The HUSH blockchain team led by Duke Leto in 2019 removed the need for full node operators to download 900+ MB of redundant privacy proving keys (zcash params). The redundancy became available because HUSH3 is a newer "sapling" Komodo (& zCash) software fork. The previous "sprout" proving keys were not needed in the blockchain record for HUSH3, therefore removed from requirements to run the chain.
Recently in the Komodo discord, a journalist who has been "deplatformed" from video sites made an enquiry on how to make the p2p streaming available to his 150,000+ audience. Not long thereafter, a new channel was created to solve the situation of a government and social media platform conspiring to silence this well known journo.
Less than two weeks since the inception of this specialized project, the Komodo dev team and an enthusiastic community member had working prototypes of the "backend" of blockchain protocol & p2p sharing for the most part sorted out. This brought some user-adoption & UX tasks to rear it's ugly head.
Reading through the developer & testing dialogues, led me to confirm the current barriers needing solutions:
Web streaming (2) is easily solvable however it becomes centrally hosted and target for DoS. An acceptable problem to not-solve at this stage in development. zCash params was a huge limitation. The everyday user cannot be expected to download 1GB of redundant privacy keys to participate in file sharing, even if it is once.
I recalled Duke Leto's feat of making a 50MB download for his team's project instead of the 1GB redundant. Bandwidth efficiency & privacy in one. Asking him how much effort is required to remove zCash params altogether, and mentioning him in the relevant parts of the chat, I woke up this morning to this PR where the efficient developer has bypassed the zCash param loading circuitry.
It's still early days in this p2p streaming application development. Less than 2 weeks since inception, and the Komodo public blockchains gain further efficiencies from ongoing community collaboration, lowering the technical requirements and setup for developer teams of the future. This new efficiency will likely make it's way into a docker container, lowering it's image size too.