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/2551#issuecomment-437718958,https://api.github.com/repos/pydata/xarray/issues/2551,437718958,MDEyOklzc3VlQ29tbWVudDQzNzcxODk1OA==,2448579,2018-11-12T00:12:46Z,2018-11-12T00:12:46Z,MEMBER,"""Perhaps we should raise our own error though instead of just passing things through to netCDF4""

This seems like a good idea","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,379177627
https://github.com/pydata/xarray/issues/2551#issuecomment-437717439,https://api.github.com/repos/pydata/xarray/issues/2551,437717439,MDEyOklzc3VlQ29tbWVudDQzNzcxNzQzOQ==,1217238,2018-11-11T23:52:34Z,2018-11-11T23:52:34Z,MEMBER,"Previously the indexing operation would sometimes return a NumPy array. Now it's always lazy, so accessing the dataset when the file is closed fails.

Perhaps we should raise our own error though instead of just passing things through to netCDF4.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,379177627
https://github.com/pydata/xarray/issues/2551#issuecomment-437707951,https://api.github.com/repos/pydata/xarray/issues/2551,437707951,MDEyOklzc3VlQ29tbWVudDQzNzcwNzk1MQ==,10050469,2018-11-11T21:46:02Z,2018-11-11T21:46:02Z,MEMBER,">  but now operates lazily.

If so, how does the dataset access the data if the file is closed? Shouldn't it raise an error instead?","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,379177627
https://github.com/pydata/xarray/issues/2551#issuecomment-437707332,https://api.github.com/repos/pydata/xarray/issues/2551,437707332,MDEyOklzc3VlQ29tbWVudDQzNzcwNzMzMg==,1217238,2018-11-11T21:37:42Z,2018-11-11T21:37:42Z,MEMBER,"My best guess is that `ds.thick.isel(x=('z', [1, 3, 4]), y=('z', [7, 2, 1]))` previously loaded data into memory (and cached it) but now operates lazily.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,379177627
https://github.com/pydata/xarray/issues/2551#issuecomment-437705060,https://api.github.com/repos/pydata/xarray/issues/2551,437705060,MDEyOklzc3VlQ29tbWVudDQzNzcwNTA2MA==,10050469,2018-11-11T21:08:13Z,2018-11-11T21:08:13Z,MEMBER,Actually I have to ask: why is it possible to to an operation on a closed dataset? What is happening under the hood which creates this error and wasn't earlier?,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,379177627
https://github.com/pydata/xarray/issues/2551#issuecomment-437664983,https://api.github.com/repos/pydata/xarray/issues/2551,437664983,MDEyOklzc3VlQ29tbWVudDQzNzY2NDk4Mw==,10050469,2018-11-11T12:10:36Z,2018-11-11T12:10:36Z,MEMBER,"> Yep, that looks a bug on your end. Were you using autoclose=True before?

No, the exact same piece of code works without error with xarray prev 0.11. Closing this now!","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,379177627
https://github.com/pydata/xarray/issues/2551#issuecomment-437618316,https://api.github.com/repos/pydata/xarray/issues/2551,437618316,MDEyOklzc3VlQ29tbWVudDQzNzYxODMxNg==,1217238,2018-11-10T20:18:31Z,2018-11-10T20:18:31Z,MEMBER,"@fmaussion Yep, that looks a bug on your end. Were you using `autoclose=True` before? If so, that could have hidden the issue, because exiting the `with` block didn't use to actually close the file in that case.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,379177627
https://github.com/pydata/xarray/issues/2551#issuecomment-437574599,https://api.github.com/repos/pydata/xarray/issues/2551,437574599,MDEyOklzc3VlQ29tbWVudDQzNzU3NDU5OQ==,10050469,2018-11-10T10:39:09Z,2018-11-10T10:40:35Z,MEMBER,"> I am slightly regretting not doing a release candidate here,

I should be better placed than anyone to anticipate and test more frequently on master :roll_eyes: 

So this reproduces the problem on my machine and [travis](https://travis-ci.org/pydata/xarray/builds/453229051?utm_source=github_status&utm_medium=notification). Note the misplaced computation out of the `with` block:

```python
import xarray as xr
import numpy as np
import netCDF4
import os

try:
    os.remove('test.nc')
except OSError:
    pass

thick = np.zeros((180, 200), np.float32)

ds = xr.Dataset()
ds['thick'] = (('y', 'x'), thick)
ds.to_netcdf('test.nc')

with xr.open_dataset('test.nc') as ds:
    dummy = ds.thick.isel(x=('z', [1, 3, 4]), y=('z', [7, 2, 1]))
    
dummy.min()  # this is the issue

with netCDF4.Dataset('test.nc', 'a') as nc:
    nc.variables['thick'][:] = thick  
```

I guess this was actually a bug in our code, but for some reason it worked fine with earlier xarray versions. I'm not sure if this requires action on the xarray side in the end...","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,379177627
https://github.com/pydata/xarray/issues/2551#issuecomment-437483895,https://api.github.com/repos/pydata/xarray/issues/2551,437483895,MDEyOklzc3VlQ29tbWVudDQzNzQ4Mzg5NQ==,1217238,2018-11-09T20:22:03Z,2018-11-09T20:22:03Z,MEMBER,"I am slightly regretting not doing a release candidate here, but hopefully this should be straightforward to fix.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,379177627
https://github.com/pydata/xarray/issues/2551#issuecomment-437471566,https://api.github.com/repos/pydata/xarray/issues/2551,437471566,MDEyOklzc3VlQ29tbWVudDQzNzQ3MTU2Ng==,1217238,2018-11-09T19:37:15Z,2018-11-09T19:37:15Z,MEMBER,"Oh my -- sorry about that!

Thinking about this a little more, I guess this should not be too surprising since I don't think we have any dask integration tests that cover appending to existing files. Maybe a good place to start would be adding one of those, e.g., adapted from this existing test:
https://github.com/pydata/xarray/blob/575e97aef405c9b473508f5bc0e66332df4930f3/xarray/tests/test_distributed.py#L66","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,379177627