User talk:Billinghurst

From Wikidata
Jump to: navigation, search

Easy creation of work item from edition... with WEF edition[edit]


I accidentally found today that it is perfectly possible to create the work item directly when editing the edition item with WEF edition...

When you edit the edition item, in the "edition of" field, you can search the work item... and if you don't, by typing a label that still doesn't exist (like "title (author)", the little pen for editing the work changes to propose the Creation of item as.... work... and automatically links it :)

This + the script for moving claims have allowed me to easily create works from editions today (which is much easier than the other way round)...

Hope you'll find it usefull :) --Hsarrazin (talk) 20:38, 8 October 2017 (UTC)

PS : this of course does not solve the problem of finding the original title of a translated work... Goethe was very mean to me this afternoon, but I finally managed to complete all plays we have on wikisource, including one that it not in wp, not in fr, nor in en, nor in de… thanks Google books… I searched with the name of characters ;) --Hsarrazin (talk) 20:38, 8 October 2017 (UTC)

Grant baronet of Monymusk[edit]

Grant baronet of Monymusk (Q41799287); could you please add more info? I have no idea what this item is about… Thanks, MisterSynergy (talk) 18:40, 12 October 2017 (UTC)

@MisterSynergy: I am trying to dig up other similar items to see how they have been classified, whether they are considered as a baronet or a baronetcy, not found a style guide for this sort of thing. I needed a placeholder. It is sitting open at home and will try to get back to it tonight.  — billinghurst sDrewth 21:18, 12 October 2017 (UTC)
Okay thanks, I just wasn’t sure whether it was forgotten. —MisterSynergy (talk) 05:18, 13 October 2017 (UTC)

Authority control[edit]

Hi Billinghurst, please do not copy authority control from English Wikisource without checking the IDs. 2014 I've tried to correct Wikisource but gave up, since my edits were reversed. KasparBot made a list with one widespread error (User:KasparBot/GND Type N). Till now nobody did a clean up. --Kolja21 (talk) 09:09, 27 October 2017 (UTC)

@Kolja21: I don't copy them from there to here, never have. As a community we haven't even had them at Wikisource for over a year, as we solely pull that data from here.  — billinghurst sDrewth 09:31, 27 October 2017 (UTC)
Authority control should identify a person. "Raleigh, Henry" for example is just a name (= placeholder; VIAF "undifferentiated", Wikipedia "disambiguation").
  1. GND 10216973X (name): Raleigh, Henry
  2. GND 189387874 (name): Raleigh, Henry P.
  3. ✓ OK GND 1142577708 (person): Raleigh, Henry (1880-1945), American illustrator
Please only ad personalized IDs. This was the discussion we had on English Wikisource in 2014. (For further information, see Help:P227.) --Kolja21 (talk) 16:24, 27 October 2017 (UTC)
I was just working from , not working directly from DNB. I must have missed the "undiff" bit at the time.  — billinghurst sDrewth 00:03, 28 October 2017 (UTC)

Removing an Id on Guillaume de Machaut (Q200580) ?[edit]

Hi Billinghurst, I notice that back in June you had removed a lot of identifiers and especially the BnF ID (P268): there. It seems to me that the record in BnF authority catalog was correct before you remove it but I am wondering if there is something I missed? Symac (talk) 14:31, 30 October 2017 (UTC)

@Symac: I was working off a whole lot of incorrectly applied VIAF and other extended authority control data and it was more effective to remove the applied data, and re-add with the authority control tool. It is possible that I removed correct data, though I would have thought that the tool would have successfully reapplied that data, I may have missed it. [We had a situation where one piece of bad data within the AC then had bots and people filling extra data. Ugly circle.]


Hi Billinghurst,

Most identifiers on that item seem to amalgamate the two persons. This is why I had left them there. I think if we add them elsewhere (or remove them), it's likely to perpetuate this.
--- Jura 09:20, 8 November 2017 (UTC)

