Recently the debugger has started to activate conda environments which is NOT what I want

@lovettchris commented on Wed Feb 06 2019

Environment data

• PTVSD version: c:\Users\clovett.vscode\extensions\ms-python.python-2019.1.0\pythonFiles\ptvsd_launcher.py
• OS and version: Win 10 version 1809, build 1773.292
• Python version (& distribution if applicable, e.g. Anaconda): Anaconda python 3.6
• Using VS Code or Visual Studio: VS Code

Actual behavior

Debugging is now doing this, which is NOT what I want, how did this start and how do I disable it? This is what I see in my Terminal output window when trying to debug one of my python scripts:

                   > cd 'd:\temp\gallery\test'; ${env:PYTHONIOENCODING}='UTF-8'; ${env:PYTHONUNBUFFERED}='1'; & 'd:\Continuum\Anaconda3\envs\tensorflow\python.exe' 'c:\Users\clovett\.vscode\extensions\ms-python.python-2019.1.0\pythonFiles\ptvsd_launcher.py' '--default' '--client' '--host' 'localhost' '--port' '51001' 'd:\git\ELL\ELL-training\Gallery\test_models.py' '--path' 'd:\temp\gallery' '--target' 'pi3' '--dataset' '\\aifs\datasets\public\ILSVRC2012 (Imagenet)\ilsvrc.json' '--skip_md' '--categories' '\\aifs\datasets\public\ILSVRC2012 (Imagenet)\ilsvrc.txt' '--max_files' '10' '--logfile' 'test.log' '--skip_md'

PS D:\temp\gallery\test> d:/Continuum/Anaconda3/Scripts/activate PS D:\temp\gallery\test> conda activate tensorflow

Expected behavior

Do not attempt to auto activate a conda environment

Steps to reproduce:

Not sure.

---

@karthiknadig commented on Wed Feb 06 2019

This is the extension activating it. Moving the issue over to extension repo.

---

@lovettchris commented on Wed Feb 06 2019

What is odd is that when I use the debug console to look at os.get_exec_path() I see this path has been added to the console environment in which I launched VS code. I have no idea how or why it was added:

os.get_exec_path()
...
00:'D:\\Continuum\\Anaconda3\\envs\\tensorflow\\Library\\bin'

[lovettchris]

Ok, turning off “python.terminal.activateEnvironment” is what I needed thanks. You can close this now.

[DonJayamanne]

Thats too specific to conda. We need to suppport pipenv, pyenv, virtualenv, etc.

Because I often run 2 or 3 different environments throughout my work day…

All i can suggest is setting the interpreter in vsc using Python: Select Interpreter command.