bug+workaround: "Error: No version of module jupyter_rfb is registered" in Jupyter / VSCode environment

[edoardogiacomello]

Hello,
I'd like to report an issue that I'm having since a couple of months while trying to use NDV on Jupyter through VSCode by following the provided setup approach.

Steps to reproduce

# Create a new conda enviroment
conda create -p $PWD/ndvtest python=3.10 ipykernel
# Activate env
conda activate $PWD/ndvtest
# Install ndv
pip install ndv[jup]

Then, in a notebook:

import ndv
import numpy as np

ndv.imshow(np.random.rand(100, 100, 3).astype(np.float32))

This provides two errors, one in the JS frame of the ndv viewport:

Failed to load model class 'RemoteFrameBufferModel' from module 'jupyter_rfb'
Error: No version of module jupyter_rfb is registered
    at WidgetManager.loadClass (https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/localscratch/vscode-server/.vscode-server/extensions/ms-toolsai.jupyter-renderers-1.1.0/out/node_modules/%40vscode/jupyter-ipywidgets8/dist/ipywidgets.js:2:2420023)
    at WidgetManager.loadClass (https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/localscratch/vscode-server/.vscode-server/extensions/ms-toolsai.jupyter-renderers-1.1.0/out/node_modules/%40vscode/jupyter-ipywidgets8/dist/ipywidgets.js:2:2803837)
    at WidgetManager.loadModelClass (https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/localscratch/vscode-server/.vscode-server/extensions/ms-toolsai.jupyter-renderers-1.1.0/out/node_modules/%40vscode/jupyter-ipywidgets8/dist/ipywidgets.js:2:2416808)
    at WidgetManager._make_model (https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/localscratch/vscode-server/.vscode-server/extensions/ms-toolsai.jupyter-renderers-1.1.0/out/node_modules/%40vscode/jupyter-ipywidgets8/dist/ipywidgets.js:2:2414769)
    at WidgetManager.new_model (https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/localscratch/vscode-server/.vscode-server/extensions/ms-toolsai.jupyter-renderers-1.1.0/out/node_modules/%40vscode/jupyter-ipywidgets8/dist/ipywidgets.js:2:2412488)
    at WidgetManager.handle_comm_open (https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/localscratch/vscode-server/.vscode-server/extensions/ms-toolsai.jupyter-renderers-1.1.0/out/node_modules/%40vscode/jupyter-ipywidgets8/dist/ipywidgets.js:2:2411259)
    at https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/localscratch/vscode-server/.vscode-server/extensions/ms-toolsai.jupyter-renderers-1.1.0/out/node_modules/%40vscode/jupyter-ipywidgets8/dist/ipywidgets.js:2:2802850
    at t._handleCommOpen (https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/localscratch/vscode-server/.vscode-server/extensions/ms-toolsai.jupyter-2025.1.0-linux-x64/dist/webviews/webview-side/ipywidgetsKernel/ipywidgetsKernel.js:3:119358)
    at async t._handleMessage (https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/localscratch/vscode-server/.vscode-server/extensions/ms-toolsai.jupyter-2025.1.0-linux-x64/dist/webviews/webview-side/ipywidgetsKernel/ipywidgetsKernel.js:3:121233)

Concurrently, VSCode alerts me with this:

Workaround

The issue seems to be related to newer version of jupyter-rfb that gets installed with ndv:
jupyter-rfb 0.5.1

If I just downgrade jupyter-rfb to jupyter_rfb==0.4.4, i get some warnings...

WARNING:wgpu:Unable to find extension: VK_EXT_swapchain_colorspace
WARNING:wgpu:No windowing system present. Using surfaceless platform
libEGL warning: DRI2: failed to create dri screen
libEGL warning: failed to open [/dev/dri/card1](https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/dev/dri/card1): Permission denied

libEGL warning: failed to open [/dev/dri/card0](https://vscode-remote+ssh-002dremote-002bvdi.vscode-resource.vscode-cdn.net/dev/dri/card0): Permission denied

WARNING:wgpu:No config found!
WARNING:wgpu:No config found!

but it ultimately works.

