https://rerun.io logo
Join Discord
Powered by
  • WebViewer Callback Documentation
    k

    kind-kitchen-2583

    06/25/2025, 1:57 PM
    Is there any documentation on webviewer callback? I see that they are available here, but just see a single example. Is there a list of callbacks with function call signature, etc anywhere? Is the callback structure stable or should we treat it as a private API that changes without warning? https://rerun.io/docs/howto/integrations/embed-web#callbacks
    Copy code
    Callbacks 
    The Viewer API also allows registering callbacks for certain events.
    
    For example, here is how you would react to entities being selected in the Viewer:
    
    
    viewer.on("selection_change", (event) => {
      for (const item of event.items) {
        if (item.type === "entity") {
          console.log(item.entity_path);
        }
      }
    });
    0
    i
    • 2
    • 4
  • Connection to GRPC in kubernets cluster
    r

    rapid-france-81005

    06/25/2025, 12:16 PM
    Previous to rerun version 0.23, I did run the rerun webviewer (served over the cli) and forwarded the ws port with nginx. That way I could access the webviewer with an open websocket connection e.g. via: http://172.31.11.40/mycluster/rerun/?url=ws%3A%2F%2F172.31.11.40%2Fmycluster%2Fws
    Copy code
    location $BASE_PATH/rerun/ {
            proxy_pass http://localhost:9090/;
            proxy_http_version 1.1;
            proxy_set_header Upgrade $http_upgrade;
            proxy_set_header Connection "Upgrade";
            proxy_set_header Host $host;
            proxy_set_header X-Real-IP $remote_addr;
            proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
            proxy_set_header X-Forwarded-Proto $scheme;
        }
    
        location = $BASE_PATH/ws {
            return 301 $scheme://$http_host$BASE_PATH/ws/;
        }
    
        location $BASE_PATH/ws/ {
            proxy_pass http://localhost:9877/;
            proxy_http_version 1.1;
            proxy_set_header Upgrade $http_upgrade;
            proxy_set_header Connection "Upgrade";
            proxy_set_header Host $host;
            proxy_set_header X-Real-IP $remote_addr;
            proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
            proxy_set_header X-Forwarded-Proto $scheme;
        }
    As of 0.23, the connection switched to grpc and I can't make it work the same way. How can I get rerun running inside the pod? My goal is to connect a vscode webviewer which can execute python code which logs to a rerun web viewer instance. Using rrd files would work but does not push data (I don't want to refresh the view).
    0
    i
    • 2
    • 7
  • clear entities with columnar api
    b

    bumpy-rose-11730

    06/23/2025, 7:41 PM
    Hi rerun team, I have a question regarding the use of the columnar api when the data have some gaps. So, in my case, I have several archetypes that I need to log (boxes, 3dpoints line series...) but all of them might have gaps in the recording. Now, with the row api, I'd just call clear on that entity when it is not available. Unfortunately, due to performance issues, I need to use the columnar api. How can I clear the content in this case? I came up with 2 solutions: - use the columnar api and log NaN values on the components of the archetype, which basically makes it disappear until a valid value is logged again. - log all valid entries with the columnar api, and use the row api to clear content on that entity using clear/cleared when needed both are not great, the first one seems like a hack, and the second one is a bit cumbersome to implement if there are many gaps. Can you think of some other way to achieve this? what would you suggest? Thanks!
    0
    e
    • 2
    • 2
  • `connect_tcp` confusion
    a

    abundant-article-60550

    06/18/2025, 6:41 PM
    Here is my current scenario: I have a CLI with an option to enable and disable Rerun logging. I also want to disable Rerun logging if I am unable to connect to a valid viewer. In order to do this I am relying on calls to
    RecordingStream
    to "early out" when disabled. However, I'm finding that
    connect_tcp
    doesn't seem to require a connection to a valid viewer for
    connect_tcp(url).is_ok()
    to return
    ErrorCode::Ok
    . For instance, if I don't open a Rerun viewer and run my application, I get the following:
    Copy code
    CLI OPTION SET: Rerun viewer IP:PORT = 127.0.0.1:9876
    [2025-06-18T18:15:12Z WARN  re_sdk_comms::buffered_client] Failed to send message after 3 attempts: Failed to connect to Rerun server at 127.0.0.1:9876: Connection refused (os error 111)
    [2025-06-18T18:15:15Z WARN  re_sdk_comms::buffered_client] Dropping messages because tcp client has timed out.
    [2025-06-18T18:15:15Z WARN  re_sdk_comms::buffered_client] Dropping messages because tcp client has timed out.
    [2025-06-18T18:15:15Z WARN  re_sdk_comms::tcp_client] Tried to flush while TCP stream was still Pending. Data was possibly dropped.
    Connected to 127.0.0.1:9876
    Rerun logging enabled.
    0
    e
    i
    • 3
    • 8
  • Is there a way to ask a running rerun viewer to open new files from the cli or SDK?
    e

