Universal asks Google to take down 127.0.0.1 for piracy | Hacker News Reader

Universal asks Google to take down 127.0.0.1 for piracy

chillingeffects.org

99 pointsclessg36 comments

I wonder how long it will be before the majority of this planet's computing power is devoted to automated processes interacting with other automated processes in such a way that the end result is complete nonsense.

Imagine spambots communicating with honeypotbots. Adbots serving clickfraudbots. Recommendations engines suggesting new content to ratingbots. Chatbots lying to each other in channels devoid of actual humans.

That's basically how the stock market works already

You might like subredditsimulator: https://www.reddit.com/r/subredditsimulator

...the call is coming from INSIDE the house...

For anyone who doesn't understand the context: [1] This is a quote from the movie "When a Stranger Calls" [2] 127.0.0.1 is the IP address for "localhost", one's own computer.

Total nitpick on my part, but the line is from an urban legend, which "When a Stranger Calls" (among others) is based upon.

A few references:

http://tvtropes.org/pmwiki/pmwiki.php/Main/TheCallsAreComing...

https://en.wikipedia.org/wiki/The_Babysitter_and_the_Man_Ups...

http://www.snopes.com/horrors/madmen/babysitter.asp

THEN WHO WAS TORRENT CLIENT

Going on a tangent: Checking DMCA notices seems like this is a good way to find sites which have a copy of the move ?

Is is, and Google even gives you a link at the bottom of search results when some results have been censored by the DCMA.

That's brilliant, even if only for it's pugnaciousness.

Considering that all DMCA notices must include various dire assertions under penalty of perjury, has anyone ever tried to sue/prosecute over a negligently false DMCA notice?

It's simply not true that DMCA claims are covered by penalty of perjury, at least not in the way that nearly everybody thinks.

The only part that is covered by penalty of perjury is that the person making the claim is, or is authorised by, the copyright holder.

This means that you commit perjury if you issue an otherwise legitimate DMCA claim for a work that is not yours, but you don't if you submit a completely false claim for a work that is.

That part of the DMCA is just smoke a mirrors, designed to make the censorship regime look like it's fair and balanced.

You still need to pay damages for lying about other parts, though.

Borrowing from others' arguments, one can argue that such selective and/or one-sided enforcement actually weakens a law or regulation. (Meaning here, no penalties for abusing the DMCA.)

Akin to speed limits. Make them unreasonable, and selectively enforce them -- perhaps for purposes, e.g. revenue generation, other than their ostensible purpose -- and people lose all respect for them.

You may attempt to enforcement through force, but you no longer have the public's buy in and willing participation -- something upon which democracy is founded.

P.S. I didn't mean for my comment to ride this high on the thread, by virtue of being a response.

Nonetheless, my point: We continue to fight e.g. artificial scarcity and strengthening threats, rather than working with efforts to actually improve the public's experience and gain willing participation.

[Meta comment on HN comment structure, not related to the topic.]

> P.S. I didn't mean for my comment to ride this high on the thread, by virtue of being a response.

As long as you're actually responding to the comment you're replying to (which you were), rather than just responding to a top comment with an unrelated comment to make yours more prominent, then I don't see how that could be taken as a bad thing. By necessity, comments are tree-structured; it'd be too confusing otherwise.

OH NO! 127.0.0.1 THATS ME!

Universal might not be as "stupid" as people think...

That address is most likely for a client that the most popular french language pirate streaming Cacaoweb site uses. http://cacaoweb.org/

Basically if you go to that site and download the client you will have a web service running on port 4001 which will accept certain parameters.

The links for that service (and other similar services) are published on the internet pointing to the local loopback address with a resource identifier of the specific content in question e.g. http://127.0.0.1:4001/?f=4a3aba6e5e7cff2825347f7293a462ae

Pressing that link on an machine on which the client is installed will allow you to instantly view the content, on machines on which the client isn't installed it won't do anything (think of it as a torrent magnet link).

So while it's a bit silly to see that you see a local loopback address on a DMCA request what they want Google to do is to not display search results that actually provide the link to the content, even if that link requires dedicated software.

