Hacker News Reader

Comment by gersg | Hacker News Reader

Google takes down content when they get notified, rarbg doesn't.

I've yet to see Google not return a piratebay link at any point in history. Is it a loophole where they take it down, but their indexer then immediately puts it back in the next pass, lol?

> I've yet to see Google not return a piratebay link at any point in history.

indeed. as it should, if it's relevant for the search.

> I've yet to see Google not return a piratebay link at any point in history. Is it a loophole where they take it down, but their indexer then immediately puts it back in the next pass, lol?

No. If you pay attention there can be a message at the bottom of the search results telling you how many results were removed due to takedown requests. IIRC, they used to even link directly to the request, but now I think you have to jump through hoops to see it.

The "loophole" is that a takedown request has to be for a specific URL, so it requires a lot of constant effort to even try to get them all. Pirate Bay always had dupes and a million mirrors.

I'm not being nit-picky or contentious - I'm asking from a genuine point of curiousity ...

but in the case of Google linking to the pirate bay, isn't the pirate bay the one linking to the pirated content? Google is 1 step removed in that node graph because they are just linking to the pirate bay.

I guess if they directly linked to a pirate bay page that had a magent link on it .... maybe (?)

A court is unlikely to care about the distinction between actually linking to pirated content, and linking to a page with both instructions and a link to the pirated content. To add, enough TPB torrents contain screenshots.

Also, Google's takedown request handling in Google Search is not a matter of DMCA or a legal matter at all - instead, it's like Content ID, where they have their own system for evaluating takedown requests separate from any law. Rights-holders can still send Google legal requests, but it's easier to go through the expedited processes Google provides that also won't increase rights-holders' liability if they happen to submit a false takedown.

Google seems to refuse removing because, according to them, "Whole-site removal is ineffective and can easily result in censorship of lawful material."

Instead of removing, they just remove links by request.

Sources: https://torrentfreak.com/google-opposes-whole-site-removal-o... and https://www.scribd.com/document/286275022/TorrentFreak-Googl...

However they did ban Pirate Bay in the Netherlands after a Dutch court ordered them.

https://www.makeuseof.com/why-google-removed-pirate-bay-from...

Isn't this the same loophole that MegaUpload used? Only removing a link to a file, not the file itself with the claim that other links belonged to potentially lawful owners of the file.

I mean, if the subpoena says "remove a link" you comply with that.

But there's also another fundamental difference: even if there's the expectation of removing all copies of the same exact file, it is "trivial" for MegaUpload to know, by using hashes. They do have access to all files, as it is in their servers.

For Google to delete all pirate links to movie X it would be much more complicated, and would put them on a position of being forced to be the internet police.

> They do have access to all files, as it is in their servers.

not if they have the encrypted content only, and the decode key is only in the hash portion of a url, which never goes to a server.

But i guess crafting a technocal "solution" to a legal problem doesn't work, since the law works off intentions, and how much money you pay lawyers...

That's Mega, the one that still exists. MegaUpload the previous one didn't really have the same encryption.

It's been 5 or 6 years ago now, but one night I searched for torrents for a particular movie, and Google returned hundreds of results from a dozen sites... and the next evening they returned 0. I think it was an October.

While I don't doubt that a torrent link shows up once in awhile, Google no longer usefully searches for such things. Or really anything, legal or not. It's more like a purchase recommendation system pretending to be a search engine.

And a terrible one at that, as it points to the lowest-quality blogspam sludge possible.

Google used to rule the net. These days they're not even the best search engine for legal content, let alone overall net searching.

Bing and Yandex will get you most everything you want.

It seems Google's tech can't keep up with the scale of the Internet anymore. It simply doesn't index a very large portion of the Internet now.

Google does not return links to the pirate bay for me https://i.imgur.com/MkAPoFl.png.

At the bottom there is a message that says "In response to a legal request submitted to Google, we have removed 4 result(s) from this page. If you wish, you may read more about the request at LumenDatabase.org" and links to https://lumendatabase.org/notices/27615507

Odd. Google always returns links to the pirate bay for me https://i.imgur.com/tNo6tbB.png

I'm in Canada though. But I did use Google.com.

that's a different search query. I did get that same result if I use your query instead. (I'm in the US)

edit: it's also for a "proxy" site. I don't really use torrents or follow TPB happenings and don't know how that is/isn't affiliated.

Linking to the site itself (especially when that’s what you searched for) isn’t the same as linking to a torrent to infringing content would be.

Canada has relatively lenient copyright laws/enforcement. It's likely that Google sees no legal need to honor DMCA-style takedown notices in Canada.

Here's a guide to the legal status of torrents here with broad categories, from most lenient to most strict (caution VPN spam): https://www.vpnmentor.com/blog/torrents-illegal-update-count...

Interesting to note that downloading copyrighted content for personal use is explicitly legal (not just overlooked) in Spain, Switzerland, and Poland.

The picture shows that downloading is illegal only in six countries.

In the US: https://imgur.com/a/VjlpwlN

Google will only remove specific URLs, not entire sites/domains. Even if every copyright holder with content on TPB sent a DMCA notice to Google today, new torrents -- at new URLs -- would pop up tomorrow.

Hmm i guess torrent sites can also counter by just not having static URLs for content?

Or by having lots and lots of duplicate static URLs (but only revealing them one by one).

Can't speak for rarbg, but plenty of piracy adjacent sites have a DMCA takedown program[0] to operate under this loophole. That way, most content survives but they are "protected"

[0]: https://annas-archive.org/copyright

Changing the url daily works as the dmca complaint refers to a specific page.

Reply on news.ycombinator.com