html_url,issue_url,id,node_id,user,created_at,updated_at,author_association,body,reactions,performed_via_github_app,issue
https://github.com/pydata/xarray/pull/1274#issuecomment-280812037,https://api.github.com/repos/pydata/xarray/issues/1274,280812037,MDEyOklzc3VlQ29tbWVudDI4MDgxMjAzNw==,6628425,2017-02-18T01:51:46Z,2017-02-18T01:51:46Z,MEMBER,"Cool, I think I have things working now.  Please let me know if you'd like me to make any changes.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,208431831
https://github.com/pydata/xarray/pull/1274#issuecomment-280643863,https://api.github.com/repos/pydata/xarray/issues/1274,280643863,MDEyOklzc3VlQ29tbWVudDI4MDY0Mzg2Mw==,6628425,2017-02-17T13:05:30Z,2017-02-17T13:05:30Z,MEMBER,"@shoyer it's entirely possible I'm in over my head here, as I'm not a Windows user, so I can't test this setup locally.  It looks as though using the existing requirements files (which install more than what was specified in the original ```appveyor.yml``` configuration file) was probably not a good idea.

If it's a matter of simply creating new requirements files for the Windows environments to mirror what was installed in the original ```appveyor.yml``` file, i.e. pip, pytest, numpy, pandas, scipy, netCDF4, matplotlib, and dask, (this time using the conda-forge channel) then I can probably handle that, but anything more is guesswork on my end, so I'd be fine closing this PR.  I was hoping for a straightforward fix.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,208431831