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/1452#issuecomment-309171587,https://api.github.com/repos/pydata/xarray/issues/1452,309171587,MDEyOklzc3VlQ29tbWVudDMwOTE3MTU4Nw==,1217238,2017-06-17T00:19:32Z,2017-06-17T00:19:32Z,MEMBER,"It looks like this variable has an attribute `_FillValue = -1`, which isn't a valid string value: 
```
In [18]: ds['tidenames'].encoding
Out[18]:
{'_FillValue': -1,
 'dtype': dtype('S1'),
 'original_shape': (38, 64),
 'source': 'http://geoport.whoi.edu/thredds/dodsC/usgs/vault0/models/tides/vdatum_gulf_of_maine/adcirc54_38_orig.nc'}
```

Potentially we could update xarray's CF conventions code to handle inconsistent dtypes and `_FillValue` attributes, or possibly this file should be fixed instead.

A work around is to set `mask_and_scale=False`:
```
In [19]: ds = xr.open_dataset(url, mask_and_scale=False)

In [21]: ds['tidenames']
Out[21]:
<xarray.DataArray 'tidenames' (ntides: 38)>
array([b'STEADY ', b'MN     ', b'SM     ', b'K1     ', b'O1     ', b'P1     ',
       b'Q1     ', b'SO1    ', b'MNS2   ', b'2MS2   ', b'N2     ', b'M2     ',
       b'2MN2   ', b'S2     ', b'K2     ', b'MSN2   ', b'2SM2   ', b'MO3    ',
       b'SO3/MK3', b'SK3    ', b'N4     ', b'3MS4   ', b'MN4    ', b'M4     ',
       b'3MN4   ', b'MS4    ', b'2MSN4  ', b'2NM6   ', b'2MN6   ', b'M6     ',
       b'MSN6   ', b'2MS6   ', b'2SM6   ', b'2(MN)8 ', b'3MN8   ', b'M8     ',
       b'2MSN8  ', b'3MS8   '],
      dtype='|S64')
Dimensions without coordinates: ntides
Attributes:
    long_name:      Tide Constituent
    missing_value:  -1
    standard_name:  tide_constituent
```","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,235687353