Are.na Dev Meetup #1 Notes

What’s Are.na working on right now?

  • User onboarding
  • Mobile app
  • GraphQL will be available in addition to REST API when the mobile app goes into beta (few months). REST API is not going away.


  • Are.na has a little less than 1M blocks
  • Slightly more that 1M connections
  • Are.na is split about 50/50, public/private


Python library

  • Francis demos his in-progress python interface. Built for his Are.na feed to Slack repo slam-dunk.
  • Most of the routes should work, but only what’s needed for slam-dunk has been tested.
  • Still missing documentation but try it out.

Sam’s Demo

  • Are.na network visualization app spider, gives you a few parameters to create a graph with live links to channels and users.
  • Interested in building applications for research as a community, see Research Tactics
  • Quick explanation of new Are.na topic modeling package, leibniz, uses MCMC and word2vec
  • What can we infer from the latent structure of the channel-channel or block-channel graph?
  • Could a bot that sees the graph differently than a typical user and packages analyses as channels be interesting?
  • Possible to make widgets that seamlessly integrate into the Are.na UI by leveraging Are.na's ability to pull in embedded content (currently uses embedly to whitelist).
  • An Are.na enhancement extension, similar to RES could also be an interesting site for developers to collaborate on community projects.

Bot discussion

  • Are.na is committed to not using algorithms to dictate what users see, but are there ways for users to surface interesting content themselves using their own algorithms?
  • What makes Are.na special is its users and how their intent structures information. Bots have the potential to dilute this.
  • Bots aren’t interesting in themselves, what users do with bots might be (intent is key).
  • Generally, it’s a positive experience when someone discovers a research collection you may have assembled in the past. A bot that kicks up this information could potentially be obnoxious.
  • Bots could do maintenance tasks, like building superchannels, to minimize disruption because users can choose to unfollow.
  • However notifications are not consensual at the moment (currently no way to suppress a notification if a bot starts connecting things you own).
  • Bots that connect to services outside of Are.na, twitter, slack, etc., could link people back in and bring more users.
  • Devs should be encouraged to build things, but mindful of the above. We should look to set a precedent that’s in keeping with the community’s values.

Are.na as a platform for agency work

  • This was something that had been discussed from the very beginning, using Are.na as a backend for arranging ideas, and then structuring that information for client presentations or full websites, see Cleopatra's.
  • Seems many people would still be interested in integrating more tightly with Are.na, just need a few solid components & documentation.

Has Are.na considered allowing users to add information to connections?

  • Yes, there is some concern the UI would be too confusing, especially for new users.
  • What are the possible use cases?
    • Specify how and when a website/bot built with Are.na posts a piece of content.
    • Contextually describe ownership and other associated properties.
  • See Categorical Are.na

Channels to follow

Are.na Dev Meetup #1 Notes