@Jura1: I wasn't planning on adding them, especially as they are already on the other items, and flagging violations. You have already added a veracity statement to deWP which should manage that side, and with the flags added to the item, that seemed sufficient to manage this without further propagating errors. Are you aware of the way that VIAF is picking up the crud mistakes that people have made? They tend to propagate, embed and become hard to fix. Apart from being embarrassing, we simply do not want known bad data to remain from local careless errors.  — billinghurst sDrewth 09:29, 8 November 2017 (UTC)
The problem is that it seems to be incorrect at LOC/GND etc, not just WP. Oddly the bot readded it to Q26222216 despite Topic:U0ad018o6jmj2sdd.
--- Jura 09:34, 8 November 2017 (UTC)
I see your issue about LoC, I have emailed them to address a fix. I have also contacted through general channels, and if that fails, we can try inside channels. I much prefer direct action to sitting and watching. That will be an old error based on the books being authored by "Colonel Churchill", it looks as though it took enWP a while to separate the two gentleman. I would almost recommend pulling all the crap assignations.
As a further general comment, I am really sick of those who bots that blithely and blindly apply data. Bot operators don't take responsibility for the added cruft and mistakes. Their lazy comments or approach that someone else can fix it is both insulting and irresponsible. Yes, it is quicker, but I hardly think that becomes appropriate for building a quality data set. Yes, I am grumpy, and yes I understand that mistakes occur, and I have had my share of those errors. Elsewhere operators are responsible for their bots, here it doesn't seem to be the case. Expediency of data addition seems to outweigh quality of addition.  — billinghurst sDrewth 10:45, 8 November 2017 (UTC)
The Wikidata way of dealing with incorrect data is to deprecate it. I don't think we should wait for everyone else to fix their problems especially if they could monitor Wikidata to improve their data (as does VIAF), but don't try to do that.
As for the bot edit, I would have brought it up on Magnus talk page if you hadn't removed the statements. Now it's just bound to re-occur for sure.
--- Jura 09:29, 10 November 2017 (UTC)
Okie dokie. I think that I have returned, deprecated and hopefully accurately reflected the ids. Had a good response from, and awaiting LoC, prior actioning, and we will look to fix the MARC records.  — billinghurst sDrewth 12:13, 10 November 2017 (UTC)
I re-opened the topic on Magnus' talk page. BTW aren't they all conflation (Q14946528)?
--- Jura 12:38, 10 November 2017 (UTC)
It wasn't MM's that I was complaining about this time, there are others who are cavalier. They could all be conflation, some just looked more the other way, but maybe I misinterpreted their authorities' pages.  — billinghurst sDrewth 14:47, 10 November 2017 (UTC)
I hope so, as without him most WMF sites would probably not be where they are. I think I moved the ones that didn't have the years in the label at VIAF, but they probably all have different criteria what they consider conflation and what it just an incorrect P569/P570 ;).
--- Jura 20:35, 10 November 2017 (UTC)

Edition and translation[edit]


I don't want to flood the Wikidata talk:WikiProject Books but I really want to understand, could you explain to me what is the problem of considering that translation are edition? and where do you draw the line between translation and edition? For instance (a bit dumb example but to know where you stand), if we are precise, Shakespeare's plays were written in Early Modern English and not in English, should we consider Shakespeare's plays in English as edition or translation?

Cdlt, VIGNERON (talk) 12:40, 9 November 2017 (UTC)

Copyright data belongs to a work, rather than an edition, as in new editions do not start new copyright clocks. Translations bring new copyright data, they have extra authors. We also have new provenance. I wouldn't have called Shakespeare's English to English works translations; though anything to another language would be. Any person who did a first language translation of S's works should be granted copyright. If you want a more complex example, the Bible. Under your system, We will have one Bible (work), and all translations are editions. The King James Version (Q623398) is the best example that springs to mind of something that needs to be managed.

If someone does a remake of an English language movie to another language, is that an edition or a new work? The model that we are building is for creative works.  — billinghurst sDrewth 20:52, 9 November 2017 (UTC)

(sorry, I was a bit away)
I hear you but I fail to see what and where is the problem with the current model.
For a movie, if it's really a remake (with different actors, places and everything), it's obviously a new work (and it is how it's done with the current model). If it's just adding subtitle and dubbing (or even some minor changes, like different cuts), it is an edition.
Cdlt, VIGNERON (talk) 19:08, 21 November 2017 (UTC)
@VIGNERON: If you say a remake of movie is a new work, then so is a new translation a work; same conditions, and intellectual property applies.  — billinghurst sDrewth 23:26, 21 November 2017 (UTC)
Can you explain a little more, I don't understand this logic. Cdlt, VIGNERON (talk) 07:52, 22 November 2017 (UTC)