[Python Environments] Environments show up under 'Environment manager' but not under 'Select interpreter'

[michael-sandler]

Type: Bug

Description

Uhhh... Using conda?

Steps to Reproduce

Expected Behavior

Actual Behavior

Environment Information

Extension Version: 1.36.0
Python Extension Version: 2026.4.0

Registered Environment Managers (6):
  - ms-python.python:system (Global)
  - ms-python.python:venv (venv)
  - ms-python.python:pyenv (PyEnv)
  - ms-python.python:pipenv (Pipenv)
  - ms-python.python:poetry (Poetry)
  - ms-python.python:conda (Conda)

Total Available Environments: 17
Environment Details:
  1. Python 3.9.6.final.0 (3.9.6.final.0) - /usr/bin/python3
  2. Python 3.12.4.final.0 (homebrew) (3.12.4.final.0) - /opt/homebrew/bin/python3
  3. naparitest3 (no-python) (no-python) - /opt/miniconda3/envs/naparitest3
  4. manim (3.13.7) (3.13.7) - /opt/miniconda3/envs/manim
  5. darpa (3.13.5) (3.13.5) - /opt/miniconda3/envs/darpa
  6. napari-mm3 (3.12.7) (3.12.7) - /opt/miniconda3/envs/napari-mm3
  7. napari7 (3.12.3) (3.12.3) - /opt/miniconda3/envs/napari7
  8. base (3.12.2) (3.12.2) - /opt/miniconda3
  9. omni (3.12.2) (3.12.2) - /opt/miniconda3/envs/omni
  10. .conda (3.11.10) (3.11.10) - /Users/michaelsandler/repos/napari-mm3/.conda
  ... and 7 more environments

Python Projects (1):
  1. /Users/michaelsandler/Documents/experiments/20260715-dyes-and-timelapse/mm-analysis
     Environment: napari-mm3 (3.12.7)

Extension Settings:
  Default Environment Manager: ms-python.python:conda
  Default Package Manager: ms-python.python:conda
Auto-activation is "command". Activation based on first 'py-env.terminal.autoActivationType' setting which is 'command' and 'python.terminal.activateEnvironment' if the first is undefined which is 'true'.

Extension version: 1.36.0
VS Code version: Code 1.126.0 (Universal) (7e7950df89d055b5a378379db9ee14290772148a, 2026-06-23T01:46:13Z)
OS version: Darwin arm64 22.2.0
Modes:

System Info

Item	Value
CPUs	Apple M1 Ultra (20 x 2400)
GPU Status	2d_canvas: enabled GPU0: VENDOR= 0x106b [Google Inc. (Apple)], DEVICE=0x0000 [ANGLE (Apple, ANGLE Metal Renderer: Apple M1 Ultra, Version 13.1 (Build 22C65))], DRIVER_VENDOR=Apple, DRIVER_VERSION=13.1 ACTIVE Machine model name: Mac Machine model version: 13.2 direct_rendering_display_compositor: disabled_off_ok gpu_compositing: enabled multiple_raster_threads: enabled_on opengl: enabled_on rasterization: enabled raw_draw: disabled_off_ok skia_graphite: enabled_on trees_in_viz: disabled_off video_decode: enabled video_encode: enabled webgl: enabled webgpu: enabled webnn: disabled_off
Load (avg)	3, 3, 3
Memory (System)	128.00GB (0.58GB free)
Process Argv	--crash-reporter-id 1f075bc2-56e3-4f11-9abb-6cdec8a7c1c2
Screen Reader	no
VM	0%

A/B Experiments

vsliv368cf:30146710
vswsl492cf:30256860
binariesv615:30325510
nativeloc1:31344060
dwcopilot:31170013
dwoutputs:31242946
copilot_t_ci:31333650
e5gg6876:31282496
pythonrdcb7:31342333
6518g693:31463988
aj953862:31281341
envsactivate1:31551504
cloudbuttont:31379625
3efgi100_wstrepl:31403338
ddidt:31399633
ec5jj548:31422691
cp_cls_t_966_ss:31526232
c3h7c220:31478652
cp_cls_c_1081:31454833
conptydll_true:31498968
e9c30283:31461165
test_treatment2:31471001
46204921:31447328
capico_a_098a3278:31555650
idci7584:31464702
e3e4d672:31494082
ei9d7968:31496641
chat:31457767
8hig5102:31480529
ah:31461780
89g7j272:31518289
7e187181:31503455
i2gc6536:31499202
52612955:31516516
ghj88844:31499326
23c7c724:31520045
ddid_c:31478207
hmra_i5g22:31518061
7df3h592:31512476
cp_cls_t_1082:31535311
logging_enabled_new:31498466
db5d2638:31499441
jb_cp_cls_t_632:31543129
748c7209:31512887
32d76977:31512328
ha629193:31508444
cdk-lw-on:31524445
jh5f2457_c:31540921
41836166:31537719
jbcp_cls_pctr_t:31531130
61138546:31518536
cp_intellij_t_nes:31548657
gf14b233:31526830
nes-curr-leftover-1500:31538366
h561g132:31533256
session_8:31532895
ihg5j128:31534457
7g2b5551:31542111
nes-ex-cf-euporie-03:31542868
fd5dc207:31551400
0dg73464:31554319
treatment-23-1:31555779
unuse_dynamic_mcp:31555281

