I evaluated a few things at the beginning. I was previously using Ruby/Rails and Node.js, and Rails was my go-to for full-stack webdev.
There was some interest in building a headless WordPress site, but the limitations on database design and performance led me back to a monolith. I have a general preference for monoliths when time is tight (time is always tight) -- the things you get for free are very helpful: deploy/rollback, DB migrations, API versioning, local dev workflow, CI, among others.
I'd deployed a couple small Elixir apps at my previous job, as well as some toy projects in Elixir/Phoenix. I've played with JVM languages like Scala and Clojure, I've tried to learn Haskell about five times. Elixir had a great developer experience story (a couple people can build a nontrivial app in ~weeks). It's on par with Ruby/Rails. We went all-in around September 2016, launching in early December of 2016.
Edit: I should have mentioned, I'm the CTO of The Outline.
Was there any particular reason why you didn't choose Clojure? From what I've heard, and based on my own experience, it scratches many of the same itches that Elixir does, especially in comparison to Node/Rails.
Personally I went for Elixir instead of Clojure for a bunch of reasons, but I'm very curious to hear how others feel about this.
Dave and I started building this app together two years ago -- we both came from different ecosystems (Dave from Node, me from Rails). It's been incredibly rewarding for us to be on Elixir/Phoenix.
There's so much more to say about why Elixir is a great choice for webdev (I love Rails but probably wouldn't go back). This post gets to some of the high-level stuff we've learned -- happy to answer any questions folks have!
I feel much the same way. I might still pick Ruby/Rails or Django/Python if what I'm building is easily solved by the ecosystem, in a way that it isn't by hex.pm, but I'd probably err on the side of Elixir.
We have Elixir/Phoenix in production on Heroku (with Heroku Pipelines) using a custom buildpack [0]. It's been great. We'll probably look at Docker. We use CircleCI, and this setup script [1] was a great starting point.
I think this was shot with a DJI S800. The RC link to fly it is usually 2.4GHz. A separate link, often at 5.8GHz, is used to send the camera feed back to the ground to help the pilot or compose the shot, or both.
Good eye ID'ing the distinctive DJI arms on the hexacopter.
The convention seems to be 2.4Ghz for telemetry back the to base-station or radio display - and something lower in the UHF band like 433 or 900Mhz for actual control.
Mainly because Zigbee/802.15.4 radios are readily available in 2.4Ghz and 433Mhz goes a lot further (and through more things) by comparison at 1W.
What kind of transmitter are you using for manual control?
I assume you aren't using a ground station all the time...
I only ask because most new 'brand name' (JR, Futaba, Airtronics, Robbe) tramsitters seem to be 2.4Ghz. I know FM radios are still available on 36 and 40 Mhz, among others, but I haven't seen many used for a while with new aircraft.
UHF transmitters can be rigged on a RC transmitter as a module or through the trainer port, depending on the radio. DragonLink and EZUHF appear to work this way. (I haven't used any of these.)
It might be interesting to persist snapshots to S3. It might be usable for small projects on Heroku/etc with ephemeral storage. It could be patched here (?) [1], but clearly there'd be a lot of issues with the long network call, how often to do it, how durable it really is, etc.