Interestingly, if I pip install --upgrade jupyter_rfb back to 0.5.1 then ndv keeps working and does not break anymore.

[gselzer]

Been debugging a little here. I was able to reproduce your error using:

# Create a new conda enviroment
conda create -p $PWD/ndvtest python=3.10 ipykernel
# Activate env
conda activate $PWD/ndvtest
# Install ndv
pip install ndv[jup]

however I was also able to reproduce the error using the conda installation instructions:

mamba create -n ndvtest -y -c conda-forge python==3.10
mamba activate ndvtest
mamba install -y -c conda-forge ndv vispy pyopengl jupyter jupyter-rfb pyglfw ipywidgets

notably, the above environment does not work in jupyter notebook either.

A quick fix that does make things work in jupyter notebook is the following, which indirectly also downgrades jupyterlab to 4.3.6

mamba install -y -c conda-forge "notebook==7.3.3"

But, this still doesn't work in VSCode.

Interestingly, if I pip install --upgrade jupyter_rfb back to 0.5.1 then ndv keeps working and does not break anymore.

Huh, this part doesn't work for me. If you have things working with the workaround, can you post your environment (i.e. pip list?)

Also see vispy/jupyter_rfb#115

[edoardogiacomello]

If you have things working with the workaround, can you post your environment (i.e. pip list?)

Package                   Version
------------------------- --------------
annotated-types           0.7.0
anyio                     4.9.0
argon2-cffi               23.1.0
argon2-cffi-bindings      21.2.0
arrow                     1.3.0
asttokens                 3.0.0
async-lru                 2.0.5
attrs                     25.3.0
babel                     2.17.0
beautifulsoup4            4.12.3
bleach                    6.2.0
certifi                   2025.1.31
cffi                      1.17.1
charset-normalizer        3.4.1
cmap                      0.6.0
comm                      0.2.1
debugpy                   1.8.11
decorator                 5.1.1
defusedxml                0.7.1
exceptiongroup            1.2.0
executing                 0.8.3
fastjsonschema            2.21.1
filelock                  3.17.0
fqdn                      1.5.1
freetype-py               2.5.1
gdown                     5.2.0
glfw                      2.8.0
h11                       0.14.0
hsluv                     5.0.4
httpcore                  1.0.7
httpx                     0.28.1
idna                      3.10
imageio                   2.37.0
ipykernel                 6.29.5
ipython                   8.30.0
ipywidgets                8.1.5
isoduration               20.11.0
jedi                      0.19.2
Jinja2                    3.1.6
json5                     0.12.0
jsonpointer               3.0.0
jsonschema                4.23.0
jsonschema-specifications 2024.10.1
jupyter                   1.1.1
jupyter_client            8.6.3
jupyter-console           6.6.3
jupyter_core              5.7.2
jupyter-events            0.12.0
jupyter-lsp               2.2.5
jupyter-rfb               0.5.1
jupyter_server            2.15.0
jupyter_server_terminals  0.5.3
jupyterlab                4.3.6
jupyterlab_pygments       0.3.0
jupyterlab_server         2.27.3
jupyterlab_widgets        3.0.13
MarkupSafe                3.0.2
matplotlib-inline         0.1.6
mistune                   3.1.3
nbclient                  0.10.2
nbconvert                 7.16.6
nbformat                  5.10.4
ndv                       0.3.1
nest-asyncio              1.6.0
notebook                  7.3.3
notebook_shim             0.2.4
numpy                     2.2.4
overrides                 7.7.0
packaging                 24.2
pandocfilters             1.5.1
parso                     0.8.4
pexpect                   4.8.0
pillow                    11.1.0
pip                       25.0
platformdirs              3.10.0
prometheus_client         0.21.1
prompt-toolkit            3.0.43
psutil                    5.9.0
psygnal                   0.12.0
ptyprocess                0.7.0
pure-eval                 0.2.2
pycparser                 2.22
pydantic                  2.11.3
pydantic_core             2.33.1
pygfx                     0.9.0
Pygments                  2.15.1
pylinalg                  0.6.7
python-dateutil           2.9.0.post0
python-json-logger        3.3.0
PyYAML                    6.0.2
pyzmq                     26.2.0
referencing               0.36.2
rendercanvas              2.0.3
requests                  2.32.3
rfc3339-validator         0.1.4
rfc3986-validator         0.1.1
rpds-py                   0.24.0
Send2Trash                1.8.3
setuptools                75.8.0
six                       1.16.0
sniffio                   1.3.1
soupsieve                 2.6
stack-data                0.2.0
terminado                 0.18.1
tifffile                  2025.3.30
tinycss2                  1.4.0
tomli                     2.2.1
tornado                   6.4.2
traitlets                 5.14.3
types-python-dateutil     2.9.0.20241206
typing_extensions         4.12.2
typing-inspection         0.4.0
uharfbuzz                 0.49.0
uri-template              1.3.0
urllib3                   2.3.0
wcwidth                   0.2.5
webcolors                 24.11.1
webencodings              0.5.1
websocket-client          1.8.0
wgpu                      0.21.1
wheel                     0.45.1
widgetsnbextension        4.0.13

