Repair Appveyor #697
Repair Appveyor #697
Conversation
|
Ok, looks like it works. |
Codecov Report
@@ Coverage Diff @@
## devel #697 +/- ##
==========================================
- Coverage 81.77% 81.66% -0.11%
==========================================
Files 36 36
Lines 3644 3644
Branches 832 832
==========================================
- Hits 2980 2976 -4
- Misses 470 474 +4
Partials 194 194
Continue to review full report at Codecov.
|
appveyor.yml
Outdated
| @@ -47,10 +47,10 @@ install: | |||
| if ($env:CONDA -eq "true") { | |||
| conda config --set always_yes yes --set changeps1 no; | |||
| conda install -q pip setuptools wheel numpy scipy; | |||
| pip install -r develop-requirements.txt; | |||
| pip install . | |||
| pip install --disable-pip-version-check -r develop-requirements.txt; | |||
There was a problem hiding this comment.
Rather than setting these options on each command I prefer to set an environment variable PIP_DISABLE_PIP_VERSION_CHECK=yes.
There was a problem hiding this comment.
In fact, I just set this variable through the interface so you don't have to worry about it. These options can simply be removed.
There was a problem hiding this comment.
Would it bother if I still set the environment variable in the config? Just so that it is explicit and people could fork the config and it would still work.
cobra/util/version_info.py
Outdated
| @@ -55,7 +55,8 @@ def get_pkg_info(): | |||
| # using requirements files that can be read in. | |||
| dependencies = frozenset(PKG_ORDER) | |||
| blob = dict() | |||
| for dist in pip.get_installed_distributions(): | |||
| dists = [d for d in pkg_resources.working_set] | |||
There was a problem hiding this comment.
Since dists is not used for anything else can you combine these two lines to:
for dist in pkg_resources.working_set:please?
There was a problem hiding this comment.
Will do. This is actually taken from #696.
So this is an attempt to repair appveyor.
I suspect the problem is that conda has not yet updated pip to version 10. So even if you update pip with conda you get pip 9. this triggers a warning from pip to stderr which appveyor interprets as failure (wrongly so).