Newpipe/yt-dlp stops working
github.com
github.com
(I'm a Premium user and even though I'm already ad-free, I do have legitimate uses for downloading DRM-free videos for my own personal use.)
The video is decrypted such that your OS is prevented from ever seeing its contents, it goes straight to the GPU. If that isn't possible, the video won't play at all (or it's going to play in a less secure mode and hence at a lower resolution).
The best you can hope for is playing the video and capturing it with a physical HDMI capture card, though that loophole is eventually also going to go away as newer HDMI standards are released and better DRM gets enforced.
In that case the true analog loophole will still be possible with a camera pointed at a monitor inside a light-proof box. If it comes to that, I will build such a device. AI can probably fix degraded video quality pretty nicely anyway.
This definitely won't happen in the next 5 years, but the next 30? I wouldn't be so sure.
This actually applies a lot less to Youtube than it does to any other service.
On other video streaming platforms, think Netflix, Max, Disney Plus etc, you have a fairly limited number of shows, each with a fairly large number of views. This means that if you break DRM, you can share your tool with a few trusted co-conspirators and only release the videos you rip, but not the tools themselves. This protects you from the platforms discovering and breaking your method.
Youtube doesn't really work like that, a lot of people who use yt-dlp and other such tools care about videos with relatively fewer views, which are unlikely to be available anywhere else. This would require any potential DRM-breakers to share their method publicly, which would make them extremely short-lived.
Good DRM algorithms nowadays aren't just some obfuscated code in a binary that you have to reverse engineer, they rely heavily on some content being decrypted in hardware. Even if you understand every single instruction in the binary, re-create a perfect open source version of it and have full read/write privileges to any part of the memory that your OS kernel can access, it's still not enough.
I doubt enough people will reach that level of fed up for it to hurt Google's profits, but that doesn't matter to me. I was happy enough to use youtube and give google my data on my own terms. If a workaround to this issue can't be found I'll be disappointed by Google's decision to cut off third party applications, but they won't be successful in trying to force me to accept their terms.
I don't think they're really that bothered about your data. They want to show you ads or sell you yt premium.
I have a strict policy against paying people to remove the problems they themselves created for me (AKA racketeering) so I was never going to be a youtube premium customer, but there are plenty of other opportunities to make money from my data, including through ads, even without those ads I won't tolerate on youtube. I'm not Google's preferred user, but that doesn't mean they can't still make money at my expense.
For people who just follow subscriptions, not the discovery page, Youtube doesn't really have a moat. PeerTube integrated into Mastodon is a borderline existential threat.
Sorry even as a big fan of PeerTube and Mastodon this is simply delusional.
Sounds bizarre, i know. But i get _some_ value out of YT, but not the price of the sub. Worse even, is that as a paying customer i still feel like their UI is hostile and i need to modulate my behavior to fight their algoritm.
God forbid i click on a video that i don't want to see 5,000 times on my home page.
I've sat in on meetings with marketing teams ordering hundreds of thousands of dollars of custom development work for ad campaigns with research entirely based on what I can only describe as "vibes".
They will hire psychologists to tell them that flashing the color yellow while playing a tone in C-flat increases the viewer's desire to spend money on floral scents and no one in the room will question it.
There are reasonable advertisements that would not annoy consumers, but those aren't what keep an industry of middle managers and scam artists alive.
Sadly some of those psychologists aren't lying or scamming. There are all kinds of weaknesses and vulnerabilities in the human brain (some universal and some cultural) that can be exploited, at least often enough for it to be worth doing.
There's an entire dark psychology industry around it and it involves human experimentation, including experiments on infants and children (https://www.nbcnews.com/health/health-news/even-3-year-old-k...), research into centuries of real world data on what tends to work well for scammers, fraudsters, preachers, cult leaders, and politicians, and the cutting edge of neuroscience.
I mean, yeah, some of it really is bullshit with very shitty science behind it, but it's easy (and tempting) to dismiss how effective and manipulative advertising is. We all want to think we're smart enough to see through it and are therefore immune, but like many optical illusions even when you know what's really going on your brain can't help but accept the lies. We're full of blind spots, cognitive biases, and hubris which is why such insane amounts of cash get pumped into ads and ad research. To an uncomfortable extent, it works.
We can arm ourselves with knowledge and critical thinking skills, but the most important thing you can do to protect yourself is to just avoid advertising wherever and whenever you can.
If they do end up doing it, I'd love to write this, it would be a lot of fun.
Otherwise you could probably fingerprint ad spots and submit them into a shared database. Like musicbrainz. For dynamic blocking of static content. Acoustic ID is probably enough and it would work with podcasts too!
I hate ads so much, I would totally volunteer to make this my hobby.
Advertisement simply sucks and is offensive. There is nothing to respect about any of it. People will always find ways around.
You can do this with ffmpeg.
No. Video files on streaming services like YouTube are served as a bunch of chunks listed in a text playlist. So all they need to do is just to add a bunch of lines in the middle of a playlist.
video_part0076.ts
video_part0077.ts
video_part0078.ts
ad_part0001.ts
ad_part0002.ts
video_part0079.ts
video_part0080.tsI guess they could prevent that if they randomize the part names (with some m:1 mapping on the server side). As long as you have control of your CDN, this would even not mess with caching.
Since the YouTube marks ads, I guess pulling the playlist twice won't really be needed. YouTube's player should somehow know when the ad starts and when it ends after all.
My friend asked me to write a userscript that simulates pressing the "Skip ad" button, so he won't stain his mouse with grease while eating.
Checking if there is a "Skip ad" button every quarter of a second and invoking its click() method eliminated the video ads completely. As it turns out, the button existed alongside the 5 second countdown, just wasn't visible.
I don't know if it still works the same way today, but it felt good to accidentally make an effective adblock with five lines of code.
AV1 encoded >fullHD sources were missing for a couple days recently, only VP9. Maybe Google was switching encoding hardware? preparing for encoding ads directly into sources?
N parameter function was rewritten - this broke yt-dlp. player b22ef6e7 https://github.com/yt-dlp/yt-dlp/commit/297b0a379282a15c80d8...
source 22 (h264 720p combined video+audio), the only HD source older clients could play, was killed a ~week ago. Now older clients are only able to play 360p (source 18) IF they can correctly decode n parameter
Doing so could ensure long-term compatibility and equal access.
1) --cookies-from-browser is actually doing something for me since I'm authed
2) just lucky
I'm not saying I like the idea of YouTube DRM. This is a technical (maybe organizational) question of whether there's something stopping Google from using Widevine for YouTube. Google is going through a ton of effort to frustrate tools like Invidious and yt-dlp when it seems (to me) that Widevine would accomplish their goal without as much effort (but there might be stuff I'm unaware of).
it drives me crazy when they create these rules that are very difficult for newbies to follow, then delete without comment when the rules aren't followed perfectly. It just discourages users from participating in the bug reporting process
"to organize the world's information and make it universally accessible and useful."
Yeah right.
[0] https://support.google.com/interconnect/answer/9058809?hl=en
For your second point, most people who have phones can access YouTube, devices are not as weak as you think.
If you went to a lemonade stand that said "Free Lemonade" and they had a guy standing next to it who drew a picture of you picking up a lemonade, that wouldn't make the lemonade not free. If you decided that lemonade was not free because someone drew a picture of you, then sure, but you can walk that all the way to having no things that are free.