    echoing-agency-30693

    06/17/2025, 3:57 PM
    I have a largish-number og rrd files and would like to explore some of them based on some data I have available in a dataframe. As such, it would be convienent to have a function in the SDK or an option from the cli to not launch a new instance of the viewer, but ask the existing view to open a specific file
    0
    e
    • 2
    • 4
  • Rerun web viewer
    q

    quick-battery-49741

    06/16/2025, 12:13 PM
    Is there a way to easily use the browser to view rerun streams? To save wifi bandwidth I want to run the viewer locally on my robot and ideally use a web page to see the data rendered on the robot when I need to. Otherwise, is there an alternative way? basically I would like to be able to opt in when to stream the data out of the robot, but keep recording it in the background. Thank you
    0
    i
    • 2
    • 8
  • How to add --drop-at-latency using C++ API as spawn option?
    v

    victorious-crayon-58185

    06/13/2025, 6:09 AM
    Hello, I am observing a lot of latency often sporadically that gradually builds up. I already use a memory limit when spawning the viewer. But this does not help with latency building since this is during the logging time. I also noticed
    --drop-at-latency
    flag: https://rerun.io/docs/howto/visualization/limit-ram#dropatlatency in the documentation, but what is the way to incorporate this using C++ similar to memory limit spawn options?
    0
    e
    • 2
    • 2
  • How to save to file and visualize live in viewer at the same time
    d

    dry-crowd-37372

    06/05/2025, 10:02 PM
    When I include rr.save('adslfkj.rrd') I no longer see any data streamed live to the viewer. I can later open up the .rrd file but that's a bit annoying. Can I view the live data and write to file at the same time?
    0
    o
    • 2
    • 2
  • alloc::raw_vec::finish_grow unbounded heap leak C++
    r

    rich-forest-94678

    06/05/2025, 6:25 PM
    We are using the C++ sdk with c++20 std. Used valgrind massif to log the heap allocations. With rerun logging on I am seeing unbounded growth and with it off I am seeing a flat profile which is expected. We have rerun actively logging to the viewer over the network and are seeing data showing up. We are filling up 8 gigs of ram in less than 30 seconds when we are not profiling. I am using version 0.22.1 include(FetchContent) FetchContent_Declare( rerun_sdk URL https://github.com/rerun-io/rerun/releases/download/0.22.1/rerun_cpp_sdk.zip ) FetchContent_MakeAvailable(rerun_sdk) I am not sure if this issue is relevant but it has no reply https://github.com/mozilla/uniffi-rs/issues/2470 Any thoughts or suggestions are appreciated! https://cdn.discordapp.com/attachments/1380251130340315146/1380251131027914793/L824KW8tjiEcBAAAAAAAAAAAAAAAAAAAAAAAAAPgS2BYqflFizSdhKtDc9ZjHcd2T7LvrPRKPAgAAAAAAAAAAAAAAAAAAAAAAAACArU46UfM8Oc2zlcecJj4t998WkW4757ZpscSjAAAAAAAAAAAAAAAAAAAAAAAAAAAAp3TWsPYsx28KSIlHAQAAAAAAAAAAAAAAAAAAAAAAAAAAzqlxfGpmj81mc2jkQWyfgAAAABJRU5ErkJggg.png?ex=6853ad49&is=68525bc9&hm=cd889c185502cfa362d0d20906aced17b9269394de874d64d60bc6180a92a1b9& https://cdn.discordapp.com/attachments/1380251130340315146/1380251131816579142/L824KW8tjiEcBAAAAAAAAAAAAAAAAAAAAAAAAAPgS2BYqflFizSdhKtDc9ZjHcd2T7LvrPRKPAgAAAAAAAAAAAAAAAAAAAAAAAACArU46UfM8Oc2zlcecJj4t998WkW4757ZpscSjAAAAAAAAAAAAAAAAAAAAAAAAAAAAp3TWsPYsx28KSIlHAQAAAAAAAAAAAAAAAAAAAAAAAAAAzqlxfGpmj81mc2jkQWyfgAAAABJRU5ErkJggg.png?ex=6853ad49&is=68525bc9&hm=e16ec250f510150646db17784a3c1097574762af6a2e6255aae914fca4b51243& https://cdn.discordapp.com/attachments/1380251130340315146/1380251132651114627/5K2BmM6gAAAABJRU5ErkJggg.png?ex=6853ad49&is=68525bc9&hm=b83de04f18164540e1672bb3b2476cf94f5df705e4f87146c18875c0af59741a&
    0
    e
    a
    i
    • 4
    • 21
  • Point3D/Ellipse3D Transparency
    a

