France Has A PRISM-Like Program With Thousands Of Trillions Of Metadata Elements | Hacker News Reader

France Has A PRISM-Like Program With Thousands Of Trillions Of Metadata Elements

techcrunch.com

148 pointspirri46 comments

Non-blog-spam story by the french newspaper that outed it: http://www.lemonde.fr/societe/article/2013/07/04/revelations...

Why are you calling it blog-spam? They rewrote the story to readable English and offered it to the Tech community. (Who reads Le Monde? ;-));

There's nothing wrong with the English in the Le Monde version, what do you mean?

Le Monde is actually quite good. I try to keep a multi-angle view of politics in Europe, so I frequent a few newspapers/sites from different European countries. After shopping around a bit, I settled for Le Monde as the IMO best among the French daily newspapers.

> The best among French Daily newspapers

Seeing how crappy the french daily newspapers are, that's a pretty meaningless claim. Over the past few years most of the "quotidiens" have been doing nothing more than rewriting the AFP reports. They don't do investigative journalism anymore, and even the published articles are full of spelling mistakes, something you didn't see 20-30 years ago.

The level has dropped significantly for the press in France, and that's also why no one buys them anymore. All these french newspapers are heavily State-subsidized for survival. Let's not wonder a second about their independence.

Of course they are, because if you could, why wouldn't you? The potential is too useful for anyone in power to not exploit.

I love the hypocrisy of French politicians at being outraged at PRISM and wanting to offer asylum to Snowden, when they're doing the same shit.

What better way to hide that you're doing the same thing. Well, right up until someone leaks this story of course...

The politicians that want to offer asylum are also against this.

The politicians that like this thing, are the same ones that stopped Evo Morales plane thinking Snowden would be on it.

They thought Snowden was on the Morales' plane because of the intel they received from the bugs they planted in the Ecuadorian embassy in London that were found weeks ago. The Ecuadorians announced they discovered the bugs within 24 hours of the forced landing of Morales' plane.

Source, please?

The connection between the bug and Snowden is pure speculation on my part. But here's a link to an article about the bug.

http://www.bbc.co.uk/news/uk-23179431

I just find it a little coincidental that Snowden is working with Wikileaks, which is led by Assange, who is holed up in the Ecuadorian embassy, where news has just been uncovered that they discovered a bug a few weeks ago, and that this news was released within 24 hours of the diversion of Morales' plane.

Personally, I think the news release of the bug was intended to send a message to the US & UK intelligence services.

I suspect that the only countries that are NOT secretly doing something like this fall into two groups:

• Those who make no secret of it.

• Those who don't have the technology to do it.

Or politicians are so corrupt they don't want to be found out...

Like, Brazil (for example, when our intelligence agency DID spy on suspected corrupt politicians, everyone involved with the investigation was fired, and all related investigations got closed, much to unhappiness of the population)

Fascinating. Do you have any links to articles describing these events in Brazil?

It is... very extensive and lots of tangled stuff (there was even Brazil-Italy government spying it seems...)

But you can find most information under the case that was forcibly closed (Operação Satiagraha) and the names of the protected criminal (Daniel Dantas) and the fired chief of investigation (Protógenes Queiroz)

And the Germans. The Germans haven't been party to most of the post-WWII spying, if I understand geopolitics correctly.

Where did you get that idea? Germany spends something like half a billion every year on signals intelligence.

It's all relative. The US spends some $75 billion: http://money.cnn.com/2013/06/07/news/economy/nsa-surveillanc... . I don't think Germany's budget will go a long way towards sweeping up global communications and analyzing them.

They also spend on their military a tiny fraction of what we spend on ours. Would you thus claim that there's no German army?

I do not believe that Geopolitics is anticommutative, that is

    Eastern Germany spying on Western Germany + Western Germany spying on Eastern Germany !=0

But I never really looked at the group aspect of international relations.

The war wasn't over when the preparations for Germany's new secret service began. The material collected about the Soviets survived the end of the war because it was hidden (on microfilm) in sealed canisters buried in the German Alps, with the intent to side with the Americans and start spying on the Soviets again... The plan totally worked. To make it even better, the organization gobbled up high-profile Nazis who wanted to avoid prosecution like nobodies business. Even today, whenever someone wants to take a closer look at the people who work(ed) there, documents vanish and nobody knows a damn thing.

Then there is the entire StaSi. Of course they lost, so we get to point fingers at them and declare them the bad guys.

Germany may have strong privacy laws, but our three branches of secret services are exempt in various ways. Don't kid yourself. We are most likely right up there with France.

