Talk:Proposed features/Relation:person (rewrite)
Rename the type ?
"religion" might be a problem. It should be better specified on the cemetery itself when usually it is divided by religions (cemetary=sector) (thus, it can be deduced for all graves inside).--Pieren (talk) 08:13, 16 October 2014 (UTC)
- +1. Should have a link to wikidata, that is a wikidata=*-element. --Gormo (talk) 12:48, 23 October 2015 (UTC)
Other possibilities to map without relation
Perhaps we could add that other possibilities to add several persons on the same grave exist. Eg. concatenate all names with ";" separator on the same node. Or trace a polygon for the grave/tomb and put one node per person (like we do for buildings with multiple shops). --Pieren (talk) 08:13, 16 October 2014 (UTC)
- I have no problem with one node per person. Don't think an polygon around is needed.--Zuse (talk) 10:34, 16 October 2014 (UTC)
- The polygon is better for the localisation. The node 'person' is inside the polygon 'grave' which is inside the polygon 'cemetery'. A grave can have its own identifier (like a ref). Of course, like for shops in a town, you can start mapping without the buildings but it's better with them. --Pieren (talk) 11:02, 16 October 2014 (UTC)
- I like your suggestion. It can make things easier to work with. Though I think most cases could be tagged on the grave itself, without an additional object. --Jgpacker (talk) 11:59, 16 October 2014 (UTC)
- We don't need a relation for that, we can simply tag the features to give some person names (or wikidata) when they are directly related to the subject and this link if locally verifiable (and not just a mere birthplace name shared by lots of people).
- If we need to cite several persons, instead of using ";" separation we can just add ":number" suffixes to key names, in order to index them in related tags for a single geographic OSM feature (the order should not matter but the number will need to match together in multiple related tags and need to be unique by person). — Verdy_p (talk) 14:20, 15 July 2017 (UTC)
For the same reasons as the first proposal:
- we don't need to collect various OSM features related to a person in a collection that could easily span over multiple continents or very wide areas on Earth, as well it's impossible to be complete and accurate on such collection of features with various nature, note necessarily named after them or known with many variants (towns, regions/states, schools, museums, tombs, pieces of art, places of birth/death, places of activity, brand names....)
- specific data about a person is actually not geographic, but will better fit in a Wikipedia article, or in Wikidata (for referencing additional wikipedias or book collections in Wikisource or media in Commons, or citations in Wikiquotes, or many other properties): Wikidata will link the various places that we will bind individually on geographic features.
- there are lot of things we can't manage in OSM about people, including privacy: it is already hard to avoid abuses only with phone numbers, emails, or URLs to social networks with unfair or illegal or damaging publications about a person.