Intellectual Property

Software piracy plummets to inconceivably low rate

Software Pirate.jpg
He wants booty.
For once, everyone agrees that China is making strides in protecting intellectual property. The Business Software Alliance (BSA) just released a study showing that software piracy in China has dropped 10% since 2003. And China's State Intellectual Property Office (SIPO) has also released its 2006 China Software Piracy Rate Investigation Report, which trumpets a 2% decline in software piracy between 2005 and 2006.

But there's no risk of an epidemic of warm, fuzzy feelings breaking out any time soon. Fifty-eight percentage points separate the piracy rates in the two reports. BSA posits that the software piracy rate is 82%; SIPO puts the number at 24%.

What explains the difference? A charitable interpretation is that the statistics are actually measuring different things. The BSA study covers only "piracy of all packaged software that runs on personal computers, including desktops, laptops and ultra-portables. The study does not include other types of software such as server- or mainframe-based software." The SIPO report, on the other hand, purports to encompass "China's entire software industry." Although the news account of the SIPO report was short on specifics, SIPO seems to have included infringement of enterprise software (which may be server-based) in its measurements.

Of course, this explanation doesn't accord with SIPO's own numbers. As reported by NetEase, SIPO pegged application and operating-system software piracy at 36%; preinstalled software piracy at 20%; piracy of preinstalled software for which fees are charged (possibly enterprise software) at 63%; work unit user piracy (also possibly enterprise software) at 39%; and individual user piracy at 78%. These numbers don't seem to add up to a 24% piracy rate for the "entire software industry".

Which leaves a less charitable explanation of SIPO's 24% figure: they made it up.

Links and Sources
There are currently 8 Comments for Software piracy plummets to inconceivably low rate.

Comments on Software piracy plummets to inconceivably low rate

haha ,smailing..
you are right.
THE copy right of softwave is pretect by the gov..

You cited the wrong reference.
Please check the following link:

Ah the old shibboleth of "software piracy" numbers.

The only figure that's at all interesting is the number of people who _could afford_ to buy the software and don't. A number which by any accounting is extremely low.

To take one example, Adobe Photoshop. Widely available as pirated software in China.

But if, from the beginning, it had been "pay full price or don't run it", can anyone say with a straight face that there's an individual, company, or sentient creature of any kind that would actually have paid the price and be running it in China?

It's _trivial_ to lock software so that only legitimate buyers can run it.

So why don't Adobe, Microsoft, etc. do that?

Because having their software running on every computer in China preps the rubes for when there is a market.

100% of zero is zero.

But even a small percentage of people harangued into needlessly paying the Microsoft Tax after they've been lulled into using Windows because it was free, adds up to big money.

The Chinese government should lead the world on this: We'll enforce copyright when you make it difficult for people to freely copy your software.

Right... And the world continues to turn...

For one thing I'd be careful around them statisticians, boy are those guys nutty or what?

Anyway perhaps they included opensource software in the equation. But it still doesn't add up. 24% come on!

Eventually, China will care as more people develop software. Unless there is a move toward some new paradigm. It will be more likely that the West will see a shift first rather than China. Or maybe not.

I kind schizo on this topic, on one hand I would like to see free software, while on the other I hope to make a living making software... Hmmm...


BTW have you see Pirates of the Silicone Valley?? Bill is thugged out baby.


It's not trivial to LOCK software. Don't believe me look at the recent Blue-ray DRM fiasco. As long as it has CODE that execute in steps on MY OWN MACHINE it will be cracked!

Although I give you that the fact is in China piracy gives these big corporations a major market share. Piracy is condoned to an certain extent. But can they stopped it by "locking" their software, hell no!

If they really wanted to they could you know pull out their operations out of China. But they'd be insane to do that.

I'm also interested in whether or not business piracies are as rampant as personal piracy. I can argue on the merit of personal IP laws but not for professional business.

[joined multiple comments --JM]

>>It's not trivial to LOCK software

For want of a better retort: yes it is!

Go read something on one-time encryption (to use just one example of how it could be done) and then we'll talk.

Full disclosure: I work in the software industry.

Hmmm one-time pad? Are you serious. One time pads are not a practical way to encrypt current problems. Straight out of wikipedia...

If software are so trivially to secure why we still have hacked computer games? You think video game makers wants people to crack their games to gain market share?

Sure when you talk about encryption system we can talk about security and all that. But the claim here is about developing software that can't be cracked. As long as someone clever can look at the assembly code, they can find a way around your "security". There are plenty of examples for this.

Dude if you can make unbreakble software I suggest you start cashing in now, before some crazy new quantum computer invalid your one time pads...

