Ryan May
In cleaning up docs, adding tests, and a few other minor things, is it ok if I just push to your branch from here?
@jhprinz Does that mean this issue is fixed for you now?
Ok, now I'm running into this, though slightly differently: ``` python from netCDF4 import Dataset import numpy as np nc = Dataset('test.nc', 'w') vlen_type = nc.createVLType(np.float64, 'vltest') nc.createDimension('x', None) v...
So that test fails because it is requesting data from a slice that is outside the actual bound of the data. Previously, this would return an empty array; with these...
I think the feature seems reasonable. Could also make it take a datetime as an optional argument, defaulting to utcnow.
Well, to be fair I'm pretty sure "most users" install pre-built binaries from linux package managers, Homebrew, conda-forge, underpants gnomes on Windows (ok, probably not, but you get the idea)....
What you've done here looks perfectly sensible to me. As far as ramifications to the TDS, I'd defer to @lesserwhirls: how much load would we induce on the server by...
I wonder if [intake-thredds](https://intake-thredds.readthedocs.io/en/latest/tutorial.html#loading-a-catalog) can do this already? `intake.open_thredds_merged()` seems particularly relevant.
@brianmapes Dask is about making virtualized numpy-compatible arrays that wrap chunks of data. This allows you to have arrays larger than fit in memory. In order to work with that...