Closed martindholmes closed 3 years ago
I've added a bunch of date attributes (@from-iso
and @to-iso
) to the UVic Campus map test data XML, and tweaked the XSLT to import them into the GeoJSON. Remaining work should take place in a branch.
The plan for the HTML generation goes roughly like this:
Now working in branch issue-28-timeline since I'm expecting to break the JS once in a while.
Various upcoming and existing projects have a particular need for a timeline filter, and our discussions in HCMC have come to some consensus about what this should look like. This is an outline of the feature:
@from
and@to
. Since historical dates are famously sketchy, we need to allow for ranges within those attributes, using the ISO 8601 slash delimiter for this, and we presumably also need to include times (to plot journeys that may take place within short periods of time). Our schema needs to be extended to support this.@to
and@from
attributes into properties of the location in the GeoJSON, and additionally, would calculate which of the defined periods a location overlaps with, encoding a list of those as a string array in the properties of the location. (Note: is this too repetitive? Should these periods be encoded somehow globally? It's not clear to me that it's even possible to add custom members to the FeatureCollection object in GeoJSON, though.)With these two components in place in the schema, we can then provide a slider on the rendered map which when moved, hides and shows features based on whether their time range includes the current slider date/time position. Additionally, in conjunction with the slider, if the second feature is present in the source XML and in the GeoJSON, then a second line control with (potentially overlapping) era/period labels would be created, enabling the user to click on a period and have all the relevant features displayed. However, bearing in mind that it's perfectly possible to construct a taxonomy/category structure for the era/period filtering, it may be less cluttered and more straightforward to treat eras/periods as just another category for location features. So initially, I propose to move forward with the simple timeline slider component, and when that's working, have some discussions about whether the era component is worth realizing in a manner distinct from a regular taxonomy.