Heads-up for the #PeerTube community, it appears Google has been blocking many tools/IP ranges that try to synchronize with YouTube. At least for the creators on #tilvids this is a massive challenge, because they can no longer sync THEIR content on multiple platforms, requiring them to manually upload to both TILvids AND #youtube We appreciate our creators going the extra mile, and please remember this is why it's so important to support alternative platforms.
https://github.com/Chocobozzz/PeerTube/issues/6522
#tech
It still lasts because there’s no easy way YT can offer their own content without the video being available as a file stream (through CDNs at googlevideos subdomains). If they centralize everything to a single, controlled domain (so to allow things as one-time HTTPS request, better session checking and so on), they’d lost the capability of load balancing allowed by the decentralized nature of CDNs. YouTube downloaders (and, by extension, third-party YT frontends such as Invidious) exploit this CDN aspect to download the videos.
It’s common to see Invidious instances momentarily blocked. The blockage can’t last forever for two reasons: firstly, IPs (especially IPv4) changes due to how ISPs offer IPv4 addresses through CGNAT, so the instance IPv4 (generally domestic servers) will eventually change (often to a completely different IPv4 range) and YouTube won’t know that the new IP is a former “offender”. Secondly, as IPv4s works through CGNAT, Google can’t keep the bans forever because this IPv4 will be eventually rotated to another client from ISP that’s completely unrelated and unaware of how their IPv4 was a former address for a downloader. It’s like how Signal/WhatsApp/Telegram/Facebook/phone-required services can’t really keep a permanent ban for a specific prepaid number (especially on countries like Brazil, where ANATEL allows for phone number rotation when the mobile plan is cancelled), because the number will be potentially owned by another person with nothing to do with the former owner.
So, in summary, Google can either end with YouTube CDNs (ditching their load balancing), or they can try to implement an innovative way to keep load balancing while serving the request one-time only, or they won’t be able to do nothing but to perpetually catch themselves drying ice cubes.
Not the OP, and I don’t actually know, but paid streaming services differ from YouTube in that everyone who accesses the content is paying for the service. On one hand, you can validate that everytime a video is served, it’s served to a paying user. On the other, you are receiving revenue directly from consumers to fund the infrastructure to store and serve the videos.
YouTube, on the other hand, stores significantly more content, for free, and can be accessed for free, without being signed in.
The “without being signed in” part of YouTube is now no longer completely true. I tried to watch a video tutorial at work the other day and it wouldn’t play because I wasn’t signed in and so “they couldn’t be sure I wasn’t a bot”. I’m not signing into any personal stuff on my work computer, or wasting time creating a “work” Google account, so I guess YT can no longer be a place where I can get helpful programming info.
I guess they have no decentralized CDNs as YouTube does, but… paid streaming services still have their weaknesses (there certainly are tools that fetches content from there because of, e.g: entire Netflix movies/series became torrents without screen recording).
It still lasts because there’s no easy way YT can offer their own content without the video being available as a file stream (through CDNs at googlevideos subdomains). If they centralize everything to a single, controlled domain (so to allow things as one-time HTTPS request, better session checking and so on), they’d lost the capability of load balancing allowed by the decentralized nature of CDNs. YouTube downloaders (and, by extension, third-party YT frontends such as Invidious) exploit this CDN aspect to download the videos.
It’s common to see Invidious instances momentarily blocked. The blockage can’t last forever for two reasons: firstly, IPs (especially IPv4) changes due to how ISPs offer IPv4 addresses through CGNAT, so the instance IPv4 (generally domestic servers) will eventually change (often to a completely different IPv4 range) and YouTube won’t know that the new IP is a former “offender”. Secondly, as IPv4s works through CGNAT, Google can’t keep the bans forever because this IPv4 will be eventually rotated to another client from ISP that’s completely unrelated and unaware of how their IPv4 was a former address for a downloader. It’s like how Signal/WhatsApp/Telegram/Facebook/phone-required services can’t really keep a permanent ban for a specific prepaid number (especially on countries like Brazil, where ANATEL allows for phone number rotation when the mobile plan is cancelled), because the number will be potentially owned by another person with nothing to do with the former owner.
So, in summary, Google can either end with YouTube CDNs (ditching their load balancing), or they can try to implement an innovative way to keep load balancing while serving the request one-time only, or they won’t be able to do nothing but to perpetually catch themselves drying ice cubes.
Maybe a stupid question, but how do paid streaming services avoid that issue?
Not the OP, and I don’t actually know, but paid streaming services differ from YouTube in that everyone who accesses the content is paying for the service. On one hand, you can validate that everytime a video is served, it’s served to a paying user. On the other, you are receiving revenue directly from consumers to fund the infrastructure to store and serve the videos.
YouTube, on the other hand, stores significantly more content, for free, and can be accessed for free, without being signed in.
Removed by mod
The “without being signed in” part of YouTube is now no longer completely true. I tried to watch a video tutorial at work the other day and it wouldn’t play because I wasn’t signed in and so “they couldn’t be sure I wasn’t a bot”. I’m not signing into any personal stuff on my work computer, or wasting time creating a “work” Google account, so I guess YT can no longer be a place where I can get helpful programming info.
That might have been because of your IT team. You can absolutely watch YouTube videos without being signed in. I do it all the time.
YT likely recognized the IP as belonging to a commercial entity and threw up the block, due to bot concerns, over that .
Possibly yeah
I guess they have no decentralized CDNs as YouTube does, but… paid streaming services still have their weaknesses (there certainly are tools that fetches content from there because of, e.g: entire Netflix movies/series became torrents without screen recording).