User talk:Jarnsax

Thanks for your work on author disambiguation

Hi - just wanted to thank you for your activity in this space. Let me know if you run into any problems. --Daniel Mietchen (talk) 05:13, 28 December 2019 (UTC)

Call for participation in a task-based online experiment

Dear Jarnsax,

I hope you are doing good,

I am Kholoud, a researcher at King's College London, and I work on a project as part of my PhD research, in which I have developed a personalised recommender system that suggests Wikidata items for the editors based on their past edits. I am collaborating on this project with Elena Simperl and Miaojing Shi.

I am inviting you to a task-based study that will ask you to provide your judgments about the relevance of the items suggested by our system based on your previous edits.

Participation is completely voluntary, and your cooperation will enable us to evaluate the accuracy of the recommender system in suggesting relevant items to you. We will analyse the results anonymised, and they will be published to a research venue.

The study will start in late January 2022 or early February 2022, and it should take no more than 30 minutes.

If you agree to participate in this study, please either contact me at kholoud.alghamdi@kcl.ac.uk or use this form https://docs.google.com/forms/d/e/1FAIpQLSees9WzFXR0Vl3mHLkZCaByeFHRrBy51kBca53euq9nt3XWog/viewform?usp=sf_link

I will contact you with the link to start the study.

For more information about the study, please read this post: https://www.wikidata.org/wiki/User:Kholoudsaa

In case you have further questions or require more information, don't hesitate to contact me through my mentioned email.

Thank you for considering taking part in this research.

Regards

Kholoudsaa (talk) 16:27, 17 January 2022 (UTC)

categories and versions

Hi! At commons, the category and title page goes with the "work" and the scan goes with the "version". At source, the transclusion goes with the "work" (unless there is more than one version there) and the index goes with the "version". This way, the book template can point to the scan file (for image page) and to the Index at wikisource and the Infobox can point to the transclusion and the transclusion can get all of the wikilinks. Even if there is not a scan, the "version" can be ready for one.

I don't know either what is right nor how I feel about having the "available at url" on the work (as opposed to having it at the version it points to). I always put them with the version that they are, and "google book id" seemed sufficient for that.

I have adjusted Farming Q114065208 and Q114065289 to work better for inter-wiki linking (interwiki links are what wikidata manages, btw). Also, have you seen the "book"? I have seen the "scan", that is the reason I put that into the description.

And a question. Have you made a category for "Haddon Hall Library" at commons and filled it with its works?--RaboKarbakian (talk) 16:15, 22 September 2022 (UTC)

Also, I was using the links that you merged at wikisource. Will you fix them there? Seriously, it will be good work when you fix them there. Right now, you have made a mess at one of the places this linking service is supposed to be "helping".--RaboKarbakian (talk) 16:43, 22 September 2022 (UTC)

I have templates at wikisource. Can you look at them before "reverting" my edits and such. Seriously, I think you want to war and I am doing other things. See s:en:Template:WD version. I did not "revert" your edits out of respect. Do tell if your "revert" is from lack of respect.... Also, do you understand what I mean by "interwiki linking"? If not, okay, it is complicated and I will attempt to explain.--RaboKarbakian (talk) 18:44, 22 September 2022 (UTC)

RaboKarbakian I was actually typing this response, and edit conflicted. The one specific edit I reverted (publication date on work item) I specifically linked you the documentation of the books data model that shows where "publication date" goes. Since the involved items The Peradventures of Private Pagett (Q114065907), The Peradventures of Private Pagett (Q114065935), and the commons cat commons:Category:The Peradventures of Private Pagett (Drury, 1904) don't show any links to Wikisource (there is not even a scan uploaded), I have no way of knowing where or how this is causing issues for you, and am willing to help sort it out, but the answer isn't to break the data model, it's to actually use and implement it consistently. A "work" doesn't get a "publication date", it has an "inception", and the publication date belongs to the edition. See https://www.wikidata.org/wiki/Wikidata:WikiProject_Books/Book_data_model

To be clear: I am working with a bibliography which has been transcribed. If you would like to see it, just ask. I did make those links yesterday. Are you following my edits and changing things purposefully without communicating with me first? Whatever purpose could such an action serve?

