![]() In Data load editor, you can optionally combine coordinate fields into a single field for convenience. When loading map data in Data manager with data profiling enabled, the data profiling service will identify country names, city names, and latitude and longitude fields and load the corresponding geometries into new fields. Fields containing non- WGS-84 coordinates (when using a custom map as the base map).Fields containing geodata in GeoJSON, LineString, or MultiLineString formats.Fields containing geopoints (latitude and longitude) in WGS-84.įields containing geopoints, polygons, or lines from a geographic data source such as a KML file.Name data in fields to place locations in map layers.To be able to create a map visualization, you need access to geographical data that connects to the data in your app. Loading point and area data from a KML file.Supported name data for fields in a map visualization.A quick survey of the landscape suggests this is the obvious best choice for place identity, should WikiTree decide to connect to an open data source for this. Update, January 2018: Mapzen is folding but Who's On First will continue. I may prototype an external tool akin to the annotation system described by Peter Thomas, but I'm under the impression that some evolution of location fields is already underway and would rather align with that if the API can support it. Nominatim is also extensive but doesn't even include the idea of historic places.Īt this point, I'm only hoping for discussion. GeoNames is very extensive and includes the idea of historic places, but does not incorporate timespans or succession. MapZen's Who's On First is basically perfect for this use, except their historic records are woefully incomplete. I've been "paper architecting" this for a few days, and have discovered only three viable open data sources for schema inspiration or data bootstrapping: ![]() I suspect, though, that the full potential of place management will require more structured data. I am not familiar with the WIkiTree codebase, so will not pretend to know. It is possible that Categories can serve this purpose, and we need only to add some UI to location field editing. In the long term, I hope WikiTree can "give back" to these data sources because we here are more motivated than most to define historic places. The gazetteer might be bootstrapped from another open data source (see below) either by pre-populating or via the location editing UI using external lookups. These places would include internationalized names, timespans, geo coordinates with precision, parent and child places, etc. Webtrees (nee PhpGedView) uses a similar approach. Rather, I propose location fields can optionally reference a Place ID, which points to the details about that place in a single "database" of places. Worse, it implies false precision: what coordinates do I enter for an ancestor born in "Germany"? This would make queries for "profiles/events nearby" much more expensive than a hierarchical reference gazetteer. To be clear, I am not suggesting the direct use of geo coordinates in location fields. Searching could be improved beyond what Categories can support.Įntry of location fields could gain auto-completion to reduce errors.ĭata validation could flag places that didn't exist at the relevant time. Maps could visualize migrations or be used as a discovery tool. I have been scouring G2G here and see that there are a lot of opinions about location tracking, but I'm afraid I want to add an option not yet explicitly discussed: a multilingual, hierarchical, historic gazetteer of place names. MyHeritage maps has renewed this interest. One of my personal goals is to visualize the geographical context of my ancestry.
0 Comments
Leave a Reply. |