Unfortunately, edx-platform depends on this package. In practice, it means that all builds of the “openedx” Docker image with Tutor will fail with the following error:
$ tutor images build openedx
...
#58 48.17 ERROR: Could not find a version that satisfies the requirement py2neo==2021.2.3 (from versions: none)
#58 48.17 ERROR: No matching distribution found for py2neo==2021.2.3
Temporary fix, for example I’m still using maple, and what i did is to apply the patch manually from fix: build error caused by removed py2neo package by regisb · Pull Request #916 · overhangio/tutor · GitHub into the Dockerfile (in my case /home/ubuntu/.pyenv/versions/3.7.17/lib/python3.7/site-packages/tutor/templates/build/openedx/Dockerfile) and then run tutor config save to update the Dockerfile in the current env, and the build process was no longer error.
Hi Mr. Regis, I am using Nutmeg Version 14.0.4 and I installed it by the binary method, how I can upgrade my current binary version 14.0.4 to 14.2.5 but with binary way? Appreciate your help.
This is a sensible answer, but if you don’t mind could you share the reason? We use a build CI that uses TVM and can only take tagged versions of tutor as input. So I’d like to understand how this could come up in the future to update our CI scripts.
If we published the tags on GitHub, then we would also have to publish the release branches. There is nothing confidential there, but it would be leading us down a slippery road, where we would have to maintain branches of past releases.
I understand this is inconvenient, and I’m prepared to change my mind on this. But it’s an important change of direction, and thus we are not going to make that change overnight.
Our CI is built to use only tagged versions from github as well. I’m sure that I can adjust the script to install from pip instead.
If the docs state that installing from binaries is supported, I would assume that it meant it is supported for all version, including any patches. I appreciate that keeping the binaries updated as well as pip for a number of past versions uses valuable time, but it would be appreciated.
Maybe the docs need to be updated to better push users to use the pip install method rather than the binaries. I will be scheduling time in the future to update our CI scripts to use pip.
I am using TVM(tutor version manager) so, Any suggestions on how I should update to the latest version?
1. Install the latest stable release of TVM.
pip install git+https://github.com/eduNEXT/tvm.git
2. Verify the installation.
tvm --version
3. Create a new project with TVM.
tvm project init tvm-palm v16.1.0
4. Open the project folder.
cd <project-name>
5. Activate the project environment.
source .tvm/bin/activate
6. Run your project.
tutor local launch
@regis Please update tutor/CHANGELOG.md at master · overhangio/tutor · GitHub to indicate the existence of v13.3.2 and this bug. I was using the changelog to find the latest version of each release as I try to do a version-by-version upgrade. So since the changelog only mentioned 13.3.1, I thought that was the latest version, so I hit this error and had to google it to find this. (Apparently the same thing with 15.3.9 which isn’t mentioned, as 15.3.7 is the last thing in the changelog…14.2.5 (14.2.3 is the latest), etc. I guess I’ll have to start using tutor · PyPI instead as the source of truth for latest versions since there’s not even a 13.3.2 tag on the github…)