No links yet. That is right. The book has not yet been transcribed. I have not requested the upload of the scan. Is there a model that says that the data is authored after the scan is uploaded? Are you to be involved in its transcription?

No, I'm not even remotely following you, I'm working from a list of thousands of entities to look at. Don't be paranoid. The "data model" is about having the information represented consistently across the database, instead of myriad variations as people hack stuff together to "make something work" on a particular wiki. I probably will upload the scans to Commons at some point, however. Jarnsax (talk) 22:44, 22 September 2022 (UTC)

About the "model"; when copyrights are reviewed information about the "first" publication (there is a "first" publisher) is required. A date of first publication is a "thing" and is required for copyright research. I just did a book where the publisher was thanked (in its first publication). So, perhaps the modellers don't research copyright. And perhaps the model is for idiots who cannot figure out how to document it when the first publication was actually in a magazine, and then a book; of which I have several.

Copyright dates should be determined from the registration, or an actual printed copyright statement that includes the date. You should not assume that the oldest copy "we know about" or "that was cataloged" was actually the first edition. In many cases (because of using Library of Congress printed cards) the publication data in the library catalog doesn't actually match that on the copy they have. The "work" and "edition" should both have copyright status properties..... the "work" is what the author created, and the "edition" is what the publisher created. They are not always the same (posthumous publication, typographic copyrights, separate copyright in illustrations or added material, etc.) Jarnsax (talk) 22:44, 22 September 2022 (UTC)

I hadn't really been worrying about Commons too much yet, other than trying to not create egregious brokenness... I was getting about a dozen or so fixed here before sorting that out. If the WD entities don't match the data model, then trying to sort out the stuff downstream doesn't help much. I am aware of how linking categories on Commons works, though, both the "commonly used way" and how it's actually supposed to work (an entity for the Commons category, linked with "topic's main category" and "category's main topic").

Start at commons. They "interwiki link". Those who start here tend to look like they don't know about the many wiki.--RaboKarbakian (talk) 19:46, 22 September 2022 (UTC)

TBH, I'm also somewhat "meh" about the "full work available at URL" thing (it seems redundant in cases where the Google or Hathi IDs are on the item) but it seems to be pretty widely done that way. When I am adding them (and the Google/Hathi IDs) to the specific "edition", though, it is because I have specifically looked at the book scan, (a lot of Hathi book scans are by Google, and on those scans Hathi gives the link to the exact Google duplicate) and verified that it is, in fact the right one..... library catalogs are very prone to having multiple "printings" cataloged as duplicates, and the confusion it creates is much of my motivation in actually trying to sort them out. In the cases where I have added "full work available at url", those urls directly equate to identifiers that are also on the same entity.

Hathi has the same year and publisher all in the same list. It is what oclc tries to do also. I have had real life trouble with "oclc work" when I requested an illustrated book (to scan the illustration) and got the unillustrated version. Some mechanism needs to acknowledge illustrators.--RaboKarbakian (talk) 19:46, 22 September 2022 (UTC)

Also, I was using the "original source at" to point to where to get the "scan" when the time comes to proof read it. Re-re-re-research is a drag.--RaboKarbakian (talk) 19:56, 22 September 2022 (UTC)

RaboKarbakian I did get rid of "full work at url" links to babel.hathitrust.org.... however, I then added the equivalent persistent urls given on the scan pages (for example https://hdl.handle.net/2027/mdp.39015010438425) and documented the scans there. You are specifically supposed to use persistent URLS, if given by the source (the source is telling you to use them instead by providing them, and they are implicitly warning you that other links to the page (the babel.hathitrust.org links) are not stable). HathiTrust itself (on the catalogue pages) uses the persistent identifiers (Handle IDs) to link to the individual scans. Jarnsax (talk) 17:24, 23 September 2022 (UTC)

In the specific cases where you moved the Commons categories over, I'm fine with that. There are cases where a "deep" category hierarchy over there is needed (multi-volume works with multiple editions) but in these cases where there were only one edition, it wasn't really clear to me which the commons category was actually supposed to represent (a single category actually somewhat conflates the idea of work and edition, but when there is only one who cares, not going to create a cat just to hold a single cat). It probably does make sense from a "data model" perspective to create cats on Commons from the top down (work itself first) though.

