Is there a GitHub repository for this? There are non-obvious features that I think it would be very good for it to have.<p>One of the core problems is that reporters might have their phones confiscated or destroyed, or the reporters might even be killed. The best solution is live uploading, but network infrastructure usually isn't good enough to stream video (especially in crowds where several people might be trying to do it at once). What I think it should do, is use the connectivity it has to live-upload limited information strategically.<p>First it should upload the location, time, device ID, and a hash of the video recorded so far. This is only a few hundred bytes (could fit in an SMS), and creates a record that video was recorded, which could be used in combination with testimony to support a claim of destruction of evidence. Then it should add, based on the amount of bandwidth available, low-quality audio, higher-quality audio, and selected still frames from the video.