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/2844#issuecomment-598016992,https://api.github.com/repos/pydata/xarray/issues/2844,598016992,MDEyOklzc3VlQ29tbWVudDU5ODAxNjk5Mg==,1217238,2020-03-12T05:43:30Z,2020-03-12T05:43:30Z,MEMBER,"We could probably make it a rule that encoding gets preserved in exactly the same way as attrs. I think there are some old issues about formalizing metadata conventions... On Wed, Mar 11, 2020 at 2:19 PM Ryan May wrote: > Thanks for the info. Based on that, I lean towards attrs. > > I think a better rationale, though, would be to formalize the role of > encoding in xarray. > > — > You are receiving this because you commented. > Reply to this email directly, view it on GitHub > , or > unsubscribe > > . > ","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,424265093 https://github.com/pydata/xarray/pull/2844#issuecomment-478298195,https://api.github.com/repos/pydata/xarray/issues/2844,478298195,MDEyOklzc3VlQ29tbWVudDQ3ODI5ODE5NQ==,1217238,2019-03-30T23:27:19Z,2019-03-30T23:27:19Z,MEMBER,"The current VariableCoder interface only works for coders that work at the level of individual variables. But coordinates only make sense on a dataset, so we'll need a different abstraction for that, e.g., a DatasetCoder? For now, it's probably fine to keep this logic where you have it currently.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,424265093