Mysteries of Police was wrong at Hathi. I made a very specific wikidata and category trees to show how it should be. I have seen some of my wikidatas at viaf!! Have you had this honor yet? Maybe the Mysteries will be fixed here and there. "Special Edition" was a better description. I did "over upload" for it, but my purpose was as already stated.--RaboKarbakian (talk) 19:46, 22 September 2022 (UTC)

I'm sorry to say, you also made mistakes. The "Profusely Illustrated" edition has a title page date of "MCMII" (1902) not 1904, and the "Special Edition" does not have a date. Looking at commons:Category:Mysteries of Police (Griffiths, 1904) right now, I see the title page of the 1902 edition. Also, you were not consistently using "has edition" and "has part" consistently.... specifically, there were "volumes" described as editions of the edition they are part of. I am cataloging these based on actually looking at the scans themselves, not Hathi's data. Jarnsax (talk) 22:44, 22 September 2022 (UTC)

The "version, edition, or translation" item should be labeled "book", not "scan" (with distribution format (P437) = printed book (Q11396303).... the entity represents the manufactured objects created by the publisher, the scans are "digital representations" if represented by an entity (though doing so is wrong unless the specific scan itself is somehow "notable", a scan is a new "distribution format" of the published book, not a new edition). If you notice, I've made a point out of using distribution format (P437) = digital representation (Q42396623) as a qualifier. Trying to create individual entities for each scan devolves into madness quickly (mass proliferation of entities, and dealing with the same digital object being available multiple places).

Library of Congress has books. Books which are not scanned, books which are scanned. They are different. If Library of Congress says they have the book and that such and such link is the scan of that book, then okay. If books are scanned and then discarded (used as toilet paper or firewood, for instance) then the scan is the scan and the book does not exist any longer. Spend sometime looking for "scans" to see my deep and real experience with this.--RaboKarbakian (talk) 19:46, 22 September 2022 (UTC)

The Library of Congress's cataloging method lumps "editions" that are identical other than publisher and date as "copies" of the catalogued version, even though they are not for our purposes. When the published cards are used by other libraries, the data often reflects the "LOC cataloged copy", not the item the library actually holds. That's why I'm actually looking at the "books" (scans) themselves, and cataloging them in the first instance, not copying the catalogs. Jarnsax (talk) 22:44, 22 September 2022 (UTC)

I don't have any actual info about the "Haddon Hall Library" itself, other than incidental stuff like https://www.abebooks.com/first-edition/Haddon-Hall-Library-Complete-9-Volumes/22645288967/bd and having seen the bindings and cover pages in scans. I think the other books are a bit farther down the list I am working from (i.e. already have WD items) and I intended to mess with collecting them together once I have all 9. Jarnsax (talk) 19:04, 22 September 2022 (UTC)

Haddon Hall Library was a special cooperation between the pubisher (who you removed), Aldine House (or some such) and Haddon Hall. I am going to do the chapter beginning and end images from Shooting, because they were reused often within the series and I think this one book contains most of them. I have been researching these books.

I watched a person proof and transcribe a chapter (a fairy tale, so also a stand alone) from another book recently. That person was able to easily install the transcribed tale into wikidata, because it was here to do that with. Once the transclusion link went to wikidata, three or four other links were made from text to hyperlink! Any model should start with interwiki linking. The Publications model makes it so that wikisource never makes it from here to wikipedia. Wikidata can provide interwiki links or be a place of war for data modellers. One is probably more prone to get funding than the other....--RaboKarbakian (talk) 19:46, 22 September 2022 (UTC)

AFAIK I didn't remove any publisher info (I might have moved it to the right place, or fixed it, tho). There wasn't an entity for the book series until I created it.

