lemm.ee

1 point

1 year ago

I pay a very small fee for debrid and VPN servers that offer exactly the same server capacity with enormous bandwidth and virtually no downtime.

Did you just compare your small private server with YouTube’s infrastructure? Jesus Christ.

Google had already been paying about 2 million a month for bandwidth in 2015 or so.

I work for a larger company as a software developer, even with a billion in gross sales, there is absolutely no chance to provide even a tenth of YouTube’s service. Especially for free (without paywalls). The company would go bust in two years.

report

[ - ]

Excrubulent@slrpnk.net

1 point

1 year ago

I didn’t, I compared globe-spanning networks of servers that serve millions of people every day to youtube. Those two things don’t seem that different to me. They scale with user numbers just fine.

I mean you work for a larger company as a software developer, and you don’t understand the concept of debrids and VPNs? Are you sure you’re not deliberately missing the point of what I’m saying?

report

[ - ]

1 point

1 year ago

VPN has absolutely nothing to do with hosting a video platform, no clue why you even bring it up.

Debrids is just a file download service, isn’t it? But even if it was a video hosting platform, a single server would never be enough. You need at least two (as a fall back). Then you need dynamic scaling for bigger user numbers, which works just fine for CPU and RAM (or even GPU resources), but doesn’t work for storage. So you need extra storage somewhere all servers have access to, but when it comes to videos you’d be paying millions in no time.

So you need your own cheap storage and datacenters around the world. And CDNs on top to serve your content worldwide (otherwise the experience would suck on another continent if your server is too far away).

Look up how Google does it, they have their own data storage centers. And if your video is crappy and you’re a nobody, it probably gets stored in a slower location on-demand. So it also loads slower. But if your video is in high-demand with millions of views it gets pushed into a more accessible location (and gets higher priority for CDNs). It’s not just hosting, there is a massive amount of logic and software behind the stack.

report

[ - ]

Excrubulent@slrpnk.net

1 point

1 year ago

You have demonstrated a complete inability to grasp what a VPN does, what a debrid service does, that they already do the things you’ve mentioned, and you have yet to acknowledge peertube even exists. I brought it up, multiple times, for a reason.

I have to ask at this point, are you curious to understand my position? I don’t see much point in continuing to explain it to you if you’re not.

I am struggling to understand yours. There doesn’t seem to be a coherent idea that you’re driving towards other than to tell me I’m wrong, which isn’t a position as much as an antiposition. If you have a position, I would appreciate you explaining it clearly.

report

[ - ]

1 point

1 year ago

You use a VPN when you either don’t trust your ISP (or the current network connection you are on) or you want to hide who you really are on the internet. Both are absolutely unnecessary when accessing a video hosting platform (you can do this, but you don’t have to). A VPN is also more on the user side of things to connect to a server, the server doesn’t care if you use a VPN.

Debrid just makes accessing files easier as far as I can see. Like you give it a torrent link and it provides you a direct download? That’s nice and all for piracy, but has absolutely nothing to do with a video hosting platform like YouTube. You could use Debrid to download the video file from a host, but we are talking about providing the actual host you store the videos on.

I absolutely do not get the points you are trying to make, do you have an example for an infrastructure like YouTube you could build out of a VPN and Debrid?

Peertube would be an alternative of course, but it obviously has tons of its own issues (mainly resources, it still costs too much to host a large instance and if you try to access one video a million times things would straight up implode). I don’t see a realistic YouTube alternative without investing millions.

report

[ - ]

Excrubulent@slrpnk.net

1 point

1 year ago

I am not saying a VPN or a debrid are necessary, only that they demonstrate the bandwidth and storage capability at scale for low cost, on which peertube could run, which would presumably scale with interest in the platform. It’s not complicated.

I won’t explain any further unless you tell me, specifically, that you are curious to understand what I am saying.

report

[ - ]

1 point

1 year ago

Now listen, Debrid isn’t actually providing any meaningful bandwidth. It’s a third party (fourth party?) service.

What they are doing is simple: For their paying users (no clue what it costs without making an account, $3 a month?) they offer fast direct downloads. But they aren’t even storing the data themselves (besides caching)! They use premium accounts for other file hosters to get around the download throttling. So instead of you being limited to 1 MB/s or less for most downloads Debrid uses their account to download at full speed, then give you the file.

So they are pretty much abusing other hosters by allowing their own users to share a premium account for various file hosting platforms. Which will work so long until these hosters start aggressively blocking accounts that use too much bandwidth.

In addition to that you are paying Debrid money, $4 or something a month? If every YouTube user even paid $1 a month there would be zero need for ads. You are right, bandwidth is relatively cheap, but getting people to pay is difficult. Your suggestion would basically be that YouTube now forces everyone to pay $2 a month or they can’t access the service (or only 480p videos or whatever), which would work! But is far less suitable than charging more for no ads and have only one out of hundred(?) users pay while the rest happily watches ads.

If every user threw in some coins per month we could have services with zero ads. But even a cheap subscription like $1 or $2 is often too much to convert users. The service has to be free, so that out of a million users maybe a few hundred actually pay.

report