> Limited use of non-approved cameras is allowable in certain circumstances (e.g. crash, POV, drone, underwater, Pan-Tilt-Zoom/Robocam etc). In all cases, such cameras must be explicitly approved by Netflix for a specific project.<p>The implications here are amazing.<p>Say someone wants to like film a mountain bike scene on a GoPro. Someone has to realize how this is going to work in advance. Then they have to tell someone to write Netflix an email to ask for permission. Someone writes that email. Someone at Netflix receives the email, and asks their team whether or not to approve the exception. A meeting is scheduled and it's discussed. They decide yes! They make a note of the approval in their project planning system. Or rather, try to... the system doesn't support that. They file a bug report. The engineering team's product manager takes a look and decides it's important, so they have a short call with the team's manager to get it prioritized. It gets prioritized. The product manager tells the Video Camera Approvals department that it will be looked at in the next couple weeks. Meanwhile an engineer picks up the ticket, decides it makes sense, and adds a new column to the ongoing_production table 'boolean unapproved_camera_exeception_granteed default false'. Some glue is put together to make this a checkbox, and in just a week, the feature is launched! The camera exceptions department is informed. They say they actually wanted it to be a list of camera models that were approved, the date that that camera model was approved, and the user who approved it. Also can you add a button to client-facing UI so that production companies can make a request there? The project manager says this is a great idea and prioritizes it. The engineer that added the checkbox is on vacation, so the more complicated version of this feature is given to the new-hire as their starter project. The new-hire puts together a small document with an overview of what they plan to do; a database migration to add a new table to store approved cameras, the UI work, a messaging system to tell the video cameras approvals team that there is a request to review, etc. A design review is scheduled for next Thursday. Don't work on it until then, we'll do the rest next sprint. Meanwhile over in the video camera exceptions approval department, the production company writes in "hey, we're running out of time, should we just use a GoPro?" The exceptions manager writes "This is approved. We're having some trouble adding a note to the CRM, but that should be fixed in the next couple months. For now, if anyone asks, this email is the record of your approval." Unfortunately, right before they press send, a cat picture is posted to Slack. Slack makes a loud sound to notify everyone of this, and the email window is closed as everyone goes over to check out the cat picture. As it turns out, this email just lives in the Drafts folder for the rest of eternity. It was the Exception Manager's last day; turns out Hulu is paying Video Camera Exceptions Managers twice as much, and in theory have better software. Meanwhile, over at the video production company, they are frustrated that they can't get ahold of anyone at Netflix to approve the camera, so they decide to just strap a Sony F65 to the rider's handlebars. There is no off-the-shelf-mount for this, so they shop around the idea to local machine shops. One offers to design and build it over the weekend so they can start using it on monday. $20,000 for the design and first prototype; $10,000 for the rush job. Approved. On Monday morning, they attach the F65 to a bicycle in the studio and pedal it around. A little wobbly, but nothing our stunt rider can't handle. They book her for a shoot tomorrow. She inspects the bike and says "whoa, this is weird" but gets used to it after a while. They head out to location and begin the technical descent. So far so good, even with the very wobbly cockpit. She hits a huge jump and get higher up in the air. The bike unexpectedly pitches downwards. The landing is not good! CRUNCH! She's off the bike. The camera flies into a ravine. The chase crew stops and throws their bikes away as quickly as possible to check on the stunt rider. She's conscious. They quickly call 911 to get a medical team out here to assist. Broken neck? Broken back? They don't know what to do. Unfortunately, there is no cell reception. Someone volunteers to hike to the summit to call. 20 minutes later, they're talking to 911. They can't get an ambulance out there, so they're sending a helicopter. Do you have the GPS coordinates? They don't, but offer to signal to the helicopter when they hear it. Someone probably has a mirror that they can signal with, right? 20 minutes pass, and the helicopter is overhead. The rest of the team doesn't know that one is coming, but someone has the bright idea of using one of their cheaper lenses as a mirror. They attract the helicopter pilot's attention. A paramedic descends through the trees, attaches a back brace to the stunt rider, and they're off to the hospital. The rest of the crew descends the slope on their bikes, reaching the production van and telling them the news. They're weeks behind schedule at this point, they might have to just cut this technical scene.<p>Some time passes. The stunt rider makes a full recovery. It looked worse than it was, and she's back in the saddle within a week. The scene is cut from the film, and it's delivered in time. The film gets amazing reviews. Subscriptions are at an all time high. The CRM software is modified to support the video camera approval workflow, and it works great. The junior engineer is promoted. A quarterly business review shows that the processes are working great, and that the scrappy Netflix is out-innovating all of the competition. High fives all around. Their share price increases and all the employees are richer.<p>This is the American economy in a nutshell, and I'll be honest -- I don't fucking get it.