My goal here is not to fuck your stuff up, which you seem to think. My goal is to fix mangled or incomplete data here. The whole point of a "data model" is so that the information is entered in a consistent format across works, which lets people use the data in a reasonable way instead of having to mangle together individual use cases to "get it to work" somewhere else. If the objects are entered correctly, stuff just works as documented. Jarnsax (talk) 22:44, 22 September 2022 (UTC)

In the collection of entities that you created for "Mysteries of Police and Crime" there was no consistency across the editions..... you had "volumes" of the work entered as editions of the edition they were part of in places, when they are parts of type "volume", and not in other places. You had entities listed as "parts" of the work itself, which they are not. You also had created entities as "instance of: digital representation"... individual book scans fail WikiData's notability criteria, and don't serve any actual purpose (not how you do it). There is nothing particularly "special" about the needed data structure, if done correctly...

• work - what is created by the author. Has an inception date, not a publication date. Should have a "copyright status" property based on first publication (if published). Gets "has edition" links to editions. Has an author, not a publisher.
  • edition - what is created by the publisher and printer. Has a publication date, and should have a "copyright status" property based on first publication of that edition. Might be issued more than once, in different "distribution formats" (paperback, hardcover, ebook for modern stuff, facsimile, microprint, digital representation). Gets "edition of" link to work, in this case gets "has parts of the class" = "volume", and "has part" links to the volumes.
    • volume - the parts the edition is divided into, inherits most properties from the edition. Gets "part of" link to edition. Since none of the exemplars are notable, we don't go below this. Any data for exemplars or scans goes here.

There should be one work entity, 3 edition entities, and 8 volume entities. Jarnsax (talk) 22:44, 22 September 2022 (UTC)

RaboKarbakian As a further note, and to ping you, if you do have a "structural need" to create entities for particular exemplars (specific copies) they should be linked to the specific volume entity, with exemplar of (P1574), and should describe the physical item held by the library...how to put it.... the claims on the volume entity that correspond to that particular exemplar should be copied to it, and claims about it's digital representations should made on that entity. Jarnsax (talk) 00:30, 23 September 2022 (UTC)

This record https://catalog.hathitrust.org/Record/008559179 links to a UIUC copy of the "Profusely Illustrated" edition for volume 1, but the links for volumes 2 and 3 are to U of Minnesota scans of the "Special Edition". Jarnsax (talk) 00:56, 23 September 2022 (UTC)

I've finished (as far as I can tell) working on the WD stuff for "Mysteries of Police and Crime". I went back through and actually looked at the title page of all the scans, again, to make sure they are all connected to the right volumes. As it stand now, the category pages infoboxes over on Commons are now correctly displaying the data I added, including links between the editions, the volumes that make them up, and the follows/followed by relationship between the volumes themselves. Hopefully that level of "then it just works" is enough to convince you that there actually is a point to following the data model. Given the "stalking" comment, I'm not actually going to mess with them on Commons, though from what the title pages say the dates are wrong. The "names" I used for the edition/version claims (Profusely Illustrated and Special Edition) are what it actually says on the title page, if you want to show something different on Commons you can override or suppress fields in the infobox template. I did not use references on most of the WD claims (because I was just looking at 'the thing itself'), but feel welcome to add referenced claims (and maybe set them preferred) if some source discusses the work and gives more or different info. Jarnsax (talk) 02:14, 23 September 2022 (UTC)

Actually, now I'm done. Added another edition (a 1899 reprint of the first edition) and used it's colophon text to determine the actual first edition date (November 1898). Jarnsax (talk) 17:11, 23 September 2022 (UTC)

@RaboKarbakian: This, by the way (Stories of King Arthur (Q114140906)) is how you are supposed to create interwiki links to categories. Doing it "the other way" creates interwiki linking problems, an example being when "some thing" has wikipedia articles, and both a gallery and a category on Commons. Cross-namespace interwiki linking (through links to a wikidata item) is bad. Done properly, the category on Commons is linked to the "Wikimedia category" WD entity, but the edit link for the WD-generated infobox takes you to the entity for the "thing" itself. Jarnsax (talk) 19:12, 23 September 2022 (UTC)

@Jarnsax: By the way. Wikipedia will never have those categories as they would hold secondary articles and/or fair use images. AND btw, you have, with those useless main categories broken my template. So, perhaps you could 1) fix the template so that it will work with these wasted effort shared categories or 2) request they be deleted. I use that method for categories here when there is stupidity in the cats at commons or when there actually might be a shared category. The templates that use those are not lightly used at source. Thanks for your further attention to this detail you have created.--RaboKarbakian (talk) 21:34, 23 September 2022 (UTC)

