3 ms·

Most HTML is undocumented and unstable, so I’d say it’s far from perfect.

matheusmoreira 3y ago
It's perfect in that there's no way for them to differentiate from normal browser traffic. This is adversarial interoperability which is exactly how we're supposed to deal with corporations and their idiotic "terms". Nobody is forced to accept their "take it or leave it" BS. https://www.eff.org/deeplinks/2019/10/adversarial-interoperability https://www.eff.org/deeplinks/2019/10/adversarial-interopera... If it's important enough, someone somewhere will care enough to fix it when it inevitably breaks. Look at youtube-dl and its variations, there's even a youtube.js now. https://news.ycombinator.com/item?id=31021611 https://news.ycombinator.com/item?id=31021611
- great_psy 3y ago
  I would be very curious to know how many engineers inside google have as their sole responsibility to fight programs like YouTube-dl and so on. I am willing to bet it’s a whole division.
  - matheusmoreira 3y ago
    They can fight all they want. In the end it doesn't matter how many engineers they throw at the problem. The only way they can win is for the world to descend ever deeper into DRM tyranny. Google must literally own your computer in order to prevent you from copying YouTube videos. It has to come pwned straight off the factory so that Google can literally dictate that you the user are prohibited from running certain bad programs that are known to hurt their bottom line. I realize we're not too far off from that dystopia but it's still a battle that deserves to be fought for the principle of it.
- moneywoes 3y ago
  Can’t they just fingerprint the incoming requests based on a litany of variables such as headers ip etc to prevent this “scraping”? Sorry I don’t follow
  - matheusmoreira 3y ago
    Sure. We can also control those variables. Youtube-dl has a small javascript interpreter to figure out the audio and video URLs that YouTube generates for every client. In this thread it was also pointed out that people can use headless chrome. It's simple. If they allow public access to their site at all, there's pretty much nothing they can do to stop any of this.

It's perfect in that there's no way for them to differentiate from normal browser traffic. This is adversarial interoperability which is exactly how we're supposed to deal with corporations and their idiotic "terms". Nobody is forced to accept their "take it or leave it" BS. https://www.eff.org/deeplinks/2019/10/adversarial-interoperability https://www.eff.org/deeplinks/2019/10/adversarial-interopera... If it's important enough, someone somewhere will care enough to fix it when it inevitably breaks. Look at youtube-dl and its variations, there's even a youtube.js now. https://news.ycombinator.com/item?id=31021611 https://news.ycombinator.com/item?id=31021611

great_psy 3y ago
I would be very curious to know how many engineers inside google have as their sole responsibility to fight programs like YouTube-dl and so on. I am willing to bet it’s a whole division.
- matheusmoreira 3y ago
  They can fight all they want. In the end it doesn't matter how many engineers they throw at the problem. The only way they can win is for the world to descend ever deeper into DRM tyranny. Google must literally own your computer in order to prevent you from copying YouTube videos. It has to come pwned straight off the factory so that Google can literally dictate that you the user are prohibited from running certain bad programs that are known to hurt their bottom line. I realize we're not too far off from that dystopia but it's still a battle that deserves to be fought for the principle of it.
moneywoes 3y ago
Can’t they just fingerprint the incoming requests based on a litany of variables such as headers ip etc to prevent this “scraping”? Sorry I don’t follow
- matheusmoreira 3y ago
  Sure. We can also control those variables. Youtube-dl has a small javascript interpreter to figure out the audio and video URLs that YouTube generates for every client. In this thread it was also pointed out that people can use headless chrome. It's simple. If they allow public access to their site at all, there's pretty much nothing they can do to stop any of this.