FactGrid:Troubleshooting: Difference between revisions

From FactGrid
Jump to navigation Jump to search
Line 57: Line 57:


= Specific problems of our Wikibase Installation =
= Specific problems of our Wikibase Installation =
=== GND Identifiers do not link ===
== GND Identifiers do not link ==


Why does it not work like a link as in Wikidata? Why can't we click at the number and access the dataset on the GND database?
Why does it not work like a link as in Wikidata? Why can't we click at the number and access the dataset on the GND database?


=== Query Service and full years ===
== Query Service and full years ==
If we enter full years with the appropriate distinction 1761-00-00etc. - why do we still get a January 1 date?
If we enter full years with the appropriate distinction 1761-00-00etc. - why do we still get a January 1 date?


=== [[Item:Q22976]] or just [[Q22976]] ===
== [[Item:Q22976]] or just [[Q22976]] ==


=== Wikidata has a function that allows quoting items via {{}} template, can we do that as well? ===
== Wikidata has a function that allows quoting items via {{}} template, can we do that as well? ==

Revision as of 11:19, 20 November 2018

Wat would be the best practice data structure

Repeated information - avoid it or go for it?

We have some 7000 documents in the database. Each document has an author, most documents have recipients and most documents mention other persons.

  • 375 documents are written by Johann Joachim Christoph Bode search.
  • 1123 documents have Bode as first recipient search
  • 108 documents refer to Bode as third person search

Organising Bode's data set Item:Q133: should we list all these documents in three sections? It might be interesting to do this because this is the point at which we show that we have far more information on the man than any other database.

I am otherwise not to eager to begin this because this was the reason why we left the Wiki and went for the database: The ideal situation is that we have every fact only once in the database. A data sheet on Bode should fill the "Bode as Author" section with a look into the database. Once we start creating lists we have to make sure that they are all constantly up to date and well organised - and a Bode data sheet with a list of 1123 documents he received will no longer be funny.

The ideal situation is a split of functions: A Bode page is created in data base retrievals. You geht the list as soon as you click at "more" freshly made from the data base. (There is a tendency to use Wikibase much like a Wikipedia page: have everything on Bode on "his page" and understand that the dataset Item:Q133 is all we know about Bode...)

Membership information: have the information on person? on the organisation or on an intermediate membership item?

We have about 1350 Illuminati serach. Is being a member a personal thing, to be stated in the person's data set? Is it a thing on the account of the Order to be stated here Item:Q10677 - or shall we create items for each membership and refer to these items on both sides, person and organisation?

A particular problem is that mere membership information will not be the end of the story. Members progress in the order, they get degrees and reach positions. We will not be able to list these details as qualifiers of the respective membership (member of the Illuminati, begin date, place, proposed by, advanced to position x, date... [Wikibase will list all dates together])

Membership items could be complex but would they be operable? Would users know what to look for? --Olaf Simons (talk) 10:45, 20 November 2018 (CET)

Changing entities and cuts through time

See Münchner Buchhandel 1500–1850 at Wikipedia for the problem.

The Publishing House of Moritz Georg Weidmann becomes eventually the Olms Verlag - if we treat this as one company - how do we create different references for different times? Books have different imprints decade after decade - especially with the early modern practice to use the publishers' personal names in the imprints. Our system must be able to make sense of practically any imprint we find - and it should at the same moment understand how the changing imprint information is actually connected to existing companies. Munich does not have 57 companies from 1500 to 1850 but six and these are interconnected through family ties.

  • How do we give the proper picture in time segments?
  • How do we best organise data on an item that goes from 1500 to the present?
  • Should we use the Wikipedia redirect function (via merge items) to create various items that are all leading the the once central item?
  • should we create different items for each imprint and state how these items are interconnected? (if so: how do we make sure that people using the QueryService know our complex data structures?

Recipients 1, 2, 3 - a good idea?

Regular communication from members into the Illuminati Order was a complex thing. Members met one a month in their local Minerval Church and handed in their monthly Quibus Licet ("to whom it may concern"), letters into the Order. The "local superior" would be the first reader. He would pass the letter to the Provincial leader who would answer as Basilus, the "Unknown Superior". The local superior of Gotha would be Christian Georg Helmolt (the first recipient), he would send the package of Gotha's "QQLL" (the abbreviation for the plural) to Johann Joachim Christoph Bode in Weimar (recipient 2) who could in turn pass the letters to Gotha's duke Ernest II. (recipient 3 in this case).

In order to link the respective places to which the letters were sent to the recipients I created different recipients which could now be spotted in different places.

One could organise this differently. User:Daniel Mietchen proposed to create an item for each transportation process. Would this be clever? How should we organise things with the aim to facilitate searches? --Olaf Simons (talk) 21:42, 6 August 2018 (CEST)

Alternative again proposed by User:Daniel Mietchen: take a look at Wikidata's P1545, series qualifier.
It all boils down to the question of best search results and optimal linking. One would love to see the bundling of information on higher levels of personal power. Search results again are a question of expectations: User have to get the information even if they do not know more about the availability. (That's why recipient 1-2-3 was probably bad - you do not know that these exist...)

FactGrid Items have an Item: prefix.

What should we do with our 1500 transcripts of documents

We are in need of a good place to offer transcripts (and eventually: scans).

Would Q22976 be a good place for the transcript of Item:Q22976?

Would it be better to have a direct input field for the transcript of full pages?

Could the reasonator as a standard unser interface grab transpripts from such wiki pages?

Could we use the Wikisource extension to offer scans side by side?

Specific problems of our Wikibase Installation

GND Identifiers do not link

Why does it not work like a link as in Wikidata? Why can't we click at the number and access the dataset on the GND database?

Query Service and full years

If we enter full years with the appropriate distinction 1761-00-00etc. - why do we still get a January 1 date?

Item:Q22976 or just Q22976

Wikidata has a function that allows quoting items via {{}} template, can we do that as well?