Kristen Thyng

`metadata` is already in `kw`, though just the original set of metadata, not the updated metadata.

I did — I tried adding it in and got: ``` In [1]: import intake_erddap ...: ...: cat = intake_erddap.ERDDAPCatalogReader(server="https://erddap.aoos.org/erddap", search_for=["aoos_204"]).read() ...: source_name = list(cat)[0] ...: cat.get_entity(source_name).metadata.update({"blah": "blahblah"}) In [2]:...

@martindurant Ping to take a look at this if you can.

I just pulled all the latest updates from github and ran this same snippet: ``` import intake_erddap cat = intake_erddap.ERDDAPCatalogReader(server="https://erddap.aoos.org/erddap", search_for=["aoos_204"]).read() source_name = list(cat)[0] cat.get_entity(source_name).metadata.update({"blah": "blahblah"}) ``` then ``` cat[source_name].metadata...

No, I didn't see that! I'll take a look, thanks.

Ok so I see your change and I've been working with it. Now my entry metadata isn't included since you moved it from a kwarg to a separate input. https://github.com/kthyng/intake-erddap/blob/0b4e66b02bd2c625ec7900a0a2f5b0510f89d925/intake_erddap/erddap_cat.py#L462-L464...

Do you mean with your change to the branch and newest changes in intake, you still see the metadata? I'm having the metadata come and go when I add and...

Ok yes, I am sorry. I see that I reinstalled `intake-erddap` locally and in doing so it reinstalled `intake` from PyPI and that was the problem. Now it is sorted...

`ERDDAPCatalogReader` is a reader, subclassing `BaseReader`. I just now tried switching `TableDAPSource` to subclass `BaseReader` too but it created a RecursionError at the `ToHvPlot` line. Are Catalogs meant to subclass...

Interesting! Thanks for this work, I appreciate it. I had been wondering if I could remove a lot of that but was worried about going too far off course; plus,...