Satellite TV Hacker Tells His Story

Wired is running a story about Christopher Tarnovsky, the man who was accused of working for NDS, a company owned by Rupert Murdoch's News Corp., to sabotage a competitor's satellite TV system. Wired had a chance to speak with Tarnovsky and get his description of how the smart-card hacking war developed. Quoting: "Tarnovsky, who was known online as 'Big Gun,' says Ereiser offered him $20,000 to fix cards that were killed by ECMs, and he agreed. Each time NDS created a countermeasure, Tarnovsky would analyze the code and find a way to circumvent the countermeasure. He did it while working full-time as a software engineer for a semiconductor company in Massachusetts. 'I'd be at work and I'd check the IRC (channel) to see if they'd launched their Thursday countermeasure yet,' he says. 'It was like a chess game for me. I couldn't wait for them to do a countermeasure because I would counter it in minutes.' It wasn't long before NDS came courting. Tarnovsky had a contact at the company to whom he'd begun passing information about holes in its software, even supplying patches to fix them."

Motivation

It was like a chess game for me. I couldn't wait for them to do a countermeasure...

Anyone developing software designed to keep content locked down needs to realize that this is the kind of person they're up against. It's hard to beat that kind of motivation. Forcing an arms race is almost always going to be counter-productive to protecting your business, this company figured that out.

Re:Motivation

This arms race deserves some indirect praise. It's like an creationist debating with an atheist on philosophical grounds, rather than the creationist just saying some crap like, "But the bible said X, therefore you are wrong and I am shutting you out." Everybody wins in a healthy pissing contest. It's a bad analogy, I took a cue from this [slashdot.org] guy.

Re:Motivation

I think that you are right. One of the groups who benefits the most are the companies that want to apply DRM to their content. Some will learn up front how much the arms race will cost them. Others will learn what is probably the point at which they should stop trying, and yet other still will learn that it is a futile business tact, and that modifying their business plan is both cheaper and garners more and loyal customers.

Additionally, with the arms race comes better code, not simply for the DRM, but for the operating systems and applications that work with the content. It is indeed evolution of both content, DRM, and code in general. The arms race in this case (not that of nuclear arms) is the catalyst of evolution, and betterment for all users in the long term. I would never call such hackers bad, simply the opposite side of the DRM coin that MUST exist, as without it, the other side cannot exist either.

Try keeping all the coins in your pocket/drawer/whatever so that you only ever see the heads side sometime. It's far easier to just allow any side to show in it's turn. It kind of makes things like pockets, coin purses, piggy banks work well.

Re:Motivation

Using a hacker of this caliber is a double edged sword. If you don't keep him busy and entertained he's going to start looking for something else to do.

Re:Motivation

Give him some Porn.

Re:Motivation

Anyone developing software designed to keep content locked down needs to realize that this is the kind of person they're up against.

I don't understand why people insist DRM is an unattainable notion. It must be all those faulty software DRM schemes that were all eventually broken. Well guess what, hardware DRM is alive and kicking - and working, when implemented correctly. Hardware hacks are orders of magnitude harder to perform than software ones.

Economically, there are two trade-offs in DRM:
1. the cost of the hardware manufacturer to implement the DRM scheme, compared to the cost of the content he's trying to distribute
2. the cost for the DRM wannabe hacker (cracks, mod chips etc.), compared to just buying a legit copy.

There's no logic fault in saying that, for a certain type of content, with a certain cost, these two tradeoffs allow a DRM system to survive. That is, to cost small enough to implement as to not increase the cost of the content significantly, and high enough to circumvent, that the users rather pay than circumvent. This is not the same as "unbreakable", especially for the types of passionate hackers like Mr. Tarnovsky, but that's irrelevant.

Note that the 2. cost can benefit tremendously from an economy of scale, if it's enough for a single user to circumvent and distribute to all others. For example when the content is in a platform independent format (distribute decrypted music), or when the DRM system is implemented in software (distribute software crack).

This is not the case with, say, live High definition TV. Maybe someone can hack his topbox and have unlimited access to live Sports coverage, but he can't feed that content to me fast enough to be useful. So I need to hack my own topbox, and that could cost much more than the subscription to the sports channel.

Also, this is not the case with a console game, where I need, again, to perform my own hardware hacks. A mod chip costs significantly today, and when the GPU, CPU, RAM and DRM chip will be integrated on a single dye, a mod chip will be impossible, and one would need to hack his own silicon.

Re:Motivation

Everyone knows that locking-down content is utterly futile, in comparison to the provisions of a well-implemented digital rights management system.

Oh yes, as if DRM is unbreakable! One quote that I have heard (don't remember where) but it was "The only DRM that doesn't get cracked are the ones that no one cares about the content on them". Just about every DRM scheme known to mankind has been broken in some way or another. Honestly, the less DRM/locked-down-content we have, the less problems you have and the less people are going to be out to crack/hack it (just look at the PS3, because Sony made installing Linux on it very easy, there has been a smaller effort to crack it compared to say, the Wii)

Other uses for his techniques?

Wow, can we get this guy to decode some of the Bluray keys used? Break HDCP? His method is pretty straight forward, easy to follow, and looks fool proof. Expose layers in the chip and read the data directly. I don't see how manufactures can stop this. As long as the key is physically somewhere in the hardware, it should be possible to access it. I guess the reason this isn't done more often is because of the expense of the high powered microscope, toxic chemicals, and fume hood.

Now we need the max headroom video Pirate to tell

Now we need the max headroom video Pirate to tell his story.

Shocking!

I was shocked when I read TFA and found that it didn't easily summarise as "I spent ages hacking the system, then got bored because there was nothing worth watching".

