Olivia GuyotJust discovered https://giro3d.org/, looks awesome and really reminds me of a discussion several years ago with @erilem 🙃
autra@oguyot_c2c part of Giro3d design is clearly inspired from discussions I had with @erilem too! He initially helped us integrate the openlayers sources, which allows us to support standard services without too much effort.
Eric Lemoine@autra @oguyot_c2c That seemed so obvious to me at that time 🙂
@autra @oguyot_c2c giro3d looks very cool!
@oguyot_c2c @erilem Maybe, but my current opinion on this: I'm not sure it's a good idea. That being said, I'd love to talk about this, because I don't know the openlayers code very well and I might be wrong on this assessment.
@autra @erilem would love to find the time and talk about this a bit more... Renderers are part of the OL API so it should definitely be possible to extend them. The biggest problem might be that an assumption for 2D is done in many places (e.g. shaders). Also maybe it just doesn't make sense to rely on another library for something so critical, as it would mean heavily dependant on it.
@oguyot_c2c @erilem I don't mind depending on another communautary project, if our voice can be heard. I'm afraid about 2D assumptions, yes! Also even for sources, we already have difficulties, because we use raster source as DEM information. For COGs, last time we checked, openlayers did process them too much for us (especially COGS with one 32 bits band if I remember correctly?)
Sébastien Guimmara@autra @oguyot_c2c @erilem for vector sources, we actually let OL rasterize the source tile/vector data into a texture, so I guess it counts as using OL renderers ?
@sguimmara @autra @erilem it does but for example being able to get webgl array buffers from renderers instead of the final image could be also interesting!