Adrian Sampson

Indeed; when I read over my comment from earlier, I found myself disagreeing with my previous self. :smiley: It does seem like conflicting "and"-joined queries would most logically just return...

Yep, let's add pagination. There was some discussion about this in https://github.com/sampsyo/beets/issues/750 and https://github.com/sampsyo/beets/issues/718. Some people (e.g., @asutherland) have suggested that start/end or offset/limit can be an efficiency problem in...

I should add that option 3 looks like a good 80% solution that would be easy to support, so I lean that way at the moment.

Very cool. I'm enthusiastic about this opaque-token pagination idea. To save any other readers a little googling, [here's a discussion of the problems with limit/offset](https://bugzilla.mozilla.org/show_bug.cgi?id=735102#c5), and [here's an implementation of...

I've put together a draft of the pagination idea: http://auraspec.readthedocs.org/en/latest/api.html#pagination Comments and criticism welcome!

Good point. Two important things were missing: the token is single-use, and the server needs to be able to expire tokens if it wants to (to allow reasonable stateful implementations)....

Hi! Thanks for looking into this. When we add new tag mappings to MediaFile, it's important to justify them by looking at existing software that uses the tags so we...

Yes, good point—I was thinking about this too. At the very least there should be a way to distinguish categories of images, and image data like the MIME type and...

I just pushed a draft of a more complete spec for images: http://auraspec.readthedocs.org/en/latest/api.html#images It might be a good idea to add a more complete list of possible "role" values (I...

One drawback to what I just proposed: it might be a pain for servers to enumerate all the images in the system at `/images`, and it certainly seems needless to...