Full Closure: I'm an undergrad math/computer science STUDENT, so I pretty clueless (I have not even taken a computer security class). I'm just making an uninformed observation on why software are not secure.

So go easy on the flame.


Sorry, my point is that as long as you have "programs" that execute... Then you can crack it.

When you encrypt a file... Sure, it's secure. Has anyone come up with a run-time encrypted system? I.E the entire O.S is under encryption.

[joined multiple posts. --JM]

I’ve written an entry about this issue, according to some info from your post and other relevant references on the internet.

My point is that statistic data from BSA and SIPO are all reliable. The trick thing is the statistic methods they used are totally different so that the results are NOT comparable.

But intentionally, SIPO put the results from BSA and SIPO together without detailed explanation. They tried to suggest that the ratio are comparable.

Let’s go into the details.
For BSA, they choose book-market-value-method,
piracy rate=(book value of pirated software) /(book value of all packaged software)
*for personal user only

But for SIPO’s first data, which they claimed 24% ONLY based on “book-market-value-method” also, the story is different.
The formula is shown as following:
piracy rate=(book value of pirated software ) / (book value of entire software industry).

I do believe that they can get 24% based on this expression. Let’s look at the value of entire software industry, which includes computer software product, pegged application and software services and exportation(计算机软件产品、嵌入式软件产品、软件服务与出口). The trick is here, software services and exportation which is independent of piracy behavior. Undoubtedly, the piracy rate will be dramatically decreased by put it in the denominator.

This is what they did−math with Chinese characteristics.

Hmm informative!

Actually if you put it that way then the Chinese figures are better IMHO.

Of course for BSA they care mainly about packaged softwares, it's what they're about.

For the Chinese it's better to look at it with respect to their entire software industry.

It depends on where you stand and who you "roll" with, haha.

One thing I learned in Probabilities and Statistics is there are no wrong answers in Probabilities and Statistics! It depends all on the claim and the interpretation.

So math with Chinese characteristics is right! For Chinese people!


Regarding crackable software, I asked an informed hacker in the known about it and here is his answer:

There is no such thing as uncrackable software as long as the user has unrestricted access to the hardware. Period.

There are some hardware implementations that store their firmware (software) in a manner that makes it very difficult to crack. For example, the HC11 micro-controller has built in eprom to store the software within the CPU itself. You can store decryption software, and a key in the chip itself and use that code to decrypt whatever is stored in off-chip ROM.

The above would stop the typical hacker, but it is not going to stop the Chinese government, assuming they had unrestricted access to the hardware. With the proper tools, you can remove the dye from the chip and reverse engineer the entire thing, including whatever is stored in ROM on the chip.

In military hardware, the circuits are designed so that if the cases are improperly opened, or the hardware is improperly removed, they chips burn up, effectively restricting access to the hardware.

> How can you encrypt a runnable program? As long as you run the program you can look at it's source code right?

I wrote a virus one time that decrypted itself on the fly. On the 80x86 hardware, there is an interrupt that is called every time an instruction is fetched. I hooked the interrupt and used it to decrypt the instruction. I was more worried about altercation than anything, and the encryption was actually for error correction. Even if an antivirus guy changed the code while it was running, it would self corrects, assuming the changes were small enough. (I played with virus code to learn, never released anything.) There always has to be a small bit of code that is unencrypted to decode the rest of the code.

Regarding one time encryption... Won't work unless you call in to get the key... and once you have the key, anybody can decrypt it, unless every single copy has a different key. Again, it is pointless... as long as I control the hardware, I can crack the software... one time encryption, public/private key encryption, etc... it doesn't matter.

... and for those really difficult cases that I have a hard time cracking the software... I can write new software that will alter the "game" or whatever at runtime to fool the "game" into thinking it is a legit copy.

[joined multiple posts. --JM]

China Media Timeline
Major media events over the last three decades
Danwei Model Workers
The latest recommended blogs and new media
From 2008
Front Page of the Day
A different newspaper every weekday
From the Vault
Classic Danwei posts
+ Culture and corporate propaganda in Soho Xiaobao (2007.11): Mid-2007 issues of Soho Xiaobao (SOHO小报), illustrating the complicated identity of in-house magazines run by real estate companies.
+ Internet executives complain about excessive Net censorship (2010.03): Internet executives complain about excessive Net censorship at an officially sanctioned meeting in Shenzhen.
+ Crowd-sourced cheating on the 2010 gaokao (2010.06): A student in Sichuan seeks help with the ancient Chinese section of this year's college entrance exam -- while the test is going on!
Danwei Archives