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/issues/890#issuecomment-409321824,https://api.github.com/repos/pydata/xarray/issues/890,409321824,MDEyOklzc3VlQ29tbWVudDQwOTMyMTgyNA==,5635139,2018-07-31T18:28:29Z,2018-07-31T18:28:29Z,MEMBER,"Closing - while the behavior could be surprising at first glance, for @shoyer 's reasons above: not much to gain from changes","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,161991202
https://github.com/pydata/xarray/issues/890#issuecomment-228408435,https://api.github.com/repos/pydata/xarray/issues/890,228408435,MDEyOklzc3VlQ29tbWVudDIyODQwODQzNQ==,1217238,2016-06-24T17:27:56Z,2016-06-24T17:27:56Z,MEMBER,"This may not be clearly documented, but it is/was intentional behavior.

To be clear, there are two things going on here:
1. Arguments like `{'a': [2, 3]}` get implicit dimensions added, e.g., `{'a': ('a', [2, 3])}`.
2. 1D `data_vars` arguments which have the same name as their sole argument are converted in `coords`.

Originally, this made sense because we didn't have a separate `coords` argument. Now, this could simply be passed as `coords` instead, which is certainly more explicit. So instead of allowing this we could raise an error. My concern such a deprecation this is that (1) there's a lot of code written with the assumption that this works and (2) there's not much be to be gained -- there's nothing else that we could sensibly use this syntax for, and (3) this is marginally convenient, allowing for slightly less typing (one less pair of curly braces).
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,161991202