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
So the renderer is basically rewritten now, though I'm still not anywhere near happy enough with it. Three.js simply isn't made to handle what DTile needs here. We are gonna ditch Three eventually, though I'm not sure for what. One option which I would really love is to have it be written in Rust with maybe gfx-rs or something like that. That would allow for some really high performance which I think is almost necessary for some of the things I have planned for DTile. Here some things that needs to happen before that can happen though:
Polymer 3: we need to have ES modules to easily be able to integrate with the rest of the JS ecosystem such as...
Parcel: while the original intention of using Polymer was that you wouldn't need to build DTile, the fact is that it has caused more problems than good. Parcel seems like the perfect middle ground here with fast, easy builds with features like...
Rust integration with parcel: parcel can already automatically build Rust files automatically but you can only pass numbers between JS and Rust which isn't gonna cut it for DTile.
If that doesn't end up being possible anytime soon though, using something like PicoGL.js and building our own renderer in JavaScript doesn't seem to be too bad of an option either.
The renderer is quite a mess currently and needs to be rewritten, probably from scratch.
Here are some things the new renderer has to be capable of:
The text was updated successfully, but these errors were encountered: