I publish today the free software #Jekyll#MapLibre (pre-release; MIT license) – a plugin for static websites for easy self-hosting of maps.
I proved to myself that privacy-friendly self-hosted (as opposed to #GoogleMaps in the cloud) #maps for displaying locations on websites can be accomplished in a few days with minimal resources. In the future, I will consider this as my new baseline when looking at #GDPR Art. 25 (data protection by design and by default).
I lost a lot of productivity this year to the assumption that Esri tools would operate consistently, have their basic functions tested, and actually work out of the box.
2024 is gonna be the year we start the Great FOSS Migration, building out some #MapLibre or #OpenLayers pages for the public, non-editable stuff, using file-based data layers like #flatgeobuf and home-grown vector tiles.
https://protomaps.com/ and PMTiles are brilliant - you can serve up a vector tile file of the entire world (100+GB) and have it display in browsers using HTTP Range header tricks to download just the subset needed for the current area
I figured out how to create a subset tile file for Half Moon Bay (just 2MB) and serve it using maplibre-gl
I am thinking about making some custom web maps for my org using #MapLibre or something, but I have a question.
Is anyone using a static site generator for a #gis site? I have some SSGs that I like, but they seem to be better suited for longer, scrolling stuff, not full-screen maps.
I remember when MapBox changed their license and subsequently Open Source version @maplibre was forked. I didn't follow the project after that, as I didn't have mapping needs at the time, but now that I looked at it, it has started very well.
#MapLibre now has a governing board and it seems pretty well run for an OSS project. Gladly checking it out now! https://maplibre.org/
I think one area where visionOS perhaps dropped the ball is MapKit. Rendering 3D maps seems like a compelling use of volumetric apps in spatial computing, and figuring this stuff out would also necessitate providing tools for view elements to make layout decisions based on whether they're running in a flat window or a volumetric context. As far as I'm aware, there's nothing like that on visionOS. A SwiftUI Map view is just a 2D map. There's both a user and a developer story here to be told
Anyone with #MapTiler experience? How can I handle overlapping markers coming from a GeoJSON source so that they have a proper distance added to each other?