On 8/31/2023 9:03 AM, Newyana2 wrote:
>| And, I didn't even know there was a "web site" called
youtube-dl.com (or
>| whatever the domain extension was); so I suspect this is a copycat site.
>|
>
> It's a legit site that had been operating but outdated for
> a long time. I don't know about the case. Maybe it's a case
> from 3 years ago that finally got a ruling. Maybe the RIAA are
> going after whoever they can, as a PR move. Microsoft now
> owns github, so that connection won't be so easy to block.
I didn't know M$ owns GitHub, which, well, maybe that's not a good thing.
But I do agree with you that the RIAA cares more about PR than about law.
The more people the RIAA can scare, the more they can claim success.
> I never got yt-dlp working properly. I had to keep restarting
> it. Recently I found 3dyd, thanks to someone in these groups,
> and it even works on my XP box. Now, that's software! :)
Thank you for suggesting 3dyd, which I had never heard of until now.
https://www.3dyd.com/
https://yd.3dyd.com/download/
It seems to be updated frequently and even seems to have a batch option.
https://www.videohelp.com/software?d=3dyd_1.19.1.exe
https://www.videohelp.com/software?d=3dyd_batch_2.11.exe
https://download.yd.3dyd.com/3dyd64_1.20.exe
Name: 3dyd64_1.20.exe
Size: 20226239 bytes (19 MiB)
SHA256: 05ADEEED0106AF971275698C27E5A4DB20585380047D11613F5C875E4C1CC2FC
https://download.yd.3dyd.com/3dyd32_1.20.exe
Name: 3dyd32_1.20.exe
Size: 18335077 bytes (17 MiB)
SHA256: F1A4659C0F1A5E5C52AAE574222E3DA596BF8F051E1781AC3BEFB4ED8DB447CA
> 3dyd seems to fail with at least some music, so maybe they're
> trying to work with the RIAA. Which is fine with me.
As you noted, all of the youtube downloaders suffer when Google does
"something" that breaks them, whether it's ClipGrab/yt-dlp or NewPipe.
I found those GUIs worked well but they must be updated often.
I suspect Google, who has all the source code, makes minor changes on
purpose to slow down their efforts - even as they all use public APIs.
> I'm not
> generally interested in Taylor Swift's latest bubble-gum-pop.
> I'm usually looking for academic lectures, car repair videos, and
> the like. On the other hand, why would the RIAA go after youtube
> downloaders, rather than Google? Why do they allow music on
> a publicly accesible website at all? Why not issue takedown
> orders to Google?
Like you, I download documentaries and repair videos, both of which I can
watch when I'm far away from my home Wi-Fi or waiting in a waiting room.
I think the RIAA just wants to get in the news to scare people who don't
understand that all these downloaders use the API that any web site does.
> It's possible that the RIAA are in bed with Google,
> on a PR campaign to do battle against basic http protocols.
I don't know if Google is "in bed" with the RIAA but Apple certainly is.
Which is why every operating system (other than iOS) allows these GUIs.
> There's a general move to turn the Internet into a subscription
> service for interactive cable TV and services.
Yup. You're on the ball. Subscriptions equate to easily forcasted profits.
I, for one (probably like you) don't have a single Internet subscription.
And yet I can rather easily get anything I want that is on the Internet.
So that's my personal proof that subscriptions aren't needed.
For example, can you think of anything that you get by subscription on
YouTube that you can't get without that subscription - if you're smart?
> That's at odds with
> the original design and intention of the Internet as an open,
> transparent communications medium. If the Hamburg ruling is
> legit it implies that their courts have been duped into believing
> that "difficult access" to a webpage or online file constitutes
> a purchase contract. That could clear the way for news sites,
> for example, who claim no one has a right to load their webpages
> without also being sent to a half dozen ad/spying domains and allowing
> the news website to control their browser via script.
I must admit that recently, they've added the Epic Privacy Browser proxies
(which Epic calls a VPN) to the cloudflare "wait a minute" domain blocks.
So, one by one, little by little, domain by domain, I have to agree with
you that the Internet is becoming more & more a "difficult access" for us.
Still... with judicious use of thousands of free openvpn servers, most
sites can be "unblocked" from the cloudflare blockers - albeit with effort.
> As it stands now, many of the popular sites I visit don't work
> without script. They cover the page with an opaque DIV. They bury
> the page text in javascript. Crazy stuff. It used to be that the idea
> was to make pages work in any browser. That's now reversed: Use
> the latest Chrome, enable script and turn off ad blockers, or we'll
> stop you accessing content.
I must agree with you that javascript is de rigeuer for most websites,
which is why the Tor Browser often has trouble with those javascript sites.
Luckily a proxy-based web browser (such as the Opera VPN Browser - again,
not a VPN but a proxy) usually has that unnecessary 'crazy stuff' blocked.
Since you knew about 3dyd (which was new to me until you mentioned it), I
wonder if you know of any other Windows-based proxy-based privacy browser?
> The other day I was reading an article at Atlantic. I don't normally
> allow script. Atlantic works fine, except that their pages are designed
> for cellphones, so I need to either read giant, serif text or turn off CSS.
> With both JS and CSS disabled I get an ugly but very readable webpage.
> A friend was having trouble getting the full article, so I revisited with
> script enabled. I got a page with a picture of 2 hardhat workers that
> said they're having problems! I toggled JS on and off, cconfirming that
> this was a lying webpage intended to block visitors without admitting
> that they're doing so. Crazy. I'm guessing that JS allowed them
> to figure out that not all files/ads were being loaded, probably due
> to my HOSTS file. (I don't actually use an ad blocker.) So they sent
> me a "broken page" page instead. And it's not a 404. Someone actually
> designed the page and illustration and they're swapping it in at that
> URL.
Since I use proxy-based browsers almost exclusively, I've found out that
sites such as the NYT, Atlantic and Financial Times (among many others)
will let me in sometimes, and then suddenly won't let me back in again.
I suspect they have an IP count (e.g., three free articles per IP per day).
The solution, which seems to work for me, is to change the browser proxy.
But I do agree with you that they are all about blocking IPs, which is
where some of the useful domain blockers seem to work well such as
https://winhelp2002.mvps.org/hosts.htm
Someday I'm going to install the Acrylic host blocker which uses wildcards!
https://mayakron.altervista.org/support/acrylic/Home.htm
> With rulings that claim it's illegal to download a file in one piece
> from a publicly accessible server, we're one step closer to making
> http a closed protocol and having a legal basis to make web browsers
> conform to a DRM model, with no settings, no saving of pages, and no
> ability to view source code. Restriction via hassle could become
> legal precedent.
You're right about the "restriction via hassle" and the tendency of the
RIAA to prosecute doomed-to-fail legal cases simply for the PR it spawns.
Luckily, for our use, the free graphical youtube downloaders all work well
using only the public APIs that Google allows any normal browser to employ.