    abundant-article-60550

    06/05/2025, 12:22 AM
    is there a way to log semi-transparent Points3D or Elipsoids3D? I tried reducing the the alpha channel of the color, but that didn't work.
    Copy code
    c++
    rec.log(
        "sphere/point",
        rerun::Points3D({{sphere.center.x, sphere.center.y, sphere.center.z}})
            .with_radii({sphere.radius})
            .with_colors({rerun::Color(0, 0, 255, 5)})
    );
    Copy code
    c++
    rec.log(
        "sphere/ellipsoid",
        rerun::Ellipsoids3D::from_centers_and_radii(
            {{sphere.center.x, sphere.center.y, sphere.center.z}},
            {sphere.radius}
        )
            .with_colors({rerun::Color(0, 0, 255, 5)})
    );
    https://cdn.discordapp.com/attachments/1379978709775286343/1379978711935357009/image.png?ex=685d3b93&is=685bea13&hm=1bd34107e68da97a68af404c17a0e601eabcf51bb7708b11358f59333491b0e9& https://cdn.discordapp.com/attachments/1379978709775286343/1379978712698851358/image.png?ex=685d3b93&is=685bea13&hm=50537b579e12a51f37dce13fb61642f92e48eaa05871de1749538023006c5ec5&
    0
    e
    • 2
    • 2
  • Viewer slow when logging ~90 entities at 10 fps
    q

    quiet-restaurant-13188

    06/04/2025, 9:09 PM
    Hello, to be fair I am not sure whether this is an actual issue of just expected behavior but for my application (I am logging the current position of a humanoid robot) I need to log 33 Transform3D and 51 Scalars at minimum 10fps. What I am observing is that after around 2 minutes of logging the Viewer accumulate a certain delay showing old data instead of the most recently logged one. Is this expected? I would like to log rgb and depth and log at a higher frequency but that makes the issue much worse. I've attached an rrd to show the amount of entities I am logging. Rerun version: 0.23.2 Tested both on windows and linux https://cdn.discordapp.com/attachments/1379930097724690492/1379930098207031407/data.rrd?ex=685d0e4d&is=685bbccd&hm=3ac912c25f2fc84d595864be788aef2fe0c3c97812906d5e995dacd14526000d&
    0
    e
    l
    v
    • 4
    • 12
  • export rerun data to LeRobot v2.1 format dataset
    r

    rough-queen-83338

    06/04/2025, 7:57 AM
    Hello, I want to use ReRun for AI robotics, to record datasets compatible with LeRobot models training. I assume I need to record my data with rerun, then implement a data pipeline to transform the data into the LeRobot format. Here are details about the LeRobot data specs: https://docs.phospho.ai/learn/lerobot-dataset Can you give me some pointers on how to do that? Would this be easy or would I end up reimplementing everything from scratch?
    0
    o
    e
    • 3
    • 41
  • Adjust timeline size
    q

    quaint-truck-38174

    06/03/2025, 11:59 AM
    Hello! Is there a way to adjust the timeline size or hidden by default? I am rendering rerun visualizations in an iframe so the scene visualization ends up being really small https://cdn.discordapp.com/attachments/1379429260976062555/1379429261185912873/Screenshot_2025-06-03_at_2.57.32_PM.png?ex=6840351c&is=683ee39c&hm=45f7dee2cded1f814ad2c2c579593a7fbad60dcdbac9e76d825f5328bdf07174&
    0
    e
    i
    • 3
    • 6
  • Unable to view data
    a

