Working on it. It is HARD, that’s why so few are out there. Small scale, easy. Increase load and requirements grow exponentially and so is need to fund it. Until someone decides to invest into that and have a team work on it full time, I don’t see it happening in near future. Simple video upload, take Handbreak, encode video MP4 for web, upload to nostr.build, enjoy. Anything more than that, and we run into multitude of complexities. As an example, recent video that went semi-viral on Nostr, consumed about 1TB of data in one day, estimated to consume another 2-3. The cheapest CDN that is available will charge $5 per TB. Now do the math and multiply by number of videos and users. 🐶🐾🫡

Reply to this note

Please Login to reply.

Discussion

I see :(

And that’s the problem, ain’t it?!

YouTube when they started burnt through millions of dollars and was still in red when Google purchased them for 5B 🐶🐾🤯

I guess this is one of those businesses that just requires a lot of upfront capital. Good to know.

We can cheap away at it with gradual growth, but don’t expect miracles. We also have day jobs that pay our actual bills. So it’s all voluntary efforts 🐶🐾🫂

Could IPFS (or perhaps a similar solution) help to solve the issue?

Same problem, needs hosting and paying for bandwidth which is not cheap. Nothing is free, nothing. You either pay in time, your attention, your data or money. 🐶🐾🫡

IPFS afaik is a solution for storage and bandwidth as that is dispersed among the provider nodes. The problem I see is that it just pushes the problem further into the future. You temporarily solve the issue by delegating the cost but at some point xou cannot scale any further because there is no incentive for provider nodes to give you storage and bandwidth for free. That in reverse could be solved via filecoin (or a similar solution), which would allow people to pay with their storage and bandwidth (which they usually have more than they need or use), or money if they don't want to provide those.

There is no need for a shit coin to pay for bandwidth and storage. We can do that with Bitcoin using lightning or chaumian ecash.

Problems are…

- not enough users will pay for content i.e. there is not enough money coming in to cover the costs.

- whoever provides the hosting may be made legally accountable for the content.

"...or another solution"

- You don't have to pay with money, just provide some storage and bandwith. instead, that's ok too. Or perhaps. ertain clients may introduce ad based monetisation models like the way users are currently used to and just pay for the storage and bandwidth through that. The point is, nothing is free, but there are more ways than the ones that are currently most popular.

You can be legally liable in some countries but, if the authorities actually try to enforce that. they will have a tough time doing so. It took a lot to take down even the most blatant copyright violatiors like Openload, Streamango or Verystream that afaik were all centralized and thus relatively easy to target. What are the state thugs going to do, if there are millions of small scale hosts dispersed across the globe?

not in a streaming environment where minimal buffer times are necessary, no. s3 with minio is about the best that i have experienced. ipfs just didnt cut it.

i would like to provide some TB storage

and i would be happy to take sats per TB for that service!

Storage is cheap, so I would go with SLA based provider who can guarantee high availability 🐶🐾🫡

What if someone made a program to allow people to run their own sever nodes like with a #umbrel with a couple TB harddrive but for video. It could be uncensorable video content! Only problem I see is that it's uncensorable content..... some videos should not be available online.

Me with almost minimal knowledge speaking, but has anyone checked out peertube or lbry (minus crypto and/or add btc)?