User talk:Laddo/Archive/2016/1

This page is an archive. Please do not modify it. Use the current page, even to continue an old discussion.

Are you removing both to add the new versions? (It seems they recently change pub to in).
--- Jura 16:12, 2 January 2016 (UTC)

@Jura1: I'll put back the "/in/" in a few moments, removing one specific URL with Autolist2 ended up removing both... -- LaddΩ chat ;) 16:23, 2 January 2016 (UTC)

I went through that too ;) . BTW, some have several accounts.
--- Jura 17:49, 2 January 2016 (UTC)

@Jura1: Noted. I'll see from the violation report and fix the last ones individually tomorrow. Thanks -- LaddΩ chat ;) 17:59, 2 January 2016 (UTC)

Dear Laddo, I restored the ChemSpider ID for sarin (Q187695) which you set to "no value". I checked both the English Wikipedia (one of the two references) and ChemSpider itself, and the value was correct. Did you set the ID to "no value" deliberately? If so, why? Egon Willighagen (talk) 07:09, 16 January 2016 (UTC)

@Egon Willighagen: Very good catch, Egon, thanks for reverting. It's rather statement NFPA Special (P877)="none" that I wanted to change to 'novalue'. These days I'm often fouled by this issue that confuses StatementSort and I sometimes end up accidentally editing the wrong statement. It seems I forgot to double-check sarin (Q187695) after-the-fact, so I'm glad you fixed it :) Thanks again -- LaddΩ chat ;) 13:43, 16 January 2016 (UTC)

Thanks for your edits! Could you also edit format in order to allow the value "novalue": these items are correct. Best regards, --Epìdosis 10:18, 8 February 2016 (UTC)

@Epìdosis: Done. -- LaddΩ chat ;) 12:26, 8 February 2016 (UTC)

hi, all this particles are not individual particles, they are types ... We should use instead subclass of (P279). author  TomT0m / talk page 18:29, 7 February 2016 (UTC)

@TomT0m: Hi, I was expecting some feedback on that change :) I really thought about it and believe that this change is correct: "subatomic particle" is a broad categorization, some kind of metaclass. Rather than obscurely stating than an electron is a type of lepton, I believe it is now correctly classified as: "it is a subatomic particle, a type of lepton". Of course, lepton is ultimately derived from that same "subatomic particle" entity. -- LaddΩ chat ;) 22:43, 7 February 2016 (UTC)

Hum, then the name is not well chosen, because I would believe that the name "particle" would refer to a specific particle. Then I would prefer to have a second item "type of subatomic particle (in the standard model)" to use for "instance of". subclass of (P279) was already used for some of the claims you instroduced, and it's a mistake to have both with the same values. author  TomT0m / talk page 11:40, 8 February 2016 (UTC)

@TomT0m: OK, will clean up duplications. Not sure about renaming to "type". Rather create a new item for that and use it? Busy today, will check that tomorrow. -- LaddΩ chat ;) 12:26, 8 February 2016 (UTC)

Please chek Help:Classification for a rationale. author  TomT0m / talk page 12:29, 8 February 2016 (UTC)

 Done Replaced with type of quantum particle (Q22675015), "Type or family" :S -- LaddΩ chat ;) 17:56, 9 February 2016 (UTC)

I don't know what you are trying to force a regeneration of, but you edit has no effect on that. If you want the constraint report to update there are two options:

1. Make sure there is a constraint violation that wasn't there earlier (you have to do that before approximately 12.00 UTC)
2. Add <!-- force update --> to the beginning of the report page on a line of its own. (Just do that before the reports get updated by the bot, it forces an update of the report at the next report generation cycle.)

If you want to force a regeneration of something else, I need to know what regeneration you want to force in order to possibly be able to help. Mbch331 (talk) 17:31, 18 February 2016 (UTC)

@Mbch331: Hi, thanks for the info to regenerate violations reports, I didn't know that one ;)

However my issue was rather the proper categorization of the property talk page itself; my edit did succeed to achieve what I was trying to do: at the bottom of Property talk:P2534, despite that Category:Properties with math-datatype correctly appeared, the page was not listed on that category. Using "?action=purge" on both the category page and the property talk page had no effect. Only that dummy edit that I did managed to have the talk page listed in the category.

