#Jetson Nano - pip install viam-sdk has no version

1 messages · Page 1 of 1 (latest)

lucid bridge
#

Running newest jetpack for Jetson Nano 4GB, but when I try pip install viam-sdk I get could not find a version (from versions:none)

#

Editing so the right answer is here

I used Jetson Hacks with a slight modification to install Python 3.9, set up a virtual environment with alias so that python and python3 both point to python3.9, and I could then pip install viam-sdk

Doing the manual build is possible, but as all the above needs to happen first, there is no point.

inland sky
#

Hey Brian, would love a bit more context on this issue so I can help troubleshoot. What version of python/pip do you have installed?

lucid bridge
#

Python is 3.6.9, pip is 21.3.1, OS is Ubuntu 18.04, hardware is Jetson Nano

#

On the build, I am also getting an error
poetry build
RuntimeError
The Poetry configuration is invalid:
- Additional properties are not allowed ('group' was unexpected)

lucid bridge
#

I installed a newer version of poetry using curl -sSL https://install.python-poetry.org | POETRY_VERSION=1.2.0a2 python3 -

lucid bridge
#

Building a newer python version came with it's own issues, dealt with here https://github.com/JetsonHacksNano/build_python/issues/2

Plus I tried newer versions of poetry (1.7.1) and that was not going to work, so went back to 1.2.0a2

Finally, poetry kept insisting that the python and python3 versions were old so it created a new venv, so I fixed my .local/venvs/python3_9 by adding the following 2 lines to the bin/activate file

alias python=python3.9
alias python3=python3.9
GitHub

Describe the issue Built 3.11,3.10 OK, not 3.9 in that order. gbp buildpackage --git-ignore-branch command fails in the script with: dpkg-checkbuilddeps: error: Unmet build dependencies: libmpdec-d...

#

I will confirm once the build is done if this worked.

lucid bridge
#

OK, viam-sdk installs using pip fine once the python is 3.9 or higher. Building from scratch is not needed.

inland sky
#

Okay glad this was figured out! It is definitely recommended to have the latest stable version of Python if possible

lucid bridge
#

Of course using Docker can fix this by running a more modern OS on top of the Nano