We all love watching videos online. We spend hours everyday watching funny Youtube videos, weekend-binging Netflix shows, late-night streaming on Twitch. Facebook recently claimed video will be the dominant content format on the internet by 2020. But with all of its recent development in video content, video software still remains costly and opaque.
Proprietary Software
Video started out as a feature in the browser, but quickly became much more. Because of the slow internet speed in the late 90s, videos on forums often took hours to download. Engineers soon figured out a new mechanism called “video streaming”, where users can download small chunks of the video and start viewing right-away. Due to the complexity of video codecs in the offline world, many video formats have been created. (flv, hls, dash, etc) This growing complexity gave birth to media servers — a piece of software that hides all the complexity and enables video streaming out-of-the-box. But most of the media servers in the market today remains close-sourced and slow-moving.
Costly Operation
The online video industry is also notorious for its high cost. Google famously spent $470 million running Youtube infrastructure in 2009. The standard CDN and cloud streaming solutions range from $3.00/hr/stream to $0.30/user/hr (this comes out to millions of dollars per month for an application with a few million users). With the new wave of technologies like 4k video, low-latency interactive video, and VR, the cost will get higher and higher. Applications have no option but to pay this high price due to the lack of open software, even though the cost of computing has lowered dramatically in the last decade.
Lack of Incentive
There has never been the correct incentive to create a high-quality, open source media server. Most open source projects stop at the university research level, when the research group eventually runs out of funding for the project. This creates a culture of proprietary software, where vendors re-implement the same code and use software licenses to lock in clients. The few open source media servers out there were never able to sign up big clients to support its development, and they remain dormant or lacking in feature parity.
Blockchain-Enabled Open Source
In the past few years, open source software itself is quietly going through a re-birth with the help of the blockchain. Rather than the traditional Redhat model, where the open source developer sells service contracts to large clients, the blockchain issues tokens to all ecosystem participants including the core developers. As the software becomes more widely adapted through the blockchain network, demand for the tokens increase, possibly leading to appreciation, which fuels the development of the software. Open source software no longer needs to survive on service contracts that required the software to be hard to use, just so the service company can make money. The best part of this, is that you don’t even need to care about the blockchain to benefit. We already see this happening with projects like IPFS and Zcash, who created libp2p and libsnark to benefit ANY peer-to-peer application. In this new model, the software creator is incentivized solely to make the software more useful.
Livepeer Media Server
With this new model, we have created the Livepeer Media Server (LPMS). With LPMS, we want to make building video applications as easy as downloading a piece of software. We have a Golang implementation that supports simple video live streaming right now, but we are committed to make it the best media server with all the features needed to build any video application. Livepeer intends to make LPMS available completely for free, while supporting this development by executing on the above token model rather than through selling consulting services.
Btw, we are looking for 1 more person to join our team. If you are excited to build the future of open source online video, get in touch here!
For further updates on Livepeer, you should follow @LivepeerOrg on twitter and join the community chat.