[TonysirShen]

I also reproduce this issue on Windows 10 with Miniconda3, and I have some additional diagnostic information that may help narrow down where the problem occurs.

Type: Bug

Description:

All Conda environments are correctly detected by the Python Environment Manager, but most of them are missing from both the Python interpreter selector and the Jupyter kernel selector.
The environments themselves appear to be discovered correctly:

• Python Environment Manager shows all Conda environments.
• pet.exe find detects all Conda environments correctly.
• Each environment can be selected using Set as Project Environment from Environment Manager.
• However, Python: Select Interpreter only shows the system Python and the currently selected project environment.
• In a Jupyter notebook, Select Kernel → Select Another Kernel... → Python Environments shows the same restricted list.

For example, PET correctly detects:
Conda:

• path/to/miniconda3/python.exe
• path/to/miniconda3/envs/DEEPLABCUT/python.exe
• path/to/miniconda3/envs/conda/_global/python.exe
• path/to/miniconda3/envs/udmt/python.exe
• path/to/miniconda3/envs/minian/python.exe

#Managers:

Conda : 1

#Environments (7):

Conda : 6
Uv : 1
All of these Conda environments are also visible in Python Environment Manager.

Steps to Reproduce:

1. Install/use Miniconda with multiple Conda environments.
2. Open VS Code.
3. Open Python Environment Manager.
4. Confirm that multiple Conda environments are detected and listed.
5. Confirm that these environments can selected in Python Environment Manager
6. Open the Command Palette and run: Python: Select Interpreter
7. Observe that the full list of environments shown in Environment Manager is not available.
8. Open a Jupyter notebook.
9. Jupyter Select Kernel Conda environments are missing
10. Run:pet.exe find that PET itself correctly finds the missing Conda environments.

Expected Behavior

All Python environments discovered by Python Environment Manager/PET should be available in:
Python: Select Interpreter and in Jupyter Select Kernel

Actual Behavior:

Python Environment Manager correctly displays all Conda environments, but the interpreter and kernel selectors display only:
System Python and The environment currently set as the Project Environment
Other Conda environments detected by Environment Manager and PET are missing.
If one of the missing environments is selected using Environment Manager,that environment then becomes the current environment and is available for use.

Environment Information

System info:
OS: Windows 10

VS Code: 1.132.0

Conda:
conda 26.1.1
VS Code extensions:
ms-python.debugpy@2026.6.0
ms-python.python@2026.4.0
ms-python.vscode-pylance@2026.3.1
ms-python.vscode-python-envs@1.36.0

ms-toolsai.jupyter@2025.9.1
ms-toolsai.jupyter-keymap@1.1.2
ms-toolsai.jupyter-renderers@1.3.0
ms-toolsai.vscode-jupyter-cell-tags@0.1.9
ms-toolsai.vscode-jupyter-slideshow@0.1.6
Conda installation:
path/to/miniconda3
Example environment detected by PET but missing from the selector:
path/to/miniconda3/envs/named_env/python.exe

A/B EXPERIMENTS:

vsliv368cf:30146710
binariesv615:30325510
nativeloc1:31344060
dwcopilot:31170013
dwoutputs:31242946
copilot_t_ci:31333650
e5gg6876:31282496
pythonrdcb7:31342333
6518g693:31463988
aj953862:31281341
4f60g487:31327383
envsactivate1:31551504
cloudbuttont:31379625
3efgi100_wstrepl:31403338
839jf696:31457053
ec5jj548:31422691
cp_cls_t_966_ss:31526232
4je02754:31466945
c3h7c220:31478652
cp_cls_t_1081:31454832
conptydll_true:31498968
e9c30283:31461165
test_treatment2:31471001
46204921:31447328
ei9d7968:31496641
nes-extended-on:31455476
chat:31457767
8hig5102:31480529
89g7j272:31518289
i2gc6536:31499202
52612955:31516516
ghj88844:31499326
h08i8180:31475369
ddid_c:31478207
hmra_i5g22:31518061
7df3h592:31512476
cp_cls_t_1082:31535311
logging_enabled_new:31498466
db5d2638:31499441
jb_cp_cls_t_632:31543129
56dj4588:31512888
32d76977:31512328
ha629193:31508444
cdk-lw-on:31524445
a1ije391_t:31540920
hgf2d445:31529595
jbcp_cls_pctr_t:31531130
61138546:31518536
cp_intellij_t_nes:31548657
d7b18187:31526828
nes-curr-leftover-1500:31538366
ahp-both-windows:31556933
sessions_3:31532893
ihg5j128:31534457
f412h606:31551416
9gg7f176:31542112
5fe6f778:31554331
allow-none:31555437
6i51c363:31551401
c7c27ce7:31554789
1h923230:31564177
1532g621_copy:31554320
treatment-23-1:31555779
unuse_dynamic_mcp:31555281
switchtoauto:31559321
vrbsty_fls:31561059
session-archive:31558132
18bic794_t_pref:31563076
5b8j3302:31564172

