>The West has a “broadcast” view of IP and ownership... China has a “network” view of IP and ownership
Probably the best way I’ve heard it described.
>You do have to read or speak Chinese to get it
This statement kind of bothers me. Mandarin is great- loads of fun to learn and speak. My problem is that I meet a lot of highly skilled professionals who don’t do more in China because knowledge of Chinese is often held up as an “obvious” prerequisite. I have met countless people who move here and waste a year with flashcards instead of making things because they are convinced that being bilingual will make them successful in China (hint: it won’t). Investing your time in doing things the Chinese can’t do very well will make you successful in China- and there is no shortage of bilingual Chinese.
You can get the Mediatek MT6260 (or just about anything else) without speaking Chinese, by using Google/Baidu/Bing Translate and relying on the spectacular patience of Chinese people for these things. Will it be a little embarrassing? Sure. Should you learn Chinese? Of course- but should you give up on doing things in the amazing development environment of China just because you don’t have a gift for languages? Certainly not.
All my coworkers are bilingual; if I spoke better Chinese, it would only get me some platitudes. It is my skills that add value to the team, not speaking Chinese doesn't detract from it (also, I work for an American company, so our biz language is English).
I would disagree a bit about what makes us different, or what Chinese can't do well. Creative independent smart local hires or experienced international savvy local hire managers exist, but there are not so many of them yet (the supply is not well developed), so why not bring in a foreigner who is easier to find? After that, diversity is still important; something we take for granted in the states.
>What Chinese can't do very well in your opinion ?
Would be a very long post, it depends in which cities, and which age groups (there is a huge cultural gap with far more risk-adverse older Chinese who grew up during the Cultural Revolution) and generalizations are always a bad idea.
That being said I’ll take a stab at it- usually Mainland Chinese are not so great at problem solving, but they are fantastic at making the solution better and cheaper (idea Iteration over Genesis). There’s the expression "别在孔子面前卖文章” roughly "Don't sell your essay in front of Confucius” akin to our "Don't try to teach your Grandma to suck eggs”. Coming up with something new- or imagining you can carries a certain air of arrogance to it. This does not let itself to encouraging creativity. While this attitude has obviously led to accusations of China just cloning products, it’s also made many Chinese to be enthusiastic proponents of Open Source (although not contributors for the above reasons). So new solutions and creative problem solving will likely be something they rely on foreigners and overseas educated Chinese to do for some time.
Software, UI/UX, marketing and outsourcing are still comparatively weak. Nearly everything is C++, Java and PHP with only a tiny sprinkling of other languages. Programmers are held in fairly low esteem and not terribly well paid.
Service at one point was a weak point but in the past two years has started to pull ahead of the West significantly. Aliexpress is the tip of the spear as far as Chinese direct sales posing a threat to Western businesses and it’s service is not even close to what you can expect from Taobao.
My wife is a Chinese UI/UX designer, and the circle of colleagues is small in a big city like Beijing. Programmers are in demand and many now know C# and even Scala; their pay has definitely gone up a lot in the last 5 years.
Gongkai is more a reference to the fact that copyrighted documents, sometimes labeled “confidential” and “proprietary”, are made known to the public and shared overtly, but not necessarily according to the letter of the law. [...] this sharing of documents helps to promote the sales of their chips
That what I've always thought too about why it's often much easier to find complete datasheets for Chinese components than Western ones - at some point, someone leaked them to the public, but the company either doesn't know because they're not the type to send lawyers after everyone, or knows and doesn't care because it's a sort of implicit advertisement.
I have an unbranded Android device based on the MT6589, and the buying decision was influenced partly because I could relatively easily find the full datasheet for it ("MT6589_HSPA+_Smartphone_Application_Processor_Datasheet.pdf", in case you're curious) along with lots of other very detailed information including schematics.
The availability of a large number of different phones based on the same SoC/reference platform has also lead to a big repair ecosystem - Bunnie has written about this before: http://www.bunniestudios.com/blog/?p=283
This is the thing that I ponder most about. Bunnie does an excellent job of describing how massive and rapid innovation can happen when data is available. I run into the roadblocks all the time when I try to approach western chip manufacturers for data about their chips. If you want to hack hardware these days you are better off knowing Chinese than you are knowing English because the most up to date and available information on the parts you want to build with, is in Chinese. It is frustrating for English speaking engineers.
Guess what : the rest of the world has had to learn English for decades to be able to access interesting knowledge.
I'm French and these days I came to the point that I don't even bother to Google anything related to programming in French. I just Google it in English and now I find useful answers (thank you Stackoverflow).
