Google has released documentation to the Mirror APIs, the API to program against Google Glass. Google Glass, if you spent the past year living under a rock — and even if you didn’t — is Google’s “I’ve got a weird little thing on my glasses that gives me useful information” upcoming product.
While Glass is supposedly powered by Android, that’s not the way to develop applications for us regular folk. Instead, we get to build web services. From what I can deduce from the documentation, the way it works is that your glass keeps a persistent connection with Google’s servers, which calls third-party web services on behalf of your glass. Similarly, if you need to push stuff to glass, you do so by calling Google’s webservice end-point, which in turn will push this data to the glass in question.
Diagrammatically it would look something like this:
Glass offers four groups of APIs:
Timeline, to shove notifications in the glasser’s eye.
Subscriptions, to receive notifications of, for instance, a glasser calling an action on your timeline item (for instance “reply”, or “delete”).
Locations, to enable stalking.
Contacts, to create, update and delete contacts (supposedly, this is what you’d use if you want to build a service that syncs Facebook or LinkedIn contacts).
For now, only a developer edition is available of glass to a select group of developers (that were willing to pay $1.5k for alpha quality hardware) — the rest of us will have to wait until 2014 to look as cool as the — clearly satisfied — glass users below.