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/1749#issuecomment-348635463,https://api.github.com/repos/pydata/xarray/issues/1749,348635463,MDEyOklzc3VlQ29tbWVudDM0ODYzNTQ2Mw==,2443309,2017-12-01T22:52:13Z,2017-12-01T22:52:13Z,MEMBER,"A practical downside to masking both values is that we won't be able to round-trip the two part fill/missing value. Once everything is a NaN, that's all we get.  That said, I'm personally fine with masking both.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,278122300
https://github.com/pydata/xarray/issues/1749#issuecomment-348374694,https://api.github.com/repos/pydata/xarray/issues/1749,348374694,MDEyOklzc3VlQ29tbWVudDM0ODM3NDY5NA==,1217238,2017-12-01T01:27:43Z,2017-12-01T01:27:43Z,MEMBER,"I can certainly see a case for taking a more relaxed approach to `_FillValue` and `missing_value`, e.g., masking on both rather than raising an error. I'm not sure that it's actually more user friendly to ignore `missing_value` for masking.

This part of xarray is very old (added in https://github.com/pydata/xarray/pull/245), before almost anyone outside Climate Corporation was using it. So I'm certainly open to revisiting this, but let's make sure we get the decision right this time.

Does anyone else with lots of experience with different netCDF files have opinions here? Maybe @dopplershift?","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,278122300