I've encountered this while doing some investigating into several breaches on some sites, the breaches didn't deface anything and they didn't seem like they cared to compromise the content of the site either. What they did leave is allot of JSON content which was looking like this:

"episodelinks": [ { "link": "http://127.0.0.1:4001/?f=73b2ca67915c6e6f7cc67dd75379bab6","...: "en","subs": "" }] }, { "episode": "2", "episodelinks": [ { "link": "http://127.0.0.1:4001/?f=4a3aba6e5e7cff2825347f7293a462ae","...: "en","subs": "" }] }] }, { "season": "2",

That content was then accessed over and over by multiple different clients from mostly french speaking countries oddly enough more from Africa and Canada than France (which there is also quite an interesting anecdotal reasons to why, Africa well can't afford and doesn't have Netflix and Canada rarely gets decent french Dubs these days as most American shows will be only broadcast in English even if they are dubbed in French for France or other regions.)

Some of the HTTP clients were sending User Agent's of various streaming apps including the cacaoweb client.

What Universal seems to want is to make sure Google doesn't display these lists on forums and various sites and there's not really a better way to do it than to include $HOME:1337/agrirpirate! currently :<

What people need to understand with the DMCA requests is that it's not intended to only remove the URL from showing up in google results as that is ineffective it prevents any website with that url in it's content from showing up as a result especially when tied to certain keywords.

Blocking say the piratebay or whateversiteahsreplacedrapidshare.com on google ins't effective making sure that all of the 10000's of forums that are used to actually publish the links wont be displayed when people search for Taylor Swift New CD Torrent on Google is.

Interesting. I wonder if this should be considered as a different category that may not be covered by the law. I mean, it's one thing to block a URL at thepiratebay.com because it's one DNS entry in the global shared DNS namespace. A URL on 127.0.0.1 can reference something unique on literally every device that uses it. I can't honestly say this is inherently terrible in this particular instance, but I worry about the precedent of censoring the entire web because of what some systems serve on certain private URLs.

You can pretty much treat "http://127.0.0.1:4001/?f=" as a quasi protocol. The DMCA takedown request is for the URL "http://127.0.0.1:4001/#/fr/", they are almost taking down a whole protocol, not just specific contents. It's like wanting to take down "magnet://" from search results.

Let's start with the question, how links starting from http://127.0.0.1/ are displayed in Google search results at all? What will Google robot get if it tries to connect to 127.0.0.1 ? Mot probably, it will get a "connection refused" error. Even if it somehow happens to connect (but why should Google run a web server on a machine where the robot is running?), it surely won't find any movie there. So how does it happen that Google "finds" something at address 127.0.0.1 ?

> Africa well can't afford and doesn't have Netflix <

And if they had there would be some crazy region locks in place to avoid someone elsewhere paying African prices while living in, say, Canada.

Might want to take down localhost too just to be sure.

And ::1 while they're at it.

Actually it's not uncommon...

https://www.chillingeffects.org/notices/search?utf8=%E2%9C%9...

Also it wasn't Universal per se, but Trident Media Group on behalf of Universal. However, in link provided by @waldirbj https://www.chillingeffects.org/notices/search?utf8=%E2%9C%9... there have been instances of Universal doing such a thing.

I’ll bite: How do you think this happened?

Automated system with little to no oversight?

Other replies have indicated, surely incontrovertibly, that this is some sort of automated framework gone wrong (or wrong-er); but, in that case, I think a question is: what's changed so that that automated framework has discovered 127.0.0.1 now, and not before?

My money's on someone running a Plex server (or something similar) on their own machine, then going to reddit and saying "I have a copy of [X], you can watch it here:" with a link to 127.0.0.1 (since that's how they watch it, so why wouldn't it work for anyone else?).

I.E. We're dealing with the intersection of an automated framework with a "normal."

Probably someone forgetting to comment out some debug code or whatever... it doesn't take much imagination to see how this might happen.

Local test servers I guess? I doubt someone goes through by hand finding these links.

Plausible explanation here: https://news.ycombinator.com/item?id=9931765 . In short, a certain filesharing client serves content on http://127.0.0.1 and links to that are distributed around the web.

Army of lawyers and business people with zero technical knowledge?

wild guess: a branch of popcorntime in French?

Now it becomes a true loopback

rm -rf $HOME - it's the only way to be sure.

Reply on news.ycombinator.com