If you know a simpler scheme and would be interested to test it, there are a bunch of other cases similar to that one: I noticed that there are many properties newly converted to external-id datatype (this list), but despite that they all display Category:Properties with external-id-datatype at the bottom of their respective Property talk pages, only 9 of them appear in the category. Unless you can suggest otherwise, I'm afraid we'll need to ask User:Hoo man to mass-purge talk pages of properties that changed type recently (as he indicated that he did for affected items). -- LaddΩ chat ;) 18:23, 18 February 2016 (UTC)

Doing a recursive purge of the module used on all these talk pages should suffice. Can be done via the API. I just did that using the following JavaScript snippet: var api = new mw.Api(); api.post( { action: 'purge', titles: 'Module:Property_documentation', forcerecursivelinkupdate: 1 } ). No need for server side "magic" here. - Hoo man (talk) 18:40, 18 February 2016 (UTC)

Didn't know that one. Thanks for the tip Hoo man. Mbch331 (talk) 18:42, 18 February 2016 (UTC)

Uuuh, looks impressive... Never done anything like that though. @Mbch331: Is it the same as this:

https://www.wikidata.org/w/api.php?action=purge&titles=Module:Property_documentation&forcerecursivelinkupdate=1

or should I use some other interface tool? -- LaddΩ chat ;) 19:07, 18 February 2016 (UTC)

I'm no API expert, but in this case I can with 100% certainty say the URL is correct. Mbch331 (talk) 19:09, 18 February 2016 (UTC)

Bonjour,

En passant Ceres (Q596), j'ai remarqué que le libellé était le nom incomplet car tu avais tronqué le nom complet et officiel en avril 2013 (Special:Diff/30749013). Est-ce volontaire et/ou issu d'une discussion ? Au fond, peu m'importe le libellé mais j'aurais une légère préférence pour le nom complet et officiel.

Cdlt, VIGNERON (talk) 10:53, 19 March 2016 (UTC)

@VIGNERON: Salut, mon erreur, j'aurais dû plutôt ajouter un alias  :( C'était nouveau pour moi en 2013 -- LaddΩ chat ;) 15:03, 19 March 2016 (UTC)

I keep removing some of the statements you're adding to properties because they seem incorrect to me. For example, Wikipedia is not a source website - we already have Wikidata item of this property (P1629) to link the property to the corresponding item (and its Wikipedia links). Even if someone put that link in the box on the talk page, do remember that those boxes may not have been filled in very well. I've also seen comment (DEPRECATED) (P2315) used for descriptions of the property, please use the description field for descriptions. - Nikki (talk) 11:36, 26 March 2016 (UTC)

Hi @Nikki:, regarding the WP sources, I indeed retained them if they seem to shed a good light on the subject. I agree WP is not a valid source and will stop that. Regarding misuse of comment (DEPRECATED) (P2315), its current role is pretty recent and I was one of the first to require a better property than "comment" (former name of P2315) but my requests failed. Only recently was Wikidata usage instructions (P2559) independently created. I also signalled that there was conversion work to do but have too little time to do it myself. -- LaddΩ chat ;) 19:04, 26 March 2016 (UTC)

I'm aware that it changed recently - I was the person who proposed Wikidata usage instructions (P2559). :) The sort of things I'm referring to are not usage instructions though, so neither comment (DEPRECATED) (P2315) nor Wikidata usage instructions (P2559) would be appropriate. See for example Special:Diff/305362039 - that's just a description of the property and it was more or less repeating what the label/description already said. - Nikki (talk) 21:47, 26 March 2016 (UTC)

@Nikki: Oh, those comments :) Well, while converting fields "allowed values" to properties, I was trying to retain everything that previous authors had though was worth mentioning in the doc template. Using "comment" for that seems appropriate to me. You may have a different opinion on this and remove them, though. The boundary of what is appropriate to enter as a "comment" is not well-defined. -- LaddΩ chat ;) 20:37, 27 March 2016 (UTC)

Hi Laddo,

Could you give your thoughts on Wikidata:Properties_for_deletion#Wikidata_example_properties ? You seem to be the one who worked with them the most.
--- Jura 23:49, 26 March 2016 (UTC)

@Jura1: Hi Jura, I'm afraid I have too little time to read the full discussion and grant it the thinking it would deserve at the moment. Sorry about this. -- LaddΩ chat ;) 20:51, 27 March 2016 (UTC)

Hi,

I see that you put two Constraint:Format on ISBN-10 (P957) and I have two questions :

• technically, does it works ? I thought you could use only one constraint by property for the format...
• more philosopicaly, is is really a good idea to allow ISBN without hyphen ?

Cdlt, VIGNERON (talk) 15:55, 7 May 2016 (UTC)

@VIGNERON: Avec le traitement actuel des contraintes par KrBot, chaque contrainte est traitée isolément des autres et elles peuvent se chevaucher sans problème. Vois par exemple Property talk:P94 qui utilise à fond cette propriété. Quant aux aux traits d'union, je n'ai pas changé le traitement préexistant... Il y a déjà eu plusieurs discussions à ce propos (ces deux-ci par exemple) et je n'ai que repris le format existant pour mieux capturer certaines instances erronées. À mon avis personnel, les ISBN devraient être stockés sans espaces ni traits d'union, mais affichés avec ces séparateurs; cependant Wikidata ne permet pas de contrôler le formatage pour le moment. A+ -- LaddΩ chat ;) 20:26, 8 May 2016 (UTC)

Hi,

I see you changed <cast member><subject item of this property><Actor> to <cast member><subject item of this property><cast member>. Can you explain the reasoning for this change?

