Andreas Reich

This is pretty much by design and document as such (although maybe not easy to find :/). `rerun::Collection` tries to borrow whenever possible to avoid copies. The way this works...

I'm not entirely sold on the idea that passing a pointer signifies borrowing data of the underlying vector well enough. Since rerun collection takes internally a pointer to `vect.data()` this...

For posterity and in defense of how `rerun::Collection` eagerly borrows: The archetype types were not meant as a storage object in any of the SDKs and should ideally only ever...

@jaidevshriram yes, the linked PR is sadly Rust only and this is a very heavy-handed way of extending the viewer. This ticket is indeed all about making smaller adjustments from...

> multiple servers feeding one client We have that already! :) Two processes (no matter where) can feed the same Viewer by specifying the same appliation & recording id

General update: 0.23.0 will introduce some basic callbacks for the npm package, jupyter notebooks and self-hosted Rust builds of the Viewer.

It's great that you want to help, but if you're not familiar with how to update crates on a Rust project, then this is probably not a good first issue...

superseeded by newer upgrade by now

@Divelix try setting the software rasterizer (or better an actual driver) explicitly via an ICD file as described [here](https://rerun.io/docs/getting-started/troubleshooting#multiple-gpus) 0.17.0 will likely run into some runtime error once you try...

Our investigation (that's, @jprochazk 's ;-)) so far yielded no workaround on our side. The next step is to file this against Firefox and link to that issue from here