Well, looks like I was the fool:

http://www.spiegel.de/international/world/edward-snowden-acc...

The phrasing is that much more confusing because France is a long scale country, so "one thousand trillion" is the valid way to say 10^21. If 1 byte = 1 metadata element (unlikely), that's a zettabyte. Which is probably not what's meant.

http://en.wikipedia.org/wiki/Long_and_short_scales

The headline of TFA says "Millions Of Trillions"

The headline of TFA's TFA (the source article) says

"La DGSE collecte ainsi des milliards de milliards de données".

That translates to 'millions of millions' pieces of data, or on the order of 10^12 (trillions in short scale) pieces of data.

I was sure milliard was billion and now you are making me doubt

In French: 1 milliard = 10^9. And 1 million = 10^6. So bnegreve is right when he says 10^18.

A milliard is a billion.

No 'milliards de milliards' translates to billions of billions. So that would be 10^18

google says

10^18 bytes in petabytes = 888.17 petabytes

or basically a million terabytes if you round up.

basically 3 million 1TB hard drives in raid 5 (1M x 3). you can probably fit a 500 1TB 2.5" drives into a single rack if you really try, and 2,000 racks can fit easily into warehouse-sized facility. (grid of 100 x 20 racks)

Ack, you're right. My French teachers would not be impressed.

Hi, I wrote this post. I see that everyone is as confused as I was with the "milliards de milliards" :) "Millions of trillions" is the right way to say it in English I believe (if you don't want to go over trillions). I edited the headline a couple of minutes after publishing to correct my mistake. Sorry!

Could this be translated as "billions and billions" or "billions upon billions"?

Literally it says billions of billions, but in context it most likely means billions upon billions.

Why? Petabytes seems like a more likely order of magnitude than gigabytes, doesn't it?

eta: From the article: "capable of managing dozens of petaoctets of data, in other words dozens of millions of gigaoctets"

No. The correct translation is "billions of billions"

Actually 'zillions' is probably the most accurate translation.

One should rather interpret this as "zillions", because "milliards de milliards" is generally used to mean a mind-boggling quantity like in "des milliards de milliards d'étoiles" (billions of billions stars).

Previous discussion of the topic: https://news.ycombinator.com/item?id=5990016

Thousands of trillions? At what point do we start expressing numbers in terms of the number of known atoms in the universe? This is big data. Civil liberties aside, this is awesome.

I'd express this as .19, or 15/78 due to the French dataset containing 10^15 and the number of atoms in the universe, at a speculated lower bound, being 10^78.

The one good thing that could come of all this surveillance, maybe we get a better version of Sim City that actually does need cloud computing.

Isn't it obvious to everyone by now? We have to wake up, basically every developed country has a program like PRISM that varies but the whole premise of capturing information isn't anything new. This is something that the US and other countries have been doing since World War 2 in some shape or form.

I could easily write a script in Python through some crafty Google searches, the Facebook API, Twitter API, LinkedIn and simple scraping could compile a report of a person and or persons activity on the Internet throw in some natural language processing and I could paint a vague but nonetheless somewhat accurate representation of a person. Capturing phone meta data and tracking locations via cell phone towers would obviously only be only something those supplied with such information would be able to store and have access to.

The difference between a surveillance scraping network that infringes upon your privacy and one that walks the grey line very closely is what information it is given. The technology behind PRISM like programs is largely the same, perhaps fluctuations in how fast information can be retrieved, sorted, tagged and stored are the only difference.

I think the surprising part about this is the following statement: "The DGSE entirely designed the program and no data protection representative is overseeing it." I differentiate between these programs more with "It's not what you collect, it's how you use it." :-)

Also known as quadrillions.

Holy millions of trillions, Batman !

(later edit: with all due respect for the seriousness of the topic, I loled when I read the title; seems like the submitter also noticed the alliteration and changed it to "thousands of trillions", which made me laugh again)

post-napoleon french number system is quite... peculiar.

Some places kept pre-revolution French (like Switzerland) and there they use "normal" numbers (ie: 79 is seventy nine, instead of sixty ten nine, and 84 is eighty four instead of four twenty and four)

"Normal" according to who? :-)

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

And then there's Danish...

I recall reading on a newsgroup sometimes in 2001 (prolly long lost since then) that the french internet was being monitored by french the government. So kind of old news to me.

With all this talk about metadata, it would be wonderful if these agencies published and standardized their ontologies.

Reply on news.ycombinator.com