Does it better reflect the purpose of <subject item of this property>, and if so why?

Thanks.

OK, changed it back. The English label was just a better match. -- LaddΩ chat ;) _to_<cast_member>?" class="ext-discussiontools-init-timestamplink">13:42, 26 June 2016 (UTC)

Hmm okay. I was just asking to better understand how the <subject item of this property> property works, since I am new to WikiData.

Thanks

Hi, after a lot of these categories were deleted in enwiki, there was a decision to recreate them, which I recently finished. When I checked a few weeks later, I was really glad to see that most of them had already been reinstated in Wikidata, by yourself and others. Thanks very much! Fayenatic london (talk) 22:21, 25 July 2016 (UTC)

Hello, this contribution was not a vandalism. It's just because I made history merge on Manikich which include the suppression and the restauration of the article. It seem that wikidata's link was deleted in the same time and was not restaured automatically. Cordially. --Gratus (talk) 12:11, 23 August 2016 (UTC)

@Gratus: I wasn't too sure, I could not figure why the link to frwiki was left dangling; however I could see from your other changes that is was not left intentionally. Sorry for my unjustified comment. -- LaddΩ chat ;) 19:27, 23 August 2016 (UTC)

Dear Laddo, I've reverted your contribution because Abon (Q35121) is for a language in Indonesia. The interwiki in frwiki seems not an language but an area* in Nigeria. --Beeyan (talk) 12:33, 1 September 2016 (UTC)

I've rechecked again, mine is wrong, you're right. I moved the information into new item. Thanks --Beeyan (talk) 12:43, 1 September 2016 (UTC)

Hey Laddo, I have become aware that you have been editing tennis-related items recently. There is a bit of a problem now, since many items of tennis events such as 2016 Brisbane International – women's doubles (Q21876593) now indirectly subclass sport (Q349), which is semantically wrong. Could you please look into this, and try to fix it? I do not want to barge in here, thus I'm asking... Thanks and regards, --MisterSynergy (talk) 17:01, 7 September 2016 (UTC)

@MisterSynergy: Hi, it looks like a mess indeed. I'll try to fix some stuff, but be aware that any event that has a date should be an "instance of" some event class. These issues were introduced by Vinkje83, who used subclass of (P279) instead of instance of (P31) on specific (dated) events. Another issue is that women's singles (Q16893403) and the like are poorly defined: these are sport disciplines, not sports. -- LaddΩ chat ;) 18:22, 7 September 2016 (UTC)

Yes I saw that it was not you who created the mess, you just did some good work on the tennis disciplines which now accidentally leads to the events as subclasses of sport. I would not worry about disciplines as subclasses of sport (Q349), since there is no clear definition about "disciplines" available. This term is used in different contexts with different sematics, so it is reasonable to permit them subclassing of sport.

It would now probably be the best to replace the subclass of (P279) "tennis discipline item" (such as women's singles (Q16893403)) statements by instance of (P31) tennis tournament (Q13219666) statements. This would technically be more correct, although I could envision even more precise ways. Shall I, or do you want to do that by yourself? Regards, MisterSynergy (talk) 18:34, 7 September 2016 (UTC)

@MisterSynergy: I'll dig to systematically fix all items described as "20* Women's Doubles" (and similar), adding P31 and removing P279, but I'm not too good at SPARQL and will check a bit further to do that with PetScan. -- LaddΩ chat ;) 18:39, 7 September 2016 (UTC)

You might want to try Autolist, which is the simplest tool for such use cases (batch removal and addition of statements). I'll then turn my attention to other things again, but feel free to ask me if you need any help. --MisterSynergy (talk) 18:52, 7 September 2016 (UTC)

Hello LaddΩ and MisterSynergy. I apologize if I wrongly applied properties. It was (and is) not my intention to create a mess. However, the discussion above might be an indication that I possibly do not understand the difference between instance of (P31) and subclass of (P279) sufficiently. I will gladly coöperate. But before I can do so, I need to obtain better understanding of the matter. Therefore I kindly request you:

1. to explain to me the difference between instance of (P31) and subclass of (P279), also comparing them with has part(s) (P527) and part of (P361)

   in particular tell me whether 'subclass' (P279) can either mean 'has subclass' or 'is subclass of' or can mean either

2. to tell me in which circumstances I should use instance of (P31); and also: why
3. to tell me in which circumstances I should use subclass of (P279); and also: why

When writing your explanation, you need not go outside the realm of tennis, as I am working on little else.

