Curie lists have random order as they are constructed as set() internally.

[mikenotfound]

Looking through the code it appears curies are initially constructed as a set() but subsequently converted to a plain list(). Therefore the order of curies is randomised.

This not only seems at odds with the expected behaviour of a list (or list-like) structure, but means tests become more complex --- having to convert the actual, and expected output to compare curies as sets is painful.

Perhaps this is more of an edge case, but would the developers be willing to discuss this possibility?