Sorry about the thumbnailer issue on this notebook! It was consistently crashing because I believe it was trying to capture this tall table. Not your fault at all, but rather ours or puppeteer's.

Thanks! Maybe it is Observable's way of suggesting I use a virtualized table grid rather than write out the entire table... :)

It seems like if we distinct by `brewery_id`, we get different counts. Some brewery names are not unique. Maybe that's intentional ;-)

There are likely errors in the data in which the same brewery has multiple entries (same name, different id), so using brewery-name is actually a bit more accurate. However, I think that is a data cleaning issue that should be resolved separately, and this example is probably better off using the id to determine distinctness here. Thanks!