    aloof-terabyte-11591

    06/02/2025, 7:58 PM
    Hey everyone I'm new to using rerun, and Im unable to get the examples to run! I'm trying to use the 3d viewer, so I ran the below code in a jupyter notebook. """Log some very simple points.""" import rerun as rr rr.init("rerun_example_points3d", spawn=True) rr.log("points", rr.Points3D([[0, 0, 0], [1, 1, 1]])) rerun looks like its receiving the data, but the actual data field is empty. I'm not seeing any points popullating. Am I missing something?
    0
    s
    • 2
    • 3
  • Timeline messing with non temporal items
    q

    quaint-truck-38174

    06/02/2025, 6:49 PM
    Hello! I am trying to create a timeline to show past and future of some elements in a 3D scene. For these temporal elements I set rr.set_time_nanos("time", t) when I log them. However, all the static elements that should always be visible seem to get messed up when I do that even though I don't specify a time for them. Is the a way to log items and make them always visible no matter what timestamp I am on? I have my temporal elements under "world/dynamic" and the others "world/static" . I saw there is an "overrides" tried doing something there like: rrb.Vertical( rrb.Spatial3DView( name="3D View", origin="/world", time_ranges=[ rr.VisibleTimeRange( "time", start=rr.TimeRangeBoundary.cursor_relative(seconds=-10.0), end=rr.TimeRangeBoundary.cursor_relative(seconds=5.0), ) ], overrides={ "/world/static": [rc.Visible(True)], # Always visible }, but I get an error. I am a bit confused reading the documentation so any help would be appreciated 🙂
    0
    e
    • 2
    • 9
  • DisconnectedSpace, View Spawning and Rust Blueprints
    n

    nutritious-ocean-76932

    05/30/2025, 1:57 PM
    I'm migrating my rust application to 0.23 from 0.20. I see that
    DisconnectedSpace
    was deprecated, then removed. One migration document suggest that a deliberately invalid
    Transform3d
    can serve the same purpose, another document suggests the view-spawning heuristic was removed. So I just want to confirm, are Blueprints the only way to spawn views now? I don't have complicated layout needs beyond spawning views, so I'd rather not go the blueprints route while it's still python only. 2nd question: I see some blueprints items exported with the rust-sdk. Is it possible to do Blueprint stuff with the rust-sdk? I'm willing to work with an underdocumented API to avoid python, but I don't want to spend a lot of time on it only to eventually learn that's currently not possible. Thanks!
    0
    i
    e
    • 3
    • 7
  • Data of different sizes are loaded in the local rerun and rerun.io/viewer
    c

    cold-rainbow-50460

    05/27/2025, 8:58 AM
    Hello. I noticed a strange thing. When I used
    rerun.io/viewer
    to view the results of data visualization using nuscenes, I noticed that the size of the loaded data was 103MB. When I saved it as an
    .rrd
    file, the size of this `.rrd
    file was **93MB**. However, when I git the code to the local computer and ran the program, I found that the size of the loaded data was **396MB**. When I saved the data as an
    .rrd` file, I found that the size of this
    .rrd
    file was 355MB. I'm very curious about why there is such a big difference in size between the data loaded in `rerun.io/viewer
    and the data I load locally with code, and there is also some difference in size between the data saved in
    .rrd` and the actual loaded data. I used the code implemented locally in python. Thanks! https://cdn.discordapp.com/attachments/1376846878133452801/1376847022241615882/data.jpg?ex=6836d036&is=68357eb6&hm=e1860eb9e02c092e68627fce6a78f37455c964878c1269c054a641440f49661e& https://cdn.discordapp.com/attachments/1376846878133452801/1376847022732087318/data_local.jpg?ex=6836d036&is=68357eb6&hm=348f94171b5e45e28853e5f8d28f04288294126f1f232b9cbc7b97d872577364& https://cdn.discordapp.com/attachments/1376846878133452801/1376847023071952896/data_rerun-io.jpg?ex=6836d036&is=68357eb6&hm=e7d11ac3be975f5c0810a8eae46f3adf7847efe2d312e1371eb0dbca2b970ab5&
    0
    e
    • 2
    • 4
  • View logs live and stream to file at the same time
    w

    wide-twilight-79510

    05/27/2025, 8:43 AM
    I want to view logs live and also stream them to a file at the same time. My goal is to have live visualisations, but also store log files in case something goes wrong. My experience is that usually, combining spawn and save results in silent failures and nothing being logged. Right now, I'm at this piece of code
    Copy code
    import rerun as rr
    