I guess the wheel is turning and soon we will have to learn a bit of Chinese to access fresh knowledge. The barrier of entry is higher though (it doesn't even involve the same parts of the brain than other languages).
Chuck isn't complaining about the language, he's complaining about the hostility of big IC manufacturers to sharing their data. For all but the most popular/trivial chips, you have to sign an NDA with the manufacturer which means you have to first get them to give you the time of day, which isn't possible if you don't have large volumes and plans to purchase tons of chips from them.
Except for companies like Alwinner (which isn't "western" by any means), most of the documentation is in English but getting it is so hard we usually have to fallback to Chinese versions, which are shared more easily due to legal and cultural norms.
If you're at 31C3 and interested in an open-to-the-RTL platform, which (yes, I'm biased) seems the logical next step, bunnie is leading a discussion on lowRISC tomorrow (Monday) at 2pm.
I've done minor reverse-engineering in the past but this is really taking it to a level of dedication that I find awe-inspiring.
The whole GongKai ecosystem is fascinating. In pure Chinese pragmatism, it's messy, relies on the power of guanxi (the force that lubricates relationships), ignores outside-imposed rules, and is terribly effective.
Keeping latest-generation ICs out of reach for hobbyists and small startups is quite usual practice from many of manufacturers geared towards chips going into consumer devices.
Another example is gigabit ethernet- it has been around for 15 years, yet I still can't buy a gigabit-capable switch IC on Digikey. The part I need does exist but manufacturers either want $10k pre-payment before even giving at least partial documentation or the sales cycle is several weeks at best.
We were aiming for 3-port (2 PHYs, 1 MII) switches, and both Realtek and Marvel do make them (Realtek part is also available on aliexpress).
Designing commercial devices (even for small production runs) based on key part available only in grey market seems to be a slippery road. Unless you live in China and can talk to vendors about previous and potential future availability of the chips ;(
If you can get someone experienced to handle logistics, it isn't that bad. I know a few companies that routinely turn to grey market parts when st hits the fan for their mass manufactured product supply chains.
In my experience, Marvell is one of the worst to work with but their fabricators make a lot of grey market copies of Marvel's chips during off hours. If you can find a reliable source (fabricator who makes chips for Marvell during the day and grey market copies during the graveyard shift), it will usually be pretty reliable.
Thus empowered by our fair use rights, we decided to
embark on a journey to reverse engineer the Mediatek
MT6260. It’s a 364 MHz, ARM7EJ-S, backed by 8MiB of RAM
and dozens of peripherals, from the routine I2C, SPI,
PWM and UART to tantalizing extras like an LCD +
touchscreen controller, audio codec with speaker
amplifier, battery charger, USB, Bluetooth, and of
course, GSM. The gray market prices it around $3/unit in
single quantities. You do have to read or speak Chinese
to get it, and supply has been a bit spotty lately due
to high Q4 demand, but we’re hoping the market will open
up a bit as things slow down for Chinese New Year.
No shortage of them on aliexpress.com, and I don't see any requirement for Chinese literacy...
Reminds of me the recent, thread on "why open source hardware is hard", whose article was written from the point of a view of a (Western) company having a hard time finding value in making it's work open source. He's looking- as an individual entity- for return for his sharing.
Bunnie crystallizes terms where we can begin to see a response- while the "why hard" author contemplated the traditional Western unidirection flow, Bunnie characterizes a loose network not of product co-development and impacts to bottom-line, but comprised largely of support and availability. The Gongkai model makes available means and platform, rather than providing the conventional expected of open source: development.
Thanks as ever Bunnie not just for the hard work and amazing worklogs, but also for helping pull together good framing to help answer "why oshw is hard" with.
Probably the best way I’ve heard it described.
>You do have to read or speak Chinese to get it
This statement kind of bothers me. Mandarin is great- loads of fun to learn and speak. My problem is that I meet a lot of highly skilled professionals who don’t do more in China because knowledge of Chinese is often held up as an “obvious” prerequisite. I have met countless people who move here and waste a year with flashcards instead of making things because they are convinced that being bilingual will make them successful in China (hint: it won’t). Investing your time in doing things the Chinese can’t do very well will make you successful in China- and there is no shortage of bilingual Chinese.
You can get the Mediatek MT6260 (or just about anything else) without speaking Chinese, by using Google/Baidu/Bing Translate and relying on the spectacular patience of Chinese people for these things. Will it be a little embarrassing? Sure. Should you learn Chinese? Of course- but should you give up on doing things in the amazing development environment of China just because you don’t have a gift for languages? Certainly not.
That being said- Bunnie is the Man