might be better to grab work type from the LinguisticObject: ``` getValueByClassification(normalizeFieldToArray(object, 'referred_to_by'), "http://vocab.getty.edu/aat/300435443");

One other thing to note here, we may come in compliance here with LinkedArt, hopefully.

Do we want to mention that we have dimensions as structured data, not just as descriptions? I'm not opposed to showing the tools to parse out strings, but it's not required to do so!

Makes sense to me! I've added a note showing how to get the dimensions directly. (At least in this dataset, it looks like some records don't have the dimension section.) I also like using the description because it has a helper function, and I think it's nice to show a little string parsing.

is it worth commenting on the fact that dimension units will vary between objects, and it may be worth checking and handling the conversion -- there are a few in mm

Yes, definitely! I will note this.

is it worth noting the Museum's collection policy and how that skews or effects the data? Paintings don't really get much past the Impressionists, and then all we have is a few sculptures and photographs in the 20th and 21st century

Yes, this is a good corrective, and I've added a short note. Would you mind checking to make sure I've characterized the collection policy correctly?

That works for me, I like how you've worked around it. thanks.

Thinking about your question in the email, I like the idea of adding this into the 1st tutorial, perhaps first getting the title or other things the easy way? Or, just doing one and adding a note for the rest saying, this helper exists?

Yes, I agree! I'll add it shortly.

Is it worth explaining how Promise.all works, or linking to the docs for it? I've got no sense of how many people know about that function.