    # Create a recording stream (acts as a log "sink")
    rec_stream = rr.RecordingStream("my_app")
    
    # Optionally, also spawn the viewer to see logs live
    #rr.init("my_app", spawn=False)
    rr.spawn(connect=True, recording=rec_stream)
    
    # Also save the stream to a file
    rec_stream.save("my_log.rrd")
    
    # Now log your data
    for i in range(1000):
        rec_stream.log("world", rr.Points3D(positions=[[1*i, 2, 3]]))
    which does not log to the saved file. Changing the order of lines changes the behaviour: sometimes I get logs in the file, sometimes in the live viewer, but never in both.
    0
    e
    • 2
    • 2
  • reduce re_renderer wgpu friction
    a

    acoustic-country-63831

    05/27/2025, 8:27 AM
    I'm looking into using wgpu code through re_renderer, but the different caches and handles are forcing me to rewrite most of the code with a bunch of rerun specific boilerplate to fetch cache (I have to use hashable wrappers and their keys), it would come a long way if I could just reuse a wgpu struct and hide the cache logic. maybe a "simple" transformation function could be used: I suspect it would greatly reduce the mental overhead of renderers too.
    Copy code
    rs
    impl RenderPipelineDesc {
        pub fn try_from(
            ctx: &RenderContext,
            // The user doesn't need much rerun knowledge, plain wgpu is enough.
            desc: wgpu::RenderPipelineDescriptor<'_>,
        ) -> Result<RenderPipelineDesc, ()> {
            Ok(Self {
                label: desc.label.into(),
                // cache key is retrieved automatically
                pipeline_layout: ctx.gpu_resources.pipeline_layouts.get_or_create(
                    ctx,
                    // Cache again
                    PipelineLayoutDesc::try_from(ctx, desc.layout.ok_or(())?)?,
                ),
                // etc...
                ...
            })
        }
    }
    posting as discussion to gauge if that's worth an issue.
    0
    e
    • 2
    • 3
  • Render a compute shader result from GPU
    a

    acoustic-country-63831

    05/26/2025, 8:30 AM
    I'm looking into using rerun stack for visualizing real-time fluid simulation: https://github.com/Vrixyz/rerun/pull/2 ; I'd like to render my GPU data, can you help me through it ? How it works currently: - I use re_renderer - each frame, I run my own compute shader to advance the simulation - each frame, I read GPU data into CPU to write it back to GPU, through
    PointCloudBuilder
    . - I'd like this step to stay in GPU, likely a custom shader step to copy my own data into the data for rerun's point cloud ? Does that make sense ? I'm not too sure where to start, and I'm not sure how to sync the batch data 🤔.
    0
    e
    • 2
    • 4
  • Use the code method to make the default theme of rerun serve light
    c

    cold-rainbow-50460

    05/26/2025, 2:20 AM
    Hello, I really like the light theme mode you added in rerun 0.24.0a1. However, I want rerun to display the light theme directly through code. I'm using python-sdk. Besides, for the time being, I can only find the Rerun Python APIs of rerun rerun 0.23.2. I hope you can provide the rerun Python APIs of rerun 0.24.0a1. Looking forward to the official release of your version 0.24.1. Thanks!
    0
    e
    • 2
    • 7
  • Instantiate a blueprint object from an rbl file in python code
    s

    some-room-85362

    05/22/2025, 6:45 PM
    Hello, is there a way to create a blueprint object in code from an rbl file? I'd like to load a blueprint saved in an rbl file when spawning a rerun viewer in code. I'm using the python sdk.
    0
    e
    e
    • 3
    • 9
  • Timeline for the web viewer to stream data
    e

    enough-greece-70683

    05/22/2025, 3:18 PM
    Hey everyone, We've got a critical performance issue with the rerun web viewer. It's trying to load multiple large videos, resulting in the rrd file being larger than 10GB+ file being pulled into memory. This makes the viewer incredibly slow, buggy, and often causes crashes. Data streaming is absolutely essential here to make the viewer usable. My Questions: When can we expect this streaming feature to be ready? Can we prioritize this development? The current state is a major bottleneck. Any insights or experience with implementing video streaming in web viewers would be hugely appreciated! Thanks!
    0
    e
    • 2
    • 5
  • Mesh3D not visible / no visualizer selected by default
    o

