Control where virtualenv is created

[caliph007]

For deploying in production it would be really helpful to have tighter control on the location where the virtual env gets created.

Just fyi a transscript from my last deploy:

[server.com] run: export PIPENV_VENV_IN_PROJECT=true
[server.com] run: pipenv --three
[server.com] out: Creating a virtualenv for this project…
[server.com] out: Using /usr/bin/python3 to create virtualenv…
[server.com] out: ⠋Running virtualenv with interpreter /usr/bin/python3
[server.com] out: Using base prefix '/usr'
[server.com] out: New python executable in /home/ubuntu/.local/share/virtualenvs/application-qO5bWt8a/bin/python3
[server.com] out: Also creating executable in /home/ubuntu/.local/share/virtualenvs/application-qO5bWt8a/bin/python
[server.com] out: Installing setuptools, pkg_resources, pip, wheel...done.
[server.com] out:
[server.com] out: Virtualenv location: /home/ubuntu/.local/share/virtualenvs/application-qO5bWt8a

Why is pipenv disrespecting the environment variable PIPENV_VENV_IN_PROJECT ?

[techalchemy]

@caliph007 if you are using docker you will need to use proper docker commands to build your environment, as export will not accomplish this. In addition you will also need to pass WORKDIR. See this link for more information or our example Dockerfile for a slightly different example. Beyond that, if you are still having issues we will need your dockerfile