Can Chinese Copycats Drive Innovation?
The spread of technology has always been more about imitation than innovation. For all of human history, technology has had a way of diffusing across continents with scant respect for political boundaries. Intellectual property, the assigning of monetary value to an idea, is itself a novel idea in the scheme of human history.
China itself has been the historical source of many innovations that have been “copied” by other countries–the litany of printing, paper, gunpowder, and the compass, familiar to all Chinese schoolchildren. The Dutch and the French engaged in what today would be called “industrial espionage” in trying to figure out how the Chinese of the Ming and Qing dynasties created their fine porcelain wares. (Turns out it was all about a bellows system with valves to allow continuous airflow that could maintain very high temperatures in the kiln–something that the Europeans of the 17th and 18th centuries, with their accordion bellows, hadn’t yet figured out).
These days, Chinese companies have earned a rather unsavory reputation as copycats. From consumer electronics like LCD TVs and MP3 players, to sophisticated telecoms networking equipment, even to automobiles, it’s no secret that Chinese factories shamelessly violate IPR.
And yes, copycat behavior is rampant on the Chinese Internet, too: business models, layouts, even logos are borrowed without much more than a change of language. In recent months a raft of Twitter clones has appeared in China: Popwu, Jiwai.de, Fanfou, Atnote, Zuosa. Even more niche products, such as 37signals’ online project management application, Basecamp, have been cloned.
This is disturbing to many. But most Chinese Web 2.0 C2C (”copy to China”) startups don’t see any problem with this kind of behavior. They point out that most of these copied tools are proven and popular concepts in developed markets–the U.S., Europe, Israel–but have no Chinese language version. Or they note that the true power of Web 2.0 is precisely that applications can be mashed and tweaked and improved. Twitter wasn’t the first short public messaging service, but its success–its massive user numbers, anyway–isn’t just due to its sleek design and its function set. Its success owes to what’s happening under the hood of the application. Twitter has had to overcome numerous scaling issues, and in doing so has contributed back to the Ruby on Rails community greatly. (Ruby on Rails is an application framework popular with many Web 2.0 companies).
This is where the challenge lies for Chinese companies. It’s easy to clone a design, a layout, a concept or a business model. But each Chinese clone will have to build its own backend. And these are often built on very different frameworks from those powering the sites or applications that the clones were modeled on. It will be interesting to see how these Chinese clones will deal with scaling issues–issues that still plague frameworks like Python, an object-oriented programming language and web framework and if they will have the same drive, to share their newfound knowledge back with the community.
Gunpowder was invented in China, some claim as early as the 2nd century. But it took a non-Chinese “borrower,” a nomadic people called the Nuzhen (Jurchen) who were ancestors to the Manchus, to improve the formula and unleash its truly explosive power on the battlefield in the 12th century, a thousand years later–ironically, against the Chinese. As Web 2.0 applications become more and more popular in China and a growing legion of Chinese developers familiarize themselves with the latest tools, we believe that China’s oft-disparaged “clones” will in fact make real contributions, and give back to the development communities they’ve borrowed from.
5 comments thus far
I think the roadmap is similar to MMOG scene, first import, localization then self development. I can see some light with xtimeline.com
Posted by Jason Vu on July 19, 2007 at 4:45 pm
The main challenge is that many Chinese sites are hacks, and therefore not scalable. However it would be the height of arrogance to think that once a hack, always a hack. Good hackers learn about scalability, and eventually build that into their programming.
The greatest challenge is what Steve Jobs would call “elegance” in hardware and software and site design. In the beginning, there is usually no economic justification for elegance, but as we have seen with the development of OSX, the MacBook and MacBook pro line of notebooks, the iPod and now the iPhone, it does eventually pay off if delivered consistently.
So when will China have its first Steve Jobs?
Posted by Paul Denlinger on July 20, 2007 at 10:31 am
[...] Filling A Market Need? Jump to Comments Digital watch had a great post this week about IPR and software
Posted by When Is Copying Filling A Market Need? « One-Eyed Panda’s Journal on July 22, 2007 at 5:55 pm
So, these companies are “driving innovation” on the other side of the browser. Getting together a stack that will scale with demand is definitely hard, but it’s a technical problem that is easy to overcome, especially when you have lots of very smart CS students graduating from universities across China.
The difference to the end user? A page that loads 2s faster.
Coming up with an original business model and design is the hard part. Where are those people?
Posted by eden on July 23, 2007 at 11:46 am
Python is not a framework, it is an object oriented and dynamic typed language, just like Ruby. It is also much faster than Ruby, performance-wise.
The analogous to Ruby on Rails in the Python world is Django, a loosely-based MVC framework.
As to the scaling point you make, any agile framework (like RoR or Django) will have scaling issues - these frameworks were designed with agility in mind, not scalability. The biggest pain points in scaling a web application lies, however, in the relational database. It is actually fairly uncomplicated to scale a web app server like Mongrel: throw in a few hardware servers, run a few hundred Mongrel servers, a load balancer and you’re good to go (performance is a whole other issue).
Relational DBs, on the other hand, do not scale cheaply. The usual way to scale an RDB is the vertical way: buy a a faster box, with more processors, more RAM, more power. So, the real issue today is: how do you scale an RDB horizontally; how do you scale a relational database on cheap commodity hardware so that you don’t have to spend all of your budget on expensive Sun and IBM servers? Flickr, LiveJournal, Twitter and a few others are working hard on that problem.
Posted by Rodrigo Landerdahl on November 9, 2007 at 9:27 pm
Post a new comment