@RaboKarbakian: Not sure what categories you're talking about "will never have", but the "wikimedia categories" entities should only be created if those categories exist, it's how you link categories to commons, and if you look at the category pages on commons for Mysteries, the book infoboxes now correctly work, when they did not before, including correctly linking to the other commons categories. That's because I fixed how the work, edition, and volume entities are linked to each other. If you are talking about the entities for works where there is only one edition, I fixed (by adding the wikimedia category entities) how the categories are linked, technically, and moved the information around to the correct entity (work or edition) according to how it's supposed to work, according to how the WD community has agreed to model data about books (which also happens to be, rather obviously, what the infoboxes on Commons rely on, since they now work). The categories on commons that are not showing the publication data in the infoboxes (where there is only one edition) are not doing so because they are linked (now, at least correctly, in a technical sense, but incorrectly) to the work entities instead of the editions. It is correct for them to be linked to the editions, because the scan is actually of a particular edition (which has a publisher). The only reason I have not fixed them is because, when I did, you complained about it. Instead, I have left your stuff over at Commons broken (by not linking the categories to the editions) and as I continue to work down my list (and adding a hell of a lot of data to these entities with Hathi link on my list) I will fix everything except (if I happen to notice it's "yours" over at Commons) I'll leave it mis-linked it to the work, as you insist it should be, go you.

Like I said before, the simple fact that fixing it, on "Mysteries", suddenly made the Commons category template correctly work, and correctly link the editions and volumes to each other, as well as showing much more information I added to the entities where it belongs, is telling. It's rather like that is how it's supposed to work on Commons, and how the templates there are designed to read the data here.

All you've told me about issues at Wikisource is a link to a template, and the statement that it broke stuff "somewhere". That's great, nowhere near enough for me to figure out what your issue is. Something like an example of where it's broken would be helpful, but the issue is the template on Wikisource, not how I moved the data around here... I moved it to the "data model" that is documented here and that makes the templates on Commons work, and the only stuff I "broke" over there is where you insisted I leave the category link here pointed to the wrong place. The simple fact is, your Wikisource template needs to play well with others, including Commons, and you showing me examples and being civil instead of calling me a stalker would probably have helped.

As it is, unless you want to "start over" here, be civil, let me move (or do it yourself) the commons category links to the editions (so that they will actually work right on Commons), and show me some examples of where pages are actually broken on Wikisource so that maybe I can help you fix your template, then you need to get off my talk page. Instead, you can RTFM like you should have to begin with at https://www.wikidata.org/wiki/Wikidata:WikiProject_Books to learn how books are described here, and how to make the templates on Commons work. 22:29, 23 September 2022 (UTC)

Okay, first, I did not get your ping. Second: I have reused the wikimedia category qids. There are so many reasons that it is not good for "interwiki links" link to reasonator and that is how the module works. At the time that the module will stop assuming (as I did and a few others at wikisource) that the "shared wikimedia categories" will always have a wikipedia article associated with it. If no article exists, the category will not be shared. Please, help me to avoid making what should be good interwiki links point instead to reasonator.

This is how the module works: it looks for an article (Main space title) at it's home wiki. Not finding that, it looks for one at the same language wikipedia. Not finding that, it looks for a category at the commons (which is the commons equiv of an article for source)). At that point, it would be silly for it to look for a shared category, no wikipedia article, no shared category! There are so many better things that the module could do (some grammar changes, some finding the named used and not the full name, etc.) that I would prefer not to the writers of the module to write for this case of a person who was following my links one day. I am not accusing of stalking; I am interested to know the non-stalking reason.

If you don't know how to find where the links are, probably a good default is not to change things. I don't make a wikidata unless I have something to link it to. I work on two wikis at one time, usually, for these entities.

I can fix your breakage, as I stated above, I am already doing so. A better task for you might be to find a list of books, or articles (but that will be quite more difficult) and try to find scans for them, find some title pages and toc scans to upload to commons, and then work with the books project how to and make your entities work here, at source and then at commons. You are smart and could very probably tell me where I went wrong. But you are not smart about this yet.

