Hacker News Reader

Comment by mozumder | Hacker News Reader

There's not enough open-source back-end tools that would allow a site to host and publish videos on their own, without going through a third-party service.

Where is the video equivalent to WordPress?

View on HN

Discovery is the problem. If they’re on YouTube they can show up in people’s recommendation lists, target users with particular history, etc.

Hosting your own video means you’re not likely to catch users who are binge watching videos, which I would imagine is a more lucrative demographic for ad dollars.

Isn't hardware (server space, bandwidth and latency) the real problem, with streaming video?

No, cost for that is constantly reducing. The problem is getting people to pay attention to what you make.

Constantly reducing is not the same as less expensive. Independent video streaming is still very expensive, both in hardware cost and bandwidth cost, and if you have a global audience, then you will have to rely on CDN, which for video is not cheap at all.

Youtube for the longest time, ran on loss - until recently. Only after Google had reliable penetration all over the globe using their on-premise GGC (aka google CDN) solution, which relies on the user (ISP) bearing the cost of hosting the hardware and the bandwidth that run through it, that now Youtube has the infrastructure to make it profitable.

I don't honestly see youtube being seriously challenged. Yes, there might be pockets of competitions - but far far from a serious challenge.

Unless an individual producer has very high marketable demand that he can get a premium price from advertisers or charges users per video, I don't see how hosting videos for individuals is profitable. Corporations are a different story, and even they often rely on Youtube.

A big CDN could do it, I think cloudfare recently launched a new video service. If they wanted to, they could build on that. More likely would be something like wordpress using this as a back-end. That feels like a great open-source project. Discovery could be done with RSS, trying to copy the podcast model.

Whilst its not great to replace one big company with another, it is at least good to have an alternative to keep youtube from being too crazy.

You do the same thing people do to get attention for their text blog - the usual publicity approaches.

Not sure if you’ve noticed, but the blogosphere ain’t what it once was.

Personal blogs, right. But I'm largely referring to professional media sites like magazines.

The world needs a WordPress for video.

Publishing is usually not the issue. The delivery is. Besides being it very expensive to re-encode, store and delivery content (in comparison to text), it's incredibly challenging to create good experience for the end user.

This is where companies like mux [0] come in the picture, but it's a paid product. YouTube is "free" with a chance to also generate, in some cases, pretty significant revenue stream.

[0] https://mux.com

Publishing is part of the problem. We need to be able to go from Final Cut Pro to web. It has the capability to deliver HTTP Live Streaming, but that's not supported in all browsers.

Everything from between "Final Cut Pro" and "Web Browser" needs to be standardized with easy-to-use self-publishing capabilities.

(The CDN can take care of the bandwidth)

All browsers (but Opera mini) support video [0]. There are standards all around video. The challenge is not in getting video from Final Cut Pro to a web browser, the challenge is to do it efficiently, economically with a good user experience.

CDN is the tiniest part of the whole process. Each video has to be re-encoded based on the device and network capabilities. CDN can the deliver anything anywhere. Also good video player is needed. Even with HLJ.js and Video.js it's a lot of work to put all of this together in a very user friendly way. And then you want to have cross device support and more. Video is just hard.

[0] https://caniuse.com/#feat=video

Right, but the whole process has to be optimized.

I mean, Final Cut pro can export videos to Facebook and YouTube, but to export on your own media site is where everything starts to fall apart in terms of user experience.

If the intent is to sell quality ads, advertisers need to see some audience analytics. A smaller publisher is less likely to get people to register, fill in the proper demographic data and stay logged in across all their devices than a large platform.

A site can't do that. You need a CDN and streaming transcoders, etc. Of course if you only have a few viewers then HTML5 will play your self-hosted video files but I don't you think that's what you mean.

setting the discovery issue aside, there are SaaS companies like cloudinary.com that take care of all the transcoding / delivery / device matching stuff

Reply on news.ycombinator.com