Preservetube - A Youtube archival site.

  • 🏰 The Fediverse is up. If you know, you know.
  • Want to keep track of this thread?
    Accounts can bookmark posts, watch threads for updates, and jump back to where you stopped reading.
    Create account
How much storage is that using right now?
There is definitely a hard cap on video size, maybe 150-200 MB? The average video size must be somewhere between 10 MB and 100 MB. At 50 MB, a million of them would be 50 terabytes. Then there should be some duplication of these across multiple servers.

PTube's giving an error 503 so maybe the real number is zero.
 
Last edited:
preservetube_metadata.png
Youtube changed the API again
 
Youtube changed the API again
It's been unusable for 2-3 days*. Fighting YouTube is a full-time job. I think I've seen ~3 GrayJay YouTube plugin updates in the last week too, and they don't even try to let you download videos without using a Google account.

*Well, apparently not unusable for somebody because the latest page has videos from today.
 
How much storage is that using right now?
~50TB, with ~12TB of available space.
I'm getting intermittent successes now, enough to make it usable. Also, download speeds are very fast, possibly as a result of necessary rate limits.
I didn't add any ratelimits, just changed how the videos are downloaded. I moved to SABR a few months (?) back, but moved back to the good ol HTTP. Seems to work better, for now.

All the code Preservetube runs is -- to some extent -- public. No need to guess about the changes made.

I also changed the limit for 480p to 25 minutes. Might change this in the future, but we'll see.
 
@PreserveTube What's up with the "page rendered by frank" and "page rendered by paris" messages that show up while archiving a video on the site?

Are there any plans to get the onion version of the site up and running again?
 
What's up with the "page rendered by frank" and "page rendered by paris" messages that show up while archiving a video on the site?
Those are the hostnames of the server you're connecting to. Helps with debugging and the such.

Are there any plans to get the onion version of the site up and running again?
Fixed. Forgot to change reverse proxy config for it.
 
Those are the hostnames of the server you're connecting to. Helps with debugging and the such.


Fixed. Forgot to change reverse proxy config for it.
Also, when I try using the site through the Tor browser on Android I end up getting the error message: "Websocket connection was closed: (1006)" after archiving some videos. After I restart the browser I can archive a few more until it shows that error message again. Hopefully you can replicate the issue on your end.
 
Those are the hostnames of the server you're connecting to. Helps with debugging and the such.


Fixed. Forgot to change reverse proxy config for it.
when is preservetube.net coming back?

Also, when I try using the site through the Tor browser on Android I end up getting the error message: "Websocket connection was closed: (1006)" after archiving some videos. After I restart the browser I can archive a few more until it shows that error message again. Hopefully you can replicate the issue on your end.
I thought you couldn't archive videos on any other domain that isn't .com?
 
Also, when I try using the site through the Tor browser on Android I end up getting the error message:
The websocket still connects via the clear web, and thus most likely Cloudflare is flagging the Tor request. I need to overhaul that whole thing and make it not dependent on Cloudflare, but it's not really at the top of prorities.

I'm also wishing to move away from Cloudflare Turnstile, so I'll have to see what Josh moves to for the forum captcha.

when is preservetube.net coming back?
It works for me.
 
suggestion:

if a video can't be archived for whatever reason, the site checks if ghost archive has it or not. Say https://www.youtube.com/watch?v=hWTFG3J1CP8, if preservetube craps itself and can't get past YouTube to download it, it can go and check ghost archive, specifically it's CDN, which uses can be easily done as all that needs to remembered is "hWTFG3J1CP8" of it, and you'd just need to check https://ghostvideo.b-cdn.net/chimurai/*.mp4/ and just download it, from what I can tell there is no anti-botting so it would be possible.
 
suggestion:
I'd like for Preservetube archives to come with a certain level of trust. I have no idea who Ghostarchive is ran by and I can't 100% confirm that they, in fact, have the intact version of the video.

In some ideal universe, I'd maybe link to them if people search for videos that aren't archived (like archive.today does), but we shall see.

I'm not running low on money, I'm running low on time. Real life duties come first :)

I also want to move away from Minio to Garage, which I'd like to spend some time on to make sure it's polished before pushing to prod.
 
Back
Top Bottom