[StellaHuang95]

TonysirShen thanks for reporting the issue. Could you open the Python output (not the python environments output), and check if you see something like Environment discovery is managed by the Python Environments extension?

Also, if you set "python.useEnvironmentsExtension": true, reload and try again, does the issue repro? This will help us find the root cause. Thanks!

[TonysirShen]

Thanks for the follow-up.I checked both points.

The Python output channel contains:

Environment discovery is managed by the Python Environments extension (ms-python.vscode-python-envs). Check the "Python Environments" output channel for environment-specific logs.

I also confirmed that Python: Use Environments Extension is already enabled in the Settings UI.

To make sure the experiment setting was not affecting the result, I explicitly added:

"python.useEnvironmentsExtension": true

to my User settings.json, then ran Developer: Reload Window.

After reloading, the issue still reproduces:

• Python Environment Manager shows all Conda environments.
• Set as Project Environment works.
• Python: Select Interpreter only shows the system Python and the current project environment.
• Select Kernel → Select Another Kernel... → Python Environments also only shows the system Python and the current project environment.

The Python Environments output also shows that PET successfully discovers the missing Conda environments,.

P.S.
I also noticed this warning in the Python Environments output:

Failed to execute conda info Os { code: 5, kind: PermissionDenied, message: "拒绝访问。" }

There are also several warnings such as:

Unable to find conda package Python in "C:\Users\PC\miniconda3\envs\minian", trying slower approach

However, PET continues after these warnings and successfully discovers the environments, so I am not sure whether these warnings are related to the picker issue.

I will also test explicitly setting:

"python.useEnvironmentsExtension": true

then reload VS Code and report whether the issue still reproduces.

[StellaHuang95]

TonysirShen Thanks for the additional diagnostics. This rules out the issue with legacy discovery path. I now suspect the issue is in the Python extension’s compatibility adapter, which translates environments from the Python Environments API into the older format used by Python and Jupyter.

If the adapter encounters a Conda environment that does not currently have Python installed, its version conversion can fail and prevent the remaining environments in that update batch from reaching the interpreter picker. Could you check whether Environment Manager shows any Conda environment labeled (no-python), or whether the Python Environments output contains Found a No-Python conda environment, or you have an environment that doesn't have Python installed ?

[TonysirShen]

Thanks — I tested this hypothesis, and I can confirm that the no-Python Conda environment is the causal trigger in my case.

I performed a reversible A/B test without modifying the contents of the environment:

1. With the no-Python Conda environment present under C:\Users\{name}\miniconda3\envs:

   • Python Environment Manager still detects the Conda environments.
   • Python: Select Interpreter only shows the system Python and the current project environment.
   • Notebook → Select Kernel → Select Another Kernel... → Python Environments shows the same restricted list.

2. I moved thethe no-Python Conda environment directory outside of the Miniconda envs directory, then reloaded VS Code:

   • Python: Select Interpreter immediately worked normally and showed the other Conda environments.
   • The Jupyter kernel picker also worked normally and showed the available Python environments.

3. I then moved the same unchanged lp directory back to:C:\Users\{name}\miniconda3\envs\

and reloaded VS Code again:

• The interpreter picker issue immediately reappeared.
• The Jupyter kernel picker issue also reappeared.

So the behavior is reproducible in both directions:

This strongly supports the compatibility-adapter failure mode you described.

I would also like to suggest that the adapter should ideally tolerate Conda environments without Python installed. A no-Python Conda environment is a valid state, so encountering one should not prevent other valid Python environments in the same update batch from reaching the interpreter or Jupyter kernel picker.

Thanks for you support!

[StellaHuang95]

TonysirShen Thanks, that confirms the root cause. I opened a fix for it. microsoft/vscode-python#26112