-
-
Notifications
You must be signed in to change notification settings - Fork 34
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
Keeping order of installs #124
Conversation
b8a5792
to
9ab940d
Compare
# We then install Numpy itself at the bottom of this script | ||
export CONDA_INSTALL="conda install $QUIET python=$PYTHON_VERSION" | ||
elif [[ $NUMPY_VERSION == stable ]]; then | ||
conda install $QUIET numpy=$LATEST_NUMPY_STABLE | ||
conda install $QUIET --no-pin numpy=$LATEST_NUMPY_STABLE |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Will it not be the case that when installing the main dependencies later, there will still be the issue that the pin file can contain numpy stable*
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
yes, probably you're right, I'll look at this again.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
No, numpy isn't listed in the pinfile, e.g. for a NUMPY_VERSION=stable CONDA_DEPENDENCIES='jinja2=2.8'
build, the pinfile contains only jinja:
++++cat /home/travis/miniconda/envs/test/conda-meta/pinned
jinja2 2.8*
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Do you think it's better to pin down the numpy version, too?
9ab940d
to
dec42fd
Compare
…he conda dependencies in the numpy and astropy steps
dec42fd
to
bf7e924
Compare
I'm merging this now, if there is a problem coming up we can still pin the numpy version, too. |
This is an attempt to address #123.
Not sure how it can be tested, but e.g. in the prerelease travis test, scipy now shouldn't be installed even though it's listed in CONDA_DEPENDENCIES.
If this works, I'll add more commits to deal with the astropy step.