-
-
Notifications
You must be signed in to change notification settings - Fork 1.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
TypeError: unorderable types: str() < NoneType() #137
Comments
Original comment by jaraco (Bitbucket: jaraco, GitHub: jaraco): I don't get that error when I run the command in my Python 3.4b2 environment on Windows. I haven't yet tried on another platform. However, given that this code is largely unchanged over the years and that this issue has not been more broadly reported, I suspect there is something unique to your environment that's triggering the error. That doesn't mean that it's not a bug, but only that it isn't a widespread bug, and because I can't replicate it, I'll need you to provide more detail about the nature of the issue. Can you confirm that you still have the issue? If so, does the issue persist if you downgrade to an older setuptools (such as 2.0.2)? If the problem persists, perhaps do this for me to help us understand the nature of the failure better:
That will launch pdb (the python debugger) with the easy_install script. It'll break right away so hit I expect it will proceed and then fail with the same traceback as above. If it does, it should leave you at a |
Original comment by lgautier (Bitbucket: lgautier, GitHub: lgautier): I experienced a similar issue when moving a project from
to
The problem disappeared after I created a new virtualenv and try to install in it. |
Original comment by david_a_joy (Bitbucket: david_a_joy, GitHub: Unknown): I just experienced this on centos6 with a source built python3.4.1. This was installing setuptools directly from setup.py (after getting the same error with a pip install --upgrade setuptools).
Running the debugger as instructed above:
So matplotlib is somehow in my site-packages twice. That's weird, but probably due to the fact that I upgraded setuptools in the middle of a deployment, then tried to upgrade matplotlib again.
Removing all the matplotlib related files from site-packages makes setup.py succeed with no errors. If I reinstall matplotlib, now I get this:
So I'm thinking this happens when there are old egg-info files laying around? |
Original comment by craigspry (Bitbucket: craigspry, GitHub: craigspry): I don't know if this helps but this is how I've managed to reproduce this, I've done this on Linux Mint,running under virtual box and on a koding.com free vm. First I installed matplotlib for python2.7 using: This got me the same result as has been discussed above. On OSX I managed to install matplotlib for python2.7 and python3, but I installed matplotlib for python2.7 so long ago I'm not sure how I installed it, but I suspect I used pip, for python3 I used pip. |
Original comment by minrk (Bitbucket: minrk, GitHub: minrk): I've seen this as a result of a mixture of non-setuptools and setuptools installs, where the non-setuptools install of a package has |
Original comment by minrk (Bitbucket: minrk, GitHub: minrk): I can reproduce this with a conda environment by doing a non-setuptools install after a setuptools install, ensuring that both are present:
The result is two |
Original comment by asif_m (Bitbucket: asif_m, GitHub: Unknown): I solved my problem (Win 8.1, Python 3.4.2) with the change suggested here: |
Detect when a venv is created from an in-tree build. Fixes #132
Originally reported by: ionelmc (Bitbucket: ionelmc, GitHub: ionelmc)
The text was updated successfully, but these errors were encountered: