@loarie – a few questions / thoughts:
-
i see the AWS open dataset out there (https://registry.opendata.aws/inaturalist-open-data/). i am able to download each of the 4 individual metadata files (looks like just a small test set of data right now), but i can’t download the bundle, and i can’t list the contents of the directory either. does that just reflect that you all are still working on things until the official April 15 launch?
-
i was thinking about the usefulness of the metadata files (vs just querying for data in iNaturalist). the photos file is already pretty big even with just a sample set of 4MM records. i wonder if breaking up that file by photo license type might make it easier to use the open dataset, while emphasizing the importance of the photo licenses?
-
i noticed that the current test versions of the photos and observations metadata files have user id, but the user metadata file has only user id + login. does it make sense to also add a real user name in the user metadata file – if the user has specified it in their profile – so that the user could theoretically receive proper attribution?
-
the license in the AWS open data registry entry reads as follows:
Creative Commons or Public Domain (CC0), varying by image. More information on how to query and properly treat licenses can be found in the documentation
i sort of think this may not be clear enough to someone just discovering iNaturalist and the dataset that photos are licensed individually by the users (as opposed to by iNaturalist), and observations are separately licensed, too. could this be rephrased or elaborated a bit?
-
in the API recommended practices doc (https://www.inaturalist.org/pages/api+recommended+practices), there was an item related to media downloads:
Downloading over 5 GB of media per hour or 24 GB of media per day may result in a permanent block
i’ve always read this to apply to downloading photos. but i’m guessing this limit no longer really applies to downloading photos that are part of the AWS open data set, right?