Announcing MapLibre Tile: a modern and efficient vector tile format
27 points by louwers
27 points by louwers
No very familiar with the space, but is this the point where the community-led version overtakes the corporate-backed version? (At least for vector tile format)
How does MapLibre Tile compare to PMTiles (which is what Protonmaps is based on)? Neither the blog post nor the paper mention this (IMHO) obvious competitor.
It's not a competitor! :-) MapLibre Tile is a vector tile format, PMTiles is a tile container format (just like MBTiles).
You can store MLT vector tiles inside PMTiles. Well, we still need to agree on a magic number. But when https://github.com/protomaps/PMTiles/pull/596 is merged you'll be able to.
Super impressive work here and in the MapLibre ecosystem in general. Thank you.
Question: How open is the door to non-Mercator projections, e.g. WGS 84 / EPSG:4326? I ask because I’ve been working on some projects that need to be able to render data beyond the limit at ~+/-85º imposed by Mercator. I know that the rendering engines use Mercator internally and that, although non-Mercator support is in the roadmap, there hasn’t been much movement on it lately as far as I’ve been able to see. I’m hoping that more doors are opening to this possibility than are being shut. It seems that the spec is mostly agnostic, but if the tile pyramid is necessarily square (that is to say, an equal number of latitudinal tiles as longitudinal), that could be an obstacle to EPSG:4326, for which tile pyramids typically have twice as many tiles longitudinally.