If someone were to tell me that another person used their entities, dropped one title into the source space and by doing so, made 3 or 4 text mentions into hyperlinks -- I should like to know how they did this and not proceed to instruct them on how to break things. But that is me and perhaps you are not like this.

Uncivil is perhaps also defined as following an edit list and making changes without even a mention. So, when I got a little peeved, the first uncivil, the second uncivil and perhaps event the third uncivil had already happened and I did not do them.

adjective: uncivil

discourteous; impolite.

"he'd been short and uncivil with her"

"Short" like nothing is short a few seconds....--RaboKarbakian (talk) 17:49, 24 September 2022 (UTC)

@RaboKarbakian: Just no. you do not hijack or "reuse" Qids. This is just totally wrong, and not how Wikidata works.... it is designed to never reuse them, merged or deleted Qids are never recreated, they are intended to be persistent identifiers. This is so wrong that if you don't fix it, or if I ever see you do it again, we'll be going straight to the Admin Noticeboard. By "recycling" an entity into one with a completely different scope, what you are doing is actively evading a discussion of the item at the deletion noticeboard.

You still, despite my repeated requests, have yet to show me an example of where my edits "broke something" (they fixed things for the Commons templates, which should show you that the lua module does indeed work that way). You've made vague comments about Wikisource, but none of the items I have worked on, that you consider "yours", had any links to wikiwource (either interwiki, or with the index page linking property), so again, I have no idea what your problem is other than you don't like the way Wikidata has decided to model dat about books. Given that you obviously don't understand the way that data about books is supposed to be structured on wikidata, and based on your earlier comments, I'm assuming the actual problem is more of you just linking to the wrong entity, and putting data on the wrong one.

Again, and last warning before this goes to AN, stay off my talk page. I have no interest at this point in 'justifying' why I decided to work on specific items to some random person, or listen to him trying to tell me what other stuff I should work on. What I do with my time is none of your fucking business, and nobody "owns" any wiki pages or WD entities, other than their own user and talk page. Go away, or go to AN. "Hijack" more entities to avoid a deletion discussion, go to AN.

As far as Stories of King Arthur (Q114140906) (the entity you hijacked) somehow not creating the right links, that's because it was, in fact, linked to the wrong entity. I didn't fix it because you changed it back when I did, and I have specifically told you that the reason it was not going to work correctly was because you were insisting on having it link to the wrong item. If you had let me fix the damn thing in the first place, then it would work to display the correct data on Commons (and, I guess, wikisource). Jarnsax (talk) 20:17, 24 September 2022 (UTC)

Template:Pinr Actually, I lied, I will tell you one more thing, which is to look at Sketches by Mark Twain (Q19039232), which, though incomplete, is how you are supposed to create interwiki links to Wikisource. 20:40, 24 September 2022 (UTC)

PMA

Hello, what are you doing here exactly? Horcrux (talk) 07:27, 20 October 2022 (UTC)

@Horcrux Basically attempting to finish what it seems someone has started, with the "copyright holder" stuff, and the "years after author's death" entities and heuristics... that had only been used a couple of hundred times total, to make my life easier researching copyright stuff.

Means you (or at least, I, and hopefully other people, and maybe bots, lol) can use based on heuristic (P887), and this stuff... heuristic for determination of copyright status of a creator (Q108697774) and copyright determination method (Q61005213).. to fill out things like "copyright holder", "copyright status", "public domain date" etc just using the heuristics and references.

Like I said, someone had 'kinda' set it up a small part of it, and someone had copied the "list of countries by copyright length" into those category "years of less" (with the classes conflated so it didin't 'make sense', lol.)

Go look at 50 years after author's death (Q114765077) and 70 years after author's death (Q114765114) to see where I did the "A"s. :) Hopefully, this will be useful, at least in easier data entry, lol. Some of it's still "unknown item" because of stuff I haven't made yet.

I'm 'keeping track' (and a lot of this, I didn't make) in my sandbox. Jarnsax (talk) 07:43, 20 October 2022 (UTC)

Some of those (like 74 pma) are for odd unique lengths... that was a unique term for Soviet WW II vets. Jarnsax (talk) 07:45, 20 October 2022 (UTC)

Eventually (ideally) the same kinda of "list" I'm making, across the "countries with xx pma" can also be connected from the other side (the "copyright law of whichever country" entities) as a "has effect", to point at the terms for various stuff. Jarnsax (talk) 07:52, 20 October 2022 (UTC)

Hopefully that all makes sense. I think the US stuff is largely done, tho apparently not documented. Jarnsax (talk) 07:53, 20 October 2022 (UTC)

I see three main problems in that item:

1. The first one (that doesn't depend on you) is the wrong usage of property contains the administrative territorial entity (P150). I think one should use has part(s) (P527) instead.
2. Then I see a wrong current usage of property has part(s) (P527), that is in contrast with subclass of (P279). You should either say that A is subclass of B or that A is contained in B.
3. Finally, I don't understand why you specified so many "unknown value"-statements for property has part(s) (P527).

--Horcrux (talk) 07:58, 20 October 2022 (UTC)

@Horcrux Yeah, the "admin territorial entity" errors aren't mine, that had already been put in (tho I actually want to make it go away into the "countries with XX years pma" cats in more detail. It hadn't shown as an constraint violation before because "jurisdictions" and actual "territories" had been conflated in the classes. If you notice, the "A"'s on the "countries with XX pma" entities (that I wrote) aren't doing that.

The has parts thing, yeah, I was actually unsure about that (I hadn't really messed with the "countries with" logic yet, they were still conflated). I had only just started looking at them in reasonator to try to debug the "inheritance" stuff... and had notice it was broken differently, lol.

The "unknown value" things are actually for stuff I haven't finished yet (and that they didn't finish). If you look at the list in my sandbox a lot of the "death" things are still missing, I intend to stick them in as I fill out the entities. Just trying to make myself less confused.

I'd intended to poke around and get feedback, once I'm gotten to the bottom, lol, before copying more actual "legal" stuff.

Then, hopefully do whatever needs to be done to the "published date" side of it, eventually. Jarnsax (talk) 08:10, 20 October 2022 (UTC)

Thank you for the reply. --Horcrux (talk) 08:52, 20 October 2022 (UTC)

Not I problem, I probably will need help at some point, lol. Jarnsax (talk) 08:56, 20 October 2022 (UTC)

@Horcrux If you look at countries with 15 years pma (Q114773644), countries with 15 years pma or shorter (Q114772623), and countries with longer than 15 years pma (Q114772416) (which are not "populated" yet) I think that's "correct", other than the "author's death" entities that are still missing. Jarnsax (talk) 23:20, 20 October 2022 (UTC)

Er, och, those should be qualities not effects. Fixing, lol. Jarnsax (talk) 23:21, 20 October 2022 (UTC)

I think how I have the "A"'s described at countries with 50 years pma (Q114763455) should also be "correct" now, if you want to look at that. Jarnsax (talk) 00:19, 21 October 2022 (UTC)

Not that sourcing them deeper than "commons said so" would be an improvement. Jarnsax (talk) 00:21, 21 October 2022 (UTC)

Part of the "plan", once all the "bits" exist, is to also make these statements works protected by copyrights (Q73555012) and copyrights on works have expired (Q71887839) an "effect of" the various nations copyright laws. Jarnsax (talk) 00:30, 21 October 2022 (UTC)

Hopefully that will make copyright research by editors more convenient. Jarnsax (talk) 00:31, 21 October 2022 (UTC)

@Horcrux See my comment at Talk:Q63705939. The class was defined as a group of jurisdiction (Q471855) (which I assume was just a mistake at some point) and as a subclass of geographic region (Q82794), which I think is just incorrect... using jurisdiction (Q5982983) embodies the same concept of "area it applies to" without it actually being a "contiguous area of land" itself. Jarnsax (talk) 05:45, 21 October 2022 (UTC)

" applies to part executive branch of the U.S. government"

Statement

Why not just "has part: executive branch of the U.S. government"? Lectrician1 (talk) 20:50, 25 October 2022 (UTC)

