciate the compromise but it strikes me now as supererogatory to introduce uniquely a wikidata format for just one of the thousands of sources used in these articles. The essential data on Tindale reflects the format used to reference all book sources in these articles, and it looks pointless to make some unique exception for Tindale. All of these articles have been (re)written using a single unified template, and the wikidata model destabilizes that. Nishidani (talk) 13:52, 5 August 2024 (UTC)

Actually, girding myself to provide the Tindale ANU direct link, I checked Amangu. There at least the problem appears to have been resolved because the archive url captures the SA Museum link.Nishidani (talk) 13:59, 5 August 2024 (UTC)

I have applied the fix to 35 pages so far and they all work perfectly. In any case I am indebted to you for drawing my attention to the defect in the original links, so that now one may correct them by replacing them all with the archival access to the SA Museum.Thanks Nishidani (talk) 14:46, 5 August 2024 (UTC)

Nishidani has asked me to comment here, mainly on technical matters. Hard to decide where to begin, but I am very clear on one point:

{{Cite Q}} should NEVER be used in conjunction with short-form referencing.

Cite Q is one of those well-intentioned templates that create more problems than they solve. (Even worse are those templates like Cite QPN – quite frequent on Aus articles – but those are a matter for discussion elsewhere.) The basic problem is that it hides (almost) everything from the editor. It might originally be in the right place within an alphabetically ordered biblio listing, but what happens when someone comes along and adds new citations? You could, I suppose refer to the displayed page to find out where and what it's supposed to be, but that's not easy. Especially if there are several of the little blighters. Then there's no way to recognise duplicates. Morevover, it seems, the editor has to manually supply a "| ref = xxxxx" parameter; if not, then there is no way to check for a harv/sfn no-target error. Contrast this with my ETVP script, which sorts a whole biblio listing in order, checks for (and removes if found) duplicates and similarly checks for no-target errors; it can even fix no-target errors in some cases; if it can't fix them, it marks the harv/sfn with a subtle red question mark. None of that is possible with Cite Q --NSH001 (talk) 15:20, 5 August 2024 (UTC)