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/1323#issuecomment-373810845,https://api.github.com/repos/pydata/xarray/issues/1323,373810845,MDEyOklzc3VlQ29tbWVudDM3MzgxMDg0NQ==,2443309,2018-03-16T18:50:56Z,2018-03-16T18:50:56Z,MEMBER,@ebo - I think your issue falls more in line with what was discussed in #1582. Can you read over that issue and perhaps comment there?,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,216621142
https://github.com/pydata/xarray/issues/1323#issuecomment-289030691,https://api.github.com/repos/pydata/xarray/issues/1323,289030691,MDEyOklzc3VlQ29tbWVudDI4OTAzMDY5MQ==,6815844,2017-03-24T14:02:24Z,2017-03-24T14:02:24Z,MEMBER,"@shoyer  Thank you for the comment.

> What metadata would you get from images? 

I'm currently use tag data in tiff files. I guess some people may use exif metadata in jpg files.

> If we added this, we would want to defer all the image reading out to an external library

Yes, I agree.
skimage is the best candidate?
I did not know much about skimage, but this library looks great.

open_image method is possibly overkill...? 
I will keep this issue just open for a while until similar request comes out.

> You're definitely not the first person to ask about this:

My second request seems more general.
I will post it as a separate issue.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,216621142
https://github.com/pydata/xarray/issues/1323#issuecomment-288947882,https://api.github.com/repos/pydata/xarray/issues/1323,288947882,MDEyOklzc3VlQ29tbWVudDI4ODk0Nzg4Mg==,1217238,2017-03-24T06:58:38Z,2017-03-24T06:58:38Z,MEMBER,"> xr.open_image(File)

What metadata would you get from images? `[row, column, channel]` for a typical color image? I guess this would usually follow skimage's [coordinate conventions](http://scikit-image.org/docs/dev/user_guide/numpy_images.html#coordinate-conventions)

If we added this, we would want to defer all the image reading out to an external library such as skimage, e.g., by leveraging [`skimage.io.imread`](http://scikit-image.org/docs/dev/api/skimage.io.html)

> `xr.DataArray.expand_dims(dim)`

You're definitely not the first person to ask about this:
http://stackoverflow.com/questions/34987972/expand-dimensions-xray

Yes, this would definitely be welcome in some form.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,216621142