The Video Shows the Holy Grail of Sat Hacking

I spent years hacking satellite television, from the early days, the glory days of the H and HU cards and then left the scene when DTV killed with the P4 card and lawsuits. I've written my own 3Ms and emulators. What Chris has done in this video really is the ultimate holy grail of smart card hacking. The security layer he is referring to, at least on NDS cards, is sort of a sticky layer that when you attempt to pull off the coating to access the bus, it simply rips up many of the thin wires on the chip and you're SOL. This is enough to discourage casual hackers and those without good resources. It also, as he mentions late in the video, eliminates the need for using "glitching", which was accomplished using a specially programmed Atmel chip and some software, to attempt to oscillate the voltage in such a manner that allows you to read/write to the card without having a properly signed packet. Dumping ROMs is exceptionally difficult to do, even with the thoroughly hacked HU cards, and he can just casually do it with his setup. Makes me think he could also dump the ASIC, something even in the heyday of DTV hacking, was never accomplished. This would eliminate the need for an access card at all- once you've dumped the ROMs, got a valid EEPROM, all you need to do is emulate the ASIC and opcodes for the processor (which on the HU card was a Texas Instruments TMS370 chip with a modified instruction set).

Re:The Video Shows the Holy Grail of Sat Hacking

Makes me think he could also dump the ASIC, something even in the heyday of DTV hacking, was never accomplished.

You can't dump an ASIC--- that's the very reason they exist in this application. It's not code, it's an Application Specific Integrated Circuit. It's essentially an unknown array of logic gates. The best you can do it try to reverse engineer it, and short of an electron microscope, you probably couldn't.

Re:The Video Shows the Holy Grail of Sat Hacking

The best you can do it try to reverse engineer it, and short of an electron microscope, you probably couldn't.

This guy is hacking smart cards with a hood, some off the shelf chemicals, a very precise scratching tool and a pile of computer & electronics gear.

Now realize that one of these days, resources like electron microscopes will be within the grasp of entities that are not a Government, University, or Corporation. It only takes one rich misanthrope...

What an arrogant douchebag

I mean...

Since NDS fired him he's been consulting for two semiconductor companies and a manufacturer of dongle tokens, but he misses his life in electronic warfare. If NDS doesn't want him, he says he'd be happy to work for Nagrastar -- jumping sides once again. "I could design a whole entire chip for them like I did for NDS," he says. "NDS thinks today that their technology is superior to everybody else's and it probably is, because they're 17 years ahead of Nagra technologically. But Nagra could catch up overnight if they used my services. "I'm a very valuable asset as far as smart-card technology goes," he adds. "I know everything about (NDS) as far as their intellectual property models go."

Then again, its Wired magazine. They exist purely to create arrogant douchebags, dont they?

Those are techniques used in failure analysis

The techniques Tarnovski used to burn the top off with acid is failure analysis stuff.
I knew a guy who worked at a chip manufacturer and that's what he did. Failure analysis.
Burn the top of the chip off with what he called "formic acid" (I think, this was over 20 years ago) which "didn't hurt the chip".
They would then look at it under a microscope and try to determine what had failed.
The second microscope Tarnovsky was using looked to be a wire bonder.
It welds wires on by hand, with a pantograph type positioner.
So you can connect the chip to the leads, for example in the package, common for eproms. You can see the little leads in the window of older eproms.
But hackers can also use those to reconnect the last link of a programmable chip like a PAL that has had the security fuses blown after programming. Then you can just read the program out of the chip. OOPS, there goes that programmable security.
I had a chance to get one of those once, but it was a big one. Too big for me.
The little tabletop one in the video would be neat. I would grab one of those if it ever presented itself.
Tarnovski used that wire bonder to grab the signals off the chip internally, where they are actually running.
Those smartcards are likely a serial device, but if you can get back to where the data bus is parallel maybe that is before the inherent security.
The guy is obviously good. Wonder if he has a college degree?

Re:Who wants to track down which company

Yeah cause there's no tech schools there or anything.

That aint nothing.

The guys at Flylogic really make some high quality micro chip reverse engineering [flylogic.net] (I was going to say porn).

Tarnovsky == Flylogic

Christopher Tarnovsky, the guy portrayed in the article, IS Flylogic. Yes, this is slashdot and nobody reads the article, but it even links to flylogic.net.

Re:Accountability?

Never. Because crimes are only for the poor.

Re:Accountability?

Silly rabbit -- It's criminal charges which are only for the poor.

Re:Accountability?

Why would Murdoch go on trial? NDS (the Murdoch company) was found guilty only of theft of satellite signals and fined like 1500 dollars. There was no proof for piracy or the like. If you want to put Murdoch on trial for something like that, that one of his many many companies did, then you're expecting the sky. No one is dumb enough to try that. There's zero chance of Murdoch actually being found guilty of anything related to this. Hell, I'm willing to wager that Murdoch himself might have actually been ignorant of this.

Re:

So claiming I don't know what's going on in my company cleans me of any guilt when my company breaks the law?

Re:Accountability?

That's pretty much the whole idea of having a corporation.

Re:

Heh, I would use a scientific argument...

Well, your honor, I thought I was working on a SETI project, you know, searching for ET. Damn if I didn't discover it was just HBO, not aliens, after all.

Re:uh, this is a PR fluff piece

The only moral of the story here is that an arrogant, ethics-free mercenary with access to any tool he pleases is given way too much admiration in the twenty first century.

Says who? You? You're just a pompous, self-righteous, moralist dickweed. Don't impose your anachronistic opinions on the rest of us. We don't agree with you.