@Lectrician1 I'm working through the laws passed in the session... they weren't originally part of the Customs Service, or the Treasury Department (neither of which existed yet)... they were just "made", then the rest sorted out several moths later. They were actually the first "officers" appointed under the Constitution, other than the Secretary of State and his Chief Clerk. Once I get to that Act I'm intending to sort it out more. There are actually enwiki articles about a few of them, and, lol, NARA's record groups informed me they all have the wrong name (United States Customs District). It was also interesting to find out that a "Naval Officer of Customs" wasn't actually a naval officer, from reading a history while I work on these. :/ Jarnsax (talk) 21:18, 25 October 2022 (UTC)

I think I got the actual "part of" "federal government" "applies to" "executive branch" from the State Department, actually. Jarnsax (talk) 21:20, 25 October 2022 (UTC)

I'm actually intending to create entities for all 50-odd of the actual "Collector of the Port of XXX" positions, one I create the districts.... due to the whole "spoils system" (and that being a tax collector was very well paid) these jobs were held by a lot of very famous people... it was basically a sinecure, and you could be in business or political office for a state at the same time, lol. Corruption city. Jarnsax (talk) 21:24, 25 October 2022 (UTC)

@Lectrician1 That better? Like I said, I'm not to that law yet, I'll fill in the dates and such. It was literally in the same session, I'm not sure yet if it was 1 Stat. 53 (which was about lighthouses, bouys, etc, that they ran) or 1 Stat. 65, that created the Treasury Department. Jarnsax (talk) 21:32, 25 October 2022 (UTC)

@Lectrician1 Also amusing, to find out from the history, that one of the early Collectors in North Carolina "went home to Ireland" for five years, came back, and started working again without his boss noticing, lol. Easy job. Jarnsax (talk) 21:35, 25 October 2022 (UTC)

Idk this is too complex for me to figure out right now, but at first-glance the statment would look weird to most contributors so that's why I asked for maybe something better. Lectrician1 (talk) 21:44, 25 October 2022 (UTC)

@Lectrician1 Yeah, I kinda knew it was weird, just don't actually quite get the history yet either, lol. It's a work in progress. :) Jarnsax (talk) 21:47, 25 October 2022 (UTC)

@Lectrician1 If you don't mind glancing at something else, actually, to see if anything looks odd. One reason I kinda started over "at the other end" is most US laws just odd stubs created from wikipedia articles.. the "help page" is quite meh, lol, so I had to sort of model them from scratch, but... Tariff of 1789 (Q7686029) and the other few laws in front of and behind it. Jarnsax (talk) 21:55, 25 October 2022 (UTC)

Uh yeah that "applies to part" is odd. I don't understand what it's trying to say. I look at "applies to part" property examples and the way you're using it (not for geographic entities) is completely different. Lectrician1 (talk) 23:23, 25 October 2022 (UTC)

@Lectrician1 I think sometimes it is somewhat vague, at least how I use it... I'm trying to imply something I don't quite know how to say, I guess. Trying to deconvolve concepts is hard, lol. Maybe I should just use "of" since it's actually intended for 'vague' stuff... I'm always half expecting someone to come along and change it and make me go "duh". It actually happens, lol. Jarnsax (talk) 23:39, 25 October 2022 (UTC)

Uh absolutely don't ever use "of". It's depreciated. If we don't have properties suitable for the relationship you're trying to make in a clear way, look into proposing one. Lectrician1 (talk) 12:29, 26 October 2022 (UTC)

@Lectrician1 I'm aware that "of" is bad.. sometimes I just don't really know a good way to "say it", which is what I think most people use it for... when I do, it's rather meant as a placeholder that will hopefully get fixed by someone if not me.

I think the most common case where I end up using it is to say something like "has effect" "rule" "of" "oath of office".... don't really know a better way to express that type of relationship. Jarnsax (talk) 17:39, 26 October 2022 (UTC)

I do the same with "object named as" quite a bit, also... I'll use it as a qualifier to some "unknown" value, for an entity I haven't created yet, with the intent to go back and fix it. Jarnsax (talk) 17:42, 26 October 2022 (UTC)