Indeed the versions of notebook and jupyterlab are the ones that you mentioned, so it might be the reason why it's working even with the latest version of jupyter-rfb

[tlambert03]

yeah, I think the state of vscode support in jupyter-rfb and ipywidgets is a bit spotty/variable. Even trying the example from the jupyter-rfb example with pygfx here it doesn't work for me.

uv venv
uv pip install ipykernel pygfx jupyter-rfb

and then open a notebook in vscode and enter:

import pygfx as gfx
import pylinalg as la

cube = gfx.Mesh(
    gfx.box_geometry(200, 200, 200),
    gfx.MeshPhongMaterial(color="#336699"),
)


def animate():
    rot = la.quat_from_euler((0.005, 0.01), order="XY")
    cube.local.rotation = la.quat_mul(rot, cube.local.rotation)


disp = gfx.Display()
disp.before_render = animate
disp.stats = True
disp.show(cube)

I see this (no output)

but the hello world juptyer-rfb example works:

I was able to get it to work with vispy:

uv pip install ipykernel jupyter-rfb vispy pyopengl glfw

running the vispy example here: https://jupyter-rfb.readthedocs.io/en/stable/examples/vispy1.html

... and that also works with ndv:

uv venv
uv pip install ipykernel 'ndv[vispy,jupyter]'

summary: I guess i'd say use the vispy backend instead of pygfx until we figure that out (ndv[vispy,jupyter])

[tlambert03]

we could update our jup convenience extra to use that for now

[gselzer]

uv venv
uv pip install ipykernel 'ndv[vispy,jupyter]'

Interestingly enough, that does not work for me:

