You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A longstanding item on the wishlist has been to enable drawing a box on the map and showing periods with spatial footprints within the box. We've put this off because finding the intersection of a box with the spatial coverage of ~7000 periods in pure JS is likely to be very slow without some kind of dedicated spatial index.
However yesterday it occurred to me that another option could be to simply add box selection to the existing UI for selecting gazetteer places to filter by. There are a lot fewer records in the gazetteer, so this could be pretty quick. It would basically work just like the existing place selector, except that rather than searching for places by name, you would draw a box on the map to get a set of places to filter by.
The text was updated successfully, but these errors were encountered:
I think this is a fantastic idea. Essentially, it'll do the same thing, right, since the set of gazetteer places within the box will then filter the period list? So you'll still get a list of periods with footprints within the box, but the footprints will be inherited from the gazetteer entries they're associated with.
Like the current version, it will only grab gazetteer entries with spatial coverage (so lots of LCSH periods, for example, won't show up -- our "Roman republic" spatial location is mapped to a Wikidata entry that is itself a period and not a place, I just realized, and in any case doesn't have an associated geometry: http://n2t.net/ark:/99152/p06c6g3hb99). This would be true of any spatial filter, but the bounding box might require us to make it clearer: since the current UI has you search by term, you can see when your search term doesn't appear, whereas drawing a bounding box might provide an illusion of comprehensiveness.
Would we replace the search by name with the bounding box, or would we have both as alternates?
We’d need to keep both, since as you point out some placenames don’t have footprints. One way to deal with this could be to associate each place without a footprint with another “containing” record for which we do have a footprint. In any case I think we still want to keep search by name. We’d definitely need to think carefully about how to make it clear what was going on in the UI to avoid confusion.
A longstanding item on the wishlist has been to enable drawing a box on the map and showing periods with spatial footprints within the box. We've put this off because finding the intersection of a box with the spatial coverage of ~7000 periods in pure JS is likely to be very slow without some kind of dedicated spatial index.
However yesterday it occurred to me that another option could be to simply add box selection to the existing UI for selecting gazetteer places to filter by. There are a lot fewer records in the gazetteer, so this could be pretty quick. It would basically work just like the existing place selector, except that rather than searching for places by name, you would draw a box on the map to get a set of places to filter by.
The text was updated successfully, but these errors were encountered: