3 ms·

How are you going to play YouTube videos without ads?

milankragujevic 11y ago
The server fetches the video page from YouTube, and embeds the VP8+Opus video URL into my site's page with a <video> tag.
- amadeusw 11y ago
  Your server will probably get blacklisted for heavy usage or some arbitrary API violation. Do you have a plan to remedy or circumvent it?
  - k33n 11y ago
    Does he really need to lay out a plan for that? Getting a new IP is absolutely trivial these days, and if you need to move around between subnets, that's easy too.
- milankragujevic 11y ago
  @amadeusw: I'm just planning to rebuild the VPS any time it gets blocked, if it gets blocked. Given that YouTube downloader websites have worked for so long without going down, I don't think that would be an issue.
- meritt 11y ago
  Ha. So you're stealing content and then offering it up for free (and presumably you're on the hook for bandwidth bills)? Exactly what's the goal of your ill-fated plan? You're going to get blocked/banned and when you continue to circumvent it, you're going to get a C&D. This is an absolutely terrible idea.
  - milankragujevic 11y ago
    My goal is to use YouTube on an Intel atom single core netbook without waiting minutes for the video to load in their stupid DASH player.
- KMag 11y ago
  Will that work? For some reason, I guessed the YouTube video URL was a function of the first two or three bytes of your IPv4 address, to prevent exactly this case. I did something like that back when I designed a CAPTCHA system for LimeWire. (The LimeWire website only ran PHP (soft requirement), the credit card processing box only ran Perl (hard requirement), no shared network drive between website and credit card box (hard requirement), so none of the existing Open Source CAPTCHA implementations worked. So, I made the CAPTCHA answer a cryptographic function of a shared secret, the user's /24 network, and a user-supplied timestamp rounded to a 32 or 64-second boundary. Timestamps older than 15 minutes were rejected, and messing with the timestamp would mean that the captcha answer wouldn't match the image you were given. I watched an attacker solve a CAPTCHA for an IP in Virginia and a couple seconds later try replaying it thousands of times from an IP in England. A few IP addresses for US navy satellite ground stations in Virginia (NAT for tons of deployed sailors) had to be white-listed from the replay checker.)

The server fetches the video page from YouTube, and embeds the VP8+Opus video URL into my site's page with a <video> tag.

amadeusw 11y ago
Your server will probably get blacklisted for heavy usage or some arbitrary API violation. Do you have a plan to remedy or circumvent it?
- k33n 11y ago
  Does he really need to lay out a plan for that? Getting a new IP is absolutely trivial these days, and if you need to move around between subnets, that's easy too.
milankragujevic 11y ago
@amadeusw: I'm just planning to rebuild the VPS any time it gets blocked, if it gets blocked. Given that YouTube downloader websites have worked for so long without going down, I don't think that would be an issue.
meritt 11y ago
Ha. So you're stealing content and then offering it up for free (and presumably you're on the hook for bandwidth bills)? Exactly what's the goal of your ill-fated plan? You're going to get blocked/banned and when you continue to circumvent it, you're going to get a C&D. This is an absolutely terrible idea.
- milankragujevic 11y ago
  My goal is to use YouTube on an Intel atom single core netbook without waiting minutes for the video to load in their stupid DASH player.
KMag 11y ago
Will that work? For some reason, I guessed the YouTube video URL was a function of the first two or three bytes of your IPv4 address, to prevent exactly this case. I did something like that back when I designed a CAPTCHA system for LimeWire. (The LimeWire website only ran PHP (soft requirement), the credit card processing box only ran Perl (hard requirement), no shared network drive between website and credit card box (hard requirement), so none of the existing Open Source CAPTCHA implementations worked. So, I made the CAPTCHA answer a cryptographic function of a shared secret, the user's /24 network, and a user-supplied timestamp rounded to a 32 or 64-second boundary. Timestamps older than 15 minutes were rejected, and messing with the timestamp would mean that the captcha answer wouldn't match the image you were given. I watched an attacker solve a CAPTCHA for an IP in Virginia and a couple seconds later try replaying it thousands of times from an IP in England. A few IP addresses for US navy satellite ground stations in Virginia (NAT for tons of deployed sailors) had to be white-listed from the replay checker.)