 ndv  uv venv
Using CPython 3.12.4 interpreter at: C:\Users\gjselzer\AppData\Local\Programs\Python\Python312\python.exe
Creating virtual environment at: .venv
Activate with: .venv\Scripts\activate
 ndv  .\.venv\Scripts\activate
(ndv)  ndv  uv pip install ipykernel 'ndv[vispy, jupyter]'
Resolved 113 packages in 1.88s
Prepared 20 packages in 10.77s
Installed 113 packages in 9.76s
 + annotated-types==0.7.0
 + anyio==4.9.0
 + argon2-cffi==23.1.0
 + argon2-cffi-bindings==21.2.0
 + arrow==1.3.0
 + asttokens==3.0.0
 + async-lru==2.0.5
 + attrs==25.3.0
 + babel==2.17.0
 + beautifulsoup4==4.13.3
 + bleach==6.2.0
 + certifi==2025.1.31
 + cffi==1.17.1
 + charset-normalizer==3.4.1
 + cmap==0.6.0
 + colorama==0.4.6
 + comm==0.2.2
 + debugpy==1.8.13
 + decorator==5.2.1
 + defusedxml==0.7.1
 + executing==2.2.0
 + fastjsonschema==2.21.1
 + fqdn==1.5.1
 + freetype-py==2.5.1
 + glfw==2.8.0
 + h11==0.14.0
 + hsluv==5.0.4
 + httpcore==1.0.7
 + httpx==0.28.1
 + idna==3.10
 + ipykernel==6.29.5
 + ipython==9.1.0
 + ipython-pygments-lexers==1.1.1
 + ipywidgets==8.1.5
 + isoduration==20.11.0
 + jedi==0.19.2
 + jinja2==3.1.6
 + json5==0.12.0
 + jsonpointer==3.0.0
 + jsonschema==4.23.0
 + jsonschema-specifications==2024.10.1
 + jupyter==1.1.1
 + jupyter-client==8.6.3
 + jupyter-console==6.6.3
 + jupyter-core==5.7.2
 + jupyter-events==0.12.0
 + jupyter-lsp==2.2.5
 + jupyter-rfb==0.5.1
 + jupyter-server==2.15.0
 + jupyter-server-terminals==0.5.3
 + jupyterlab==4.4.0
 + jupyterlab-pygments==0.3.0
 + jupyterlab-server==2.27.3
 + jupyterlab-widgets==3.0.13
 + kiwisolver==1.4.8
 + markupsafe==3.0.2
 + matplotlib-inline==0.1.7
 + mistune==3.1.3
 + nbclient==0.10.2
 + nbconvert==7.16.6
 + nbformat==5.10.4
 + ndv==0.3.1
 + nest-asyncio==1.6.0
 + notebook==7.4.0
 + notebook-shim==0.2.4
 + numpy==2.2.4
 + overrides==7.7.0
 + packaging==24.2
 + pandocfilters==1.5.1
 + parso==0.8.4
 + platformdirs==4.3.7
 + prometheus-client==0.21.1
 + prompt-toolkit==3.0.50
 + psutil==7.0.0
 + psygnal==0.12.0
 + pure-eval==0.2.3
 + pycparser==2.22
 + pydantic==2.11.3
 + pydantic-core==2.33.1
 + pygments==2.19.1
 + pyopengl==3.1.9
 + python-dateutil==2.9.0.post0
 + python-json-logger==3.3.0
 + pywin32==310
 + pywinpty==2.0.15
 + pyyaml==6.0.2
 + pyzmq==26.4.0
 + referencing==0.36.2
 + requests==2.32.3
 + rfc3339-validator==0.1.4
 + rfc3986-validator==0.1.1
 + rpds-py==0.24.0
 + send2trash==1.8.3
 + setuptools==78.1.0
 + six==1.17.0
 + sniffio==1.3.1
 + soupsieve==2.6
 + stack-data==0.6.3
 + terminado==0.18.1
 + tinycss2==1.4.0
 + tornado==6.4.2
 + traitlets==5.14.3
 + types-python-dateutil==2.9.0.20241206
 + typing-extensions==4.13.1
 + typing-inspection==0.4.0
 + uri-template==1.3.0
 + urllib3==2.3.0
 + vispy==0.14.3
 + wcwidth==0.2.13
 + webcolors==24.11.1
 + webencodings==0.5.1
 + websocket-client==1.8.0
 + widgetsnbextension==4.0.13

[tlambert03]

Well shoot. I kinda do think these are all issues for Jupyter-rfb rather than ndv (though obviously we’ve got incentive to figure them out)

[gselzer]

Yup, totally agree. For what it's worth, it does work in jupyter notebook - I spent more time than I probably should have trying to figure out what was different (figured it likely had to do with a different set of widget sources), but I just know too little about requireJS/Jupyter in general

[tlambert03]

This is one thing that using anywidget might help with a bit (over ipywidgets at least)… but it would also require us to reimplement Jupyter rfb (which currently relies on ipywidgets for the model/view sync)

[CatEek]

Unfortunately neither capping the jupyter-rfb version like was proposed, nor playing around with glfb version didn't help. Keep getting in VSCode
GLFWError: (65550) b'X11: The DISPLAY environment variable is missing' GLFWError: (65537) b'The GLFW library is not initialized' warnings.warn(message, GLFWError)

[tlambert03]

Thanks Igor, maybe we can look at this this week :)