    orange-house-80427

    05/22/2025, 1:14 PM
    Using
    0.22.0
    , I am logging lots and lots of
    rr.Mesh3D
    . When looking at the
    rr.SpatialView3D
    , none of the meshes are visible. If I click through them, and open up the entity in the tree, I see that no Visualizers are selected. If I manually add a Mesh3D visualizer, it shows up. (other options are rr.Arrow3D and rr.Points3D. Is there any way I can make it the default? I couldn't really figure out the overrides regarding visualizers yet. Note, that I have hundreds of meshes in a deeply nested tree, so manually clicking through it would be a pain.
    0
    e
    • 2
    • 2
  • Duration of GeoPoints and GeoLineStrings
    e

    enough-greece-70683

    05/21/2025, 6:31 PM
    How to set duration if I want some geopoints to stay longer than just the setTime, the way I am getting past this right now is setting a new entityPath for each drawing, even though it should be the same and it slows down the application quite a bit doing it this way.
    0
    e
    • 2
    • 2
  • Gradio_Rerun Performance Issues
    q

    quiet-restaurant-13188

    05/21/2025, 8:17 AM
    When I log a lot of data to rerun (a few hundred entities every 50ms), the visualizer embedded in gradio becomes laggy and some logged data is lost. If I visualize the data just with rerun (with no gradio) everything works perfectly.
    0
    o
    i
    • 3
    • 14
  • Timeline issues
    e

    enough-greece-70683

    05/20/2025, 8:13 PM
    I am having issues where my timeline is not showing up. Using 23.2 and linux, I am sending in two timestamps using
    Copy code
    dt = datetime.datetime.fromtimestamp(t_ns / 1e9, tz=datetime.timezone.utc)
                    rr.set_time("timestamp_gimbaled_camera", timestamp=dt)
    Copy code
    dt = datetime.datetime.fromtimestamp(int(ts_ns) / 1e9, tz=datetime.timezone.utc)
                    rr.set_time("timestamp_vehicle", timestamp=dt)
    https://cdn.discordapp.com/attachments/1374480038803800154/1374480039038550156/image.png?ex=682e33c9&is=682ce249&hm=e0812f7e54f033dbebbfc109b6f11074ec0ef93bbf3b439dce29dcfbb69bf9ee&
    0
    e
    i
    • 3
    • 6
  • Visualizing different timelines at once
    w

    wonderful-pizza-10806

    05/20/2025, 5:12 PM
    Hi! We have a use case where we want to log information from various sources that use different timestamps (ticks vs actual time). We have at least one source that contains values for both timeline. Right now we're using a manual asOf join to put everything onto the same timeline before logging, but can't rerun also take care of that for me? I feel like that would be one of the main reasons why someone would use multiple different timelines in Rerun, but I can't find anything about that documented.
    0
    e
    • 2
    • 3
  • General Entity Path Filters
    q

    quiet-restaurant-13188

    05/19/2025, 4:41 PM
    Hello, I want my view to display the entities (they are scalars)
    Copy code
    poses/action/components/x
    poses/proprio/components/x
    and right now I'm using the filters
    Copy code
    + poses/action/components/x/**
    + poses/proprio/components/x/**
    I would like to have a more general filter like
    + poses/**/x/**
    I know I could restructure the entity paths to end with the values I want to generalize on e.g.
    Copy code
    + poses/x/action
    + poses/x/proprio
    but that doesn't fit my visualizer logic.
    0
    e
    • 2
    • 5
  • Visual bounds 2D are all NaN
    h

    helpful-cricket-73183

    05/19/2025, 12:36 PM
    Hi All, I'm trying to visualize some simple 2D data in rerun, but nothing is showing up inside the main viewport, even though if I check the data in rerun, it all looks correct, except for one thing: All values in my "visual bounds 2D" are NaN. My guess is that these visual bounds are the reason why I'm not seeing anything, but I can't figure out why this is happening. Any idea what I'm doing wrong? A minimal example with which it happens is this
    Copy code
    auto rerun_recording_stream = std::make_unique<rerun::RecordingStream>("Navigation");
        rerun_recording_stream->connect_tcp(fmt::format("{}:{}", rerun_ip, rerun_port)).exit_on_failure();
        rerun_recording_stream->log("path", rerun::LineStrips2D{ rerun::LineStrip2D{ {
                                                { 1.55f, 3.03f },
                                                { 11.27f, 6.05f },
                                                { 8.07f, 14.41f },
                                                { 15.25f, 23.05f },
                                            } } });
    0
    e
    • 2
    • 5