Tracker: deck.gl v9.2

[ibgreen]

deck.gl v9.2

• Target Release Date: Q2 2025
• Follows v9.1 and v9.0

New v9.2 Features under consideration

General

• luma.gl 9.2 - Production WebGPU support - Tracker
• WebGPU - A few layers running on WebGPU - Tracker
• Widgets - More widgets, more capabilities, better docs - Tracker

Geospatial domain

• ❌ TBD: PMTiles support? - e.g. via "TileSourceLayer" from @deck.gl-community
• ❌ TBD: Dynamic vector tiling layer? - render very big geojsons (via TableTileSource)
• ❌ TBD: GeoParquet support? - e.g. example showing Overture or Foursquare datasets

Infovis domain

• 🚧 See @deck.gl-community/infovis-layers

Biology domain?

• ❌ TBD

Port Layers to WebGPU

Geospatial layers

• ScatterplotLayer
• PointCloudLayer

Chores

• Cut deck 9.1-release branch, move master to 9.2.0-alpha
• Lock deck 9.1-release to luma.gl@~9.1.0

[ouzhou]

Is this plan going to be postponed？

[ibgreen]

Is this plan going to be postponed？

@ouzhou Thanks for reaching out.

I updated the target date, and yes we are sometimes too optimistic about our release dates.

However, it is not for a lack of activity. We are currently busy putting the final touches on deck.gl v9.1. (In fact you can follow along in the #openvis-v9 channel in our slack if you are interested: https://www.openvisualization.org/#get-involved)

[ouzhou]

Thank you for your reply
I use DeckGL to do some graph visualization to render a large number of points and edges. The performance of DeckGL is excellent.
But I found that the performance advantages of WebGL were not enough, so I started trying to use LumaGL with webgpu to directly implement it
I would like to ask why the LumaGL project does not all in webGPU, so that there is no need to maintain webgl?