Wikidata

Wikidata:Property proposal/Wikisource index page

< Wikidata:Property proposal

Wikisource index edit

Wikisource index page edit

Originally proposed at Wikidata:Property proposal/Sister projects

   Not done
Descriptionitem for page in Wikisource containing digital and paper pagination of the source
RepresentsHelp:Index pages (Q15628590)
Data typeItem
Example 1Use here on Wikidata: The Kiss and Other Stories by Anton Tchekhoff (Q15839163) → Index:The Kiss and Other Stories by Anton Tchekhoff, 1908.pdf (Q89675998)
Example 2Use at Wikimedia Commons: File:The Kiss and Other Stories by Anton Tchekhoff, 1908.pdf → Index:The Kiss and Other Stories by Anton Tchekhoff, 1908.pdf (Q89675998)
Example 3MISSING
See alsoWikisource index page URL (P1957)

Wikisource pagination string edit

Originally proposed at Wikidata:Property proposal/Sister projects

   Not done
Descriptionpagination of a file in the pagelist format used by Wikisource index pages. See property talk page for formatting details.
RepresentsHelp:Index pages (Q15628590)
Data typeString
Allowed valuessee s:Help:Index_pages#pagelist
Example 1Index:The Kiss and Other Stories by Anton Tchekhoff, 1908.pdf (Q89675998) → <pagelist 1to5=- 6=1 37=33 58=55 69=67 70=69 80=81 107=109 120=123 133=137 172=177 181=187 188=195 189=197 212=221 221=231 222=233 233=245 252=265 253=267 />
Example 2MISSING
Example 3MISSING
See alsofile page (P7668), title page number (P4714)

Motivation edit

This was initially done with URL datatype (P1957), but, as for other sitelinks to sister sites, using actually item datatype seems preferable. I don't think these pages are currently linked from Wikidata. Accordingly, we don't really now if P1957 values are complete/how incomplete they are. Eventually P1957 could be replaced by this. Except for the pagination, I think all other information is included in Wikidata. Please help complete the proposal. (Add your motivation for this property here.) --- Jura 15:06, 7 April 2020 (UTC)[reply]

BTW, this could also be used on Commons (see 2nd sample above). --- Jura 17:13, 7 April 2020 (UTC)[reply]


@Mike Peel, Beleg Tâl, Samwilson: --- Jura 15:06, 7 April 2020 (UTC)[reply]

Discussion edit

  • If the Index page property was of type Item, we'd have to create separate items for each Index page, but Wikisource index page URL (P1957) is already meant to be used on items describing editions. I don't know there's much to be gained from having to have separate items for e.g. each volume in a multi-volume edition, let alone a separate item for every single edition (there's already enough confusion with creating edition and work items). My understanding is that Wikisource index page URL (P1957) is of type URL because there is no other type that's suitable. Really, it'd be good to have something like Commons media but for any wiki; that's a whole other thing though, and it's not possible at the moment. —Sam Wilson 00:23, 8 April 2020 (UTC)[reply]
    • I don't really have a good explanation for why P1957 is using URL datatype, but it seems to date back to the time when Wikisource was just rolled out and it wasn't clear how sitelinks were done in general. We gradually started changing properties for Commons from odd datatypes to the usual item ones. I don't really see a potential for confusing items like Index:The Kiss and Other Stories by Anton Tchekhoff, 1908.pdf (Q89675998) with others (note the "Index:" prefix), at least if we define them adequately. --- Jura 06:11, 8 April 2020 (UTC)[reply]
      • @Jura1: I don't think we should have sitelinks to Index pages. That would mean we'd have to have separate items for every Index page. The reason Wikisource index page URL (P1957) is of type URL is that there's no general 'wiki page' data type (only the special purpose Commons media type one, which is why Commons properties get special treatment). —Sam Wilson 23:09, 8 April 2020 (UTC)[reply]
        • It's actually an advantage to have items for index pages. It would allow to determine for which editions ones we currently lack links (compare e.g. only s:Category:Index Validated has 3500 pages, with 2900 uses of Wikisource index page URL (P1957)). Clearly there is a lot of content at Wikisource that isn't discoverable from Wikidata. BTW we also started changing the property for Commons to item properties (e.g. Property:P3722 was replaced by Property:P7867 recently). @Mike Peel: what are your thoughts on this? --- Jura 09:08, 9 April 2020 (UTC)[reply]
        • While the new property involves some work (I'd be doing), I don't really see any downside for existing users of the current property. Are there some uses that wouldn't be possible? --- Jura 09:08, 9 April 2020 (UTC)[reply]
          • @Jura1: I'm somewhat neutral here. On one hand, sitelinks and links between QIDs are way better than strings for keeping track of interwiki links, on the other hand you would have to create a lot of new items and I'm not sure they would have many other links. I think the case for sitelinks would become a lot stronger if Wikidata information was being used on the pages being sitelinked to - then the information can easily be accessed without messing around with QIDs locally. This is what I consider the key use case to be for Commons and why I wanted the new category for maps (P7867) (each one of those items is used to display an infobox in the sitelinked Commons categories). I'm not sure what the current Wikidata situation is on Wikisource, @Billinghurst: would know more. Thanks. Mike Peel (talk) 09:26, 9 April 2020 (UTC)[reply]
            • @Mike Peel: Thanks for your feedback. I think the Russian Wikisource is fairly well integrated with Wikidata. Information there can be found through Wikidata (with items for Wikisource articles and "described by" statements). Its community seems to be fairly proficient in Wikidata tools. A lot of work is also being done for French Wikisource. I don't think this is necessarily the case for other languages, e.g. poems by Emily Dickinson were mostly malformed entries. So some of the wealth of information on Wikisource is lost to the wider Wikimedia community. I'm trying to support the integration of an encyclopedia with more than 25,000 entries. I'm not sure how much use the Russian or French Wikisource editions make of Wikidata. Maybe storage of some of the metadata. It might also be helpful for some checks. In general, the advantage for them is probably rather the accessibility of the items about the sources through queries. --- Jura 09:52, 9 April 2020 (UTC)[reply]
  • For the Wikisource pagination string proposal: I think this has merit, but this data belongs in Commons where the scans are kept. The items here are for editions, and do not correspond to particular scans of particular books. I'm not familiar yet with the process of proposing a new property for Commons; is this that process? If so, then great. —Sam Wilson 00:23, 8 April 2020 (UTC)[reply]
    • It would need to be proposed here as well, but they seem to be reluctant to support string properties on Commons. So the only option seems to be to host this here. The main property can link the item here from Commons (see 2nd sample above). --- Jura 06:11, 8 April 2020 (UTC)[reply]

  Comment Wikisource Index: pages are not content pages at the WSes, they are simply constructs/carriers to manage a transclusion and they link a file to a transclusion providing page numbers on the way, so they have no particular importance outside of the WS. The thing about Index: and File: pages are they may not be complete edition, so for a work at a Wikisource there may be multiple Index: or you may find that an Index:/File: contains multiple works, depending on the scanner [So an edition may have multiple Index:, or an Index may have editions of multiple works]. To also note that an Index: / File: may not be unique to just one WS, where a WS allows for the Index: / Page: nss to be able to be utilised for a crowd-sourced translation of the published work you could find the same file used at two places, and have two connected Index:.

I don't think that Wikisource index page URL (P1957) is used much, and I haven't particularly used it for a while. [If a field isn't in WEF framework tool for "FRBR Edition", I don't add it.] Please don't make it harder, or more work for the WSers, especially without the deliver of a greater benefit. There is already need for better tools to make the additions, and there has been little effort to undertake that, so we are stuck with brute force edits. We still cannot make easy good WORKS from ITEMS, and ITEMS from WORKS. You cannot easily load a work to Commons, and get the resulting componentry elsewhere without a lot of work. [My epistles on this three way linking of editions to Commons/WS/WD, then linking to works for WP are around. Now the suggestion is for another item for the Index:. I hope that one day the thinking will focus on giving better tools to WSes and that will give better data and compliance.] At this time there are just too many hurdles in getting good data from WSes into WD.  — billinghurst sDrewth 10:33, 9 April 2020 (UTC)[reply]

  Comment I am not seeing the point of the proposal and what it achieves. The data in the Property Proposal is locally created, and should be considered as dynamic, rather than static, and pagination and numerous means of presentation, so why would you inhale the <pagelist> data and how would you continually check it for accuracy? The Index namespace at the Wikisources is our work zone, not a content namespace, and only truly has relevance to a WS working edition, not to anything external to wikimedia wikis. I just see a time pit; not a proposal of clear value to Wikisource.  — billinghurst sDrewth 00:38, 13 April 2020 (UTC)[reply]

  • It all depends on your needs. It could simplify creating the edition items you mention. If the index page is complete, it could include the equivalent information included. I'd assume it remains static once the work is proofread and validated. I can understand that, for Wikisource contributors who have the index page, there is no need for any item(s) at Wikidata as the Index page already holds all information. However, users of Commons/Wikidata can't access that information easily nor know it's there. --- Jura 11:38, 13 April 2020 (UTC)[reply]
I am also kind of   Neutral. I agree with billinghurst that index page is something relevant wot Wikisource and not that relevant to anybody else. I do not understand how we would use it. I do think we should link to index page, but I would rather skip the step of creating wikidata item for it and just store URLs, as I can not imagine what other metadata we would add to that item. --Jarekt (talk) 16:02, 21 July 2020 (UTC)[reply]
@Jarekt: I think the pagination of the PDF is important (the main feature of these pages IMO). This is true independently of a transcription at Wikisource.
There are obviously several ways to store the pagniation. I don't think having it at an unconnected Wikisource page is one of them. --- Jura 16:21, 21 July 2020 (UTC)[reply]
I agree that it is important, but we do need some idea about how to use it before we decide what we need to store and how. I would not want to be donating my time curating metadata nobody will ever use, and I think others might feel the same. If we do come up with some realistic use-cases for it than I agree that we should store it. --Jarekt (talk) 17:10, 21 July 2020 (UTC)[reply]
@Jon Harald Søby, Tpt: I found you through https://fr.wikisource.org/wiki/Discussion_MediaWiki:Proofreadpage_index_template . Is it possible to display the pagination on file description pages if we add a statement on Commons? I couldn't really figure ou from the French module ( https://fr.wikisource.org/wiki/Module:Index_template ) how it does that. Obviously, the interested would be for the many files on Commons without any Wikisource using them. --- Jura 17:29, 21 July 2020 (UTC)[reply]
  Comment @Jarekt, Jura1: Note about pagination. It is part of the summer coding project and new pagination processes are coming through. I will also note that the pagination style is very arbitary and changing, so having it statically recorded in WD is the wrong way to go about it. Just like our transcriptions can update through time, so can our jottings on page numbering, and I certainly do not want another task of transferring that data when it would be better stored in a json file on the WS as part of the Index  – The preceding unsigned comment was added by [[User:|?]] ([[User talk:|talk]] • contribs).
  • I’d like to see bilingual texts mentioned here (e.g. Q72743941):
    – Switching to an item for the index might help navigation between index pages once setup (assuming the two indexes be sitelinks of the same items, not sure the two indexes can be considered the same concept, though?).
    – What would pagination be if shared by two index pages? Q72743941 is maybe not the best example here, because the polish version could be used as a common one: What about unshareable conventions, like pages named “Titre” (= “Title”), “illust”, “tdm” (= “toc”), etc. on the French Wikisource? (or would pagination-on-Wikidata be a third version completely independent from Wikisource?)
    Jura1, could you clarify what you expect here?
    — Ltrlg (talk), 15:28, 13 September 2020 (UTC)[reply]
    • @Ltrlg: if it couldn't be shared, I suppose one would have two add to statements with the property to differentiate the two. Oddly I'm more concerned about the many Commons files that wont ever get an index page on Wikisource. --- Jura 15:19, 22 September 2020 (UTC)[reply]
  •   Support, an important property for Wikisource.--Arbnos (talk) 17:05, 19 February 2021 (UTC)[reply]
  • @Jura1, GZWDer, Samwilson, Mike Peel: @billinghurst, Jarekt, Ltrlg, Arbnos:   Not done Main body of discussion did not indicate support, proposal now stale. ArthurPSmith (talk) 17:20, 9 September 2021 (UTC)[reply]
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Property_proposal/Wikisource_index_page&oldid=1495137732"