As to the above suggestion to link items like 2016 Brisbane International – women's doubles (Q21876593) to tennis tournament (Q13219666): that would be a redundant specification, as 2016 Brisbane International – women's doubles (Q21876593) is part of (P361) 2016 Brisbane International (women) (Q21876143) which instance of (P31) Brisbane International (women) (Q14551987) which is part of (P361) Brisbane International (Q264806) which instance of (P31) tennis tournament (Q13219666). Kind regards, Vinkje83 (talk) 14:58, 8 September 2016 (UTC)

Hey Vinkje83, no need to apologize. Mistakes are there to be made, so don't feel bad. We all make mistakes every now and then, particularly if we start doing something new. We only get angry if someone "creates mess" although they know better, or if they are extremely reluctant to learn.

Regarding the usage fo P31 and P279: this is indeed a complicated problem for many Wikidata users. P31 is a little easier to understand, thus I start there. You can use it basically for anything which is an individual entity of any type. Fictive examples: you are an instance of human; the device you are using to edit Wikidata is an instance of "electronic device"; your wedding event is an instance of wedding; and so on. Anything that you can touch, and beyond that many individual concepts such as the wedding example (which you can't touch) are instances of something and therefore require a P31 statement. Now to the P279: this property is used if an (Wikidata) entity does not describe a specific expression of something, but rather the concept; fictive examples: "car" is a subclass of "machine"; "tennis tournament" is a subclass of "sports competition"; on the abstract side we are really talking of classes which define the instances within this class, on the more specific side the class turn to something which is rather a quantity of similar instances of entities: an example would be: Brisbane International (Q264806) is a subclass of "tennis tournament"; it is not an instance since it does not describe a specific edition of this tournament, but it rather describes the type of this tournament in Brisbane which is held annually.

"Part of" and "has part" properties are easier to use -- just follow your intuition. They are useful to relate items which are actually part of a bigger entity, but split into individual items for whatever reason. --MisterSynergy (talk) 19:38, 8 September 2016 (UTC)

@Vinkje83: Indeed, mistakes resulting from good faith edits are well understandable from editors that are less familiar with the concepts. No worries! You may find an interesting overview at Help:Basic membership properties.

Note that has part(s) (P527) and part of (P361) can create relations but do not help in understanding the nature of what they relate. P31 and P279 are used for precisely that.

In the case of tennis tournaments, let's consider WP category Category:Eastbourne_International: that category lists all articles relating in some manner with that "event class". Note the main article, Eastbourne International, representing the entire family of such events. In Wikidata, the correct relations would be:

• Eastbourne International (Q431569) subclass of (P279) tennis tournament (Q13219666), creating a "class of tennis competitions" that applies to all other articles that where added to that WP category. Eastbourne International (Q431569) is NOT (not P31) a specific event of a given tennis tournament, but rather a type, or class, of such tournaments. Thus Eastbourne International (Q431569) should not bear P31 at all.
• Say, 2014 Aegon International (Q16973113) instance of (P31) Eastbourne International (Q431569), indicating that it is an instance of that tournament class (one of the member of that family of tournaments); it indeed appears in Category:Eastbourne_International; such specific competition instances should additionally bear some time component (point in time (P585)=2014) and some other properties;
• Things are not too clear regarding 2014 Aegon International (women) (Q17122089) or 2014 Aegon International (men) (Q17252919): these are "sub" tennis competitions, often held separately and organized by different associations (ATP vs. WTA). If I was to get this straight, I would create Wikidata items WTA tennis tournament and ATP tennis tournament (both subclasses of tennis tournament, but having different organizer (P664)) and make both 2014 Aegon International (women) (Q17122089) and 2014 Aegon International (men) (Q17252919) respective "instance of (P31)" these new high-level classes;
• Finally you got items like 2014 Aegon International – men's doubles (Q17150201), that up to now I was just making instance of (P31) men's doubles (Q17299700) and "part of" their respective tournaments.

Running out of time, got to go, hope it helped. LaddΩ chat ;) 20:23, 8 September 2016 (UTC)

Hi,

In June 2013, you add these two claims :

• ⟨ Cephissus (Q1235405)    ⟩ named after (P138) ⟨ Cephissus (Q1478704)    ⟩

and the reverse

• ⟨ Cephissus (Q1478704)    ⟩ named after (P138) ⟨ Cephissus (Q1235405)    ⟩

This seems very strange to me as this is self-referencing (A named after B named after A named after B, and so on forever) ; I guess only one is correct, but not sure which one... (maybe the river god before the river ; or maybe P138 is not the right property?)

When you'll have some times, could you look into it?

Cdlt, VIGNERON (talk) 11:53, 14 September 2016 (UTC)

@VIGNERON: Merci pour le signalement, sans doute une distraction de ma part, en 2013... Ça m'a permis de corriger deux autres erreurs dans Cephissus (Q1235405). A+ -- LaddΩ chat ;) 22:06, 14 September 2016 (UTC)