/tech/ - Tech


Mode: Reply

Max message length: 8192


Max file size: 80.00 MB

Max files: 5


(used to delete files and postings)


Remember to follow the rules

(7.00 KB 497x111 yt.png)
youtube-dl DMCA takedown thread Comrade 10/23/2020 (Fri) 23:51:42 No. 5575
We all should have known this was coming https://github.com/github/dmca/blob/master/2020/10/2020-10-23-RIAA.md https://news.ycombinator.com/item?id=24872911 This should be an indication that all git repos need to be mirrored *somewhere* At the moment there are many mirrors on gitlab.com, including https://gitlab.com/pintert3/youtube-dl.git Last known commit according to IRC is 416da574ec0df3388f652e44f7fe71b1e3a4701f The last few commits via git log: commit 416da574ec0df3388f652e44f7fe71b1e3a4701f (HEAD, origin/master, origin/HEAD, master) Author: Sergey M․ [email protected]> Date: Fri Oct 23 21:31:37 2020 +0700 [ytsearch] Fix extraction (closes #26920) commit 48c5663c5f7dd9ecc4720f7c1522627665197939 Author: Toan Nguyen [email protected]> Date: Thu Oct 22 19:15:05 2020 +0700 [afreecatv] Fix typo (#26970) commit 7d740e7dc7149cfd93dde1fa47e9f314e72582c2 Author: Hannu Hartikainen [email protected]> Date: Mon Oct 19 17:56:23 2020 +0000 [23video] Relax _VALID_URL (#26870) commit 4eda10499e8db831167062b0e0dbc7d10d34c1f9 Author: Kevin O'Connor [email protected]> Date: Sat Oct 17 13:10:41 2020 -0400 [utils] Don't attempt to coerce JS strings to numbers in js_to_json (#26851) The current logic in `js_to_json` tries to rewrite octal/hex numbers to decimal. However, when the logic actually happens the `"` or `'` have already been trimmed off. This causes what were originally strings, that happen to look like octal/hex numbers, to get rewritten to decimal and returned as a number rather than a string. In practive something like: ```js { "0x40": "foo", "040": "bar", } ``` would get rewritten as: ```json { 64: "foo", 32: "bar } ``` This is problematic since this isn't valid JSON as you cannot have non-string keys.
What is even the DMCA claim being entertained by these fucking cretins? Any youtube video can be downloaded with a little tinkering through a page's source, youtube-dl simple does that for you.
Reference 1 is present in the notice, but not reference 2. Where is reference 2, "the attached court decision from the Hamburg Regional Court"?
(104.46 KB 913x490 memri riaa.jpg)
>>5577 >>5580 as far as I understand the claim is: "youtube (music?) licenses only allow streaming, not downloading, therefore this software is for violating copyrights/license agreements" the Landgericht Hamburg 19.07.2017, 308 O 230/17 thing they cite was apparently a ruling about how one webservice for extracting/dl-ing audio stream from YT vids was cracking a fuctional copy protection, because youtube hides the address of the stream somewhat. that court is notorious in german IT circles for being anti-user since forever does the claim even matter, though? they're prob. just salty that this allows people to circumvent ads
>>5585 Maybe someone should explain to them how streaming actually works.
Thats what you get for still using Github tbh
>>5585 Please state the full name of that "webservice for extracting/dl-ing audio stream from YT vids" because my search results are flooded with the youtube v GEMA saga, wich does not apply here.
>>5589 I don't know the name. Here is the ruling: http://www.landesrecht-hamburg.de/jportal/portal/page/bsharprod.psml?showdoccase=1&doc.id=JURE180006255&st=ent Schutzgesetzverletzung: Umgehung von wirksamen technischen Maßnahmen auf Streaming-Portalen zum Schutz urheberrechtlich geschützter Musikwerke durch Konvertierungssoftware they only say: "ist jedenfalls seit dem 16.06.2017 im Impressum der unter der URL www. c..net abrufbaren Website verzeichnet" just "c..net" which is obviously a reference to a now forbidden domain. If you want the name you will have to contact this court.
>>5587 sadly, reality and technology don't interact much with law when this court is involved. material reality just doesn't matter to them(and political results prove them right), I've been following this for years now
>>5590 >Here is the ruling Thank you, friend. May the gods double your yak herd.
Here is the Hamburg Regional Court decision via google translate. http://0x0.st/ikbr.txt All the web service stuff is irrelevant here and all the OP DMCA references is the circumvention of the effective technical measure. The claimant's argument for "effective technical measure" is in [16], the defendant's counterargument is in [21], and the court's decision is in [34] through [42]. All of the defendant's points are conceded and then dismissed as irrelevant, and the only point that causes the court to rule that the technical measure is effective is that the court considers that copying the correct URL out of the network tab of dev tools is too much for the average user. There is virtually zero chance that a US court would rule differently because the US legal standard for the average person is known as "an idiot in a hurry".
Did Invidious Snopyta recently become unreachable for anyone else but me? http://c7hqkpkpemu6e7emz5b4vyz7idjgdvgaaa3dyimmeojqbgpea3xqjoid.onion/ It just gives me "502 Bad Gateway" on several circuit renewals
there is no way to make the code disappear, there are plenty of mirrors and forks out there. they can easily host it themselves, preferably in russia - similar to scihub. the real question is how this will influence the development of ytdl. When it inevitably breaks in a couple of months, will there be an update? If devs get sued personally, that would reduce the amount of people willing to commit code. also, ytdl is used by other programs, like mpv for example. Will they feel pressured to remove easy integration of ytdl into their program, thus reducing the usefulness of ytdl itself? I use it a lot more to stream youtube content without opening their dogshit website than actually saving content from there, and i would hate to go back to that - especially since mpv uses a lot less bandwith and resources to view a yt vid
>>5619 It works intermittently on the clearnet for me with the occasional 500 Internal Server Error, which is fairly common for invidious instances under heavy load. I'm not seeing the 502s.
>>5629 I'm the anon that posted what you replied to; fortunately it went back to "normal" (Bad Gateway "only" once every ~10 circuit-renewals) the day after. Can access it now too. (This really is an incredible resource, I hope this becomes a TPB kind of thing, with mirrors always reappearing). We also should shill it a bit honestly, I was pretty surprised no one in /ITG/ or /leftytrash/ being immediately familiar with "Invidious" when I mentioned the downtime (except one based Hoxhaist Posadist poster in the latter thread).
>>5630 Right now invidiou.site seems to be more responsive than invidious.snopyta.org.
Funniest shit is retards STILL using github after this to host ytdl forks.
is there any way I can still download playlists? I need to download an album.
>>5716 Just download it from their website. https://youtube-dl.org/
>>5575 i wouldn't be surprised if this ((DMCA)) is just a cover-up IIRC ytdl can also grab other sites maybe they are trying to prevent others from migrating to other platforms
>>6380 Nice PR move from M$.
(5.17 MB 1500x1853 image.jpg)
'''The 2017 decision of the Hamburg Regional Court in Germany that RIAA references, which refers to YouTube’s “signature” mechanism, was wrongly decided and is not binding nor even persuasive under U.S. law. The court in that case apparently reasoned that since the judge was not familiar with JavaScript, using the “signature” code was beyond the capabilities of the average user. It was on this basis that the court declared the code to be an effective technical measure under Germany’s analogue of Section 1201. The court’s analysis overlooks the ubiquity of JavaScript, which is embedded in every browser and similar software, making use of the “signature” mechanism well within the capabilities of the average user. The Hamburg court’s analysis sweeps too broadly: it would cause anti- circumvention law to apply to any web content except the simplest plain-text pages, because all such content can appear obscure to the average user in source-code form but is easily read and used in a browser. The Hamburg court’s decision is not consistent with the U.S. DMCA and would not be followed by a U.S. court.''' While I'm all for restoring yt-dl and taking the MAFIAA down a notch, I was disappointed to see that the EFF made no attempt to address how their "ubiquity of JavaScript" argument allows the average user to use the browser and the js code provided by YouTube to save a permanent copy, as opposed to ephemeral playback. Hamburg said >>5593 that the saving specifically is beyond the average user. '''As federal appeals court recently ruled, one does not “circumvent” an access control by using a publicly available password. Digital Drilling Data Systems, L.L.C. v. Petrolink Services, 965 F.3d 365, 372 (5th Cir. 2020). Circumvention is limited to actions that “descramble, decrypt, avoid, bypass, remove, deactivate or impair a technological measure,” without the authority of the copyright owner. “What is missing from this statutory definition is any reference to ‘use’ of a technological measure without the authority of the copyright owner.” Egilman v. Keller & Heckman, LLP., 401 F. Supp. 2d 105, 113 (D.D.C. 2005). Because youtube-dl simply uses the “signature” code provided by YouTube in the same manner as any browser, rather than bypassing or avoiding it, it does not circumvent''' This part was great. I did not know that for the US DMCA 'use' does not qualify as 'circumvention'. This seems to void the RIAA's claim. My worry here is that this was in federal appeals court, and if it goes to the supreme court in the future, it will be reversed. The current 6-3 supreme court is sure to rule against the freedom of the users, and for the "freedom" of the rights-holders to impose restrictions on the users.
What's stopping youtube from using widevine?
>>6389 nice ass


no cookies?