Slashdot Mirror
Can China Pull An India?
ricst writes: "The New York Times has a story about how China is trying to leap ahead of India as the world's second-largest producer of software. Apparently the Chinese are trying to learn everything they can from the Indian software developers. It's not clear that if China becomes a strong competitor to India that 'jobs will be lost or simply not created' in the U.S. My guess is that the most creative software opportunities will remain in the US for some time, and the more routine development efforts will continue to be transfered overseas."
People may wonder how this type of outsourced programming works, and I'll run down a few examples here.
One is the fixed API method. A function or functions are needed that perform X on data Y. This requirement is simply farmed out, and code is produced that does this. This code is then integrated into the larger code by the contracting company. In essence, this is the grunt work of programming, and it's where India started, and where China will likely start too.
Another is code upgrade. Legacy code in one language is handed to a programming team, and the requirement that it be ported to new language X on system Y is given. India does a lot of this now, and their technology parks have a plethora of older hardware to mimic these legacy systems for developers to work with. The advantage here is that Indian's speak english fluently and reading native code with it's comments and documentation presents no problem at all for them. A legacy of British Colonialism that the Indians have turned to their advantage. I don't see the Chinese doing well here very quickly, as periodic reviews will be done in english, and communication could be a total headache.
The third is the requirement style. Software must perform X,Y,Z and run on systems A,B,C. This is becoming more common. In this case, the entire software suite, from the core to the interface is handled by the Indian company. This is where India finds itself today, and it's pretty good at it. I've reviewed some results from projects like these, and the coding style is uniform, properly commented and compact. It's also a unique kind of style, and takes some getting used to, but any given company will produce the same style each time, so it's certainly very acceptable. On average I'd say it's less buggy, BUT!, I only see the end result, it may have been hell just weeks before, and I never saw it. This is somewhere the Chinese could do well, as a final pass to translate comments isn't terribly hard, but
The last style is market need. This is where a perceived need is seen, and software is made to meet this need unsolicited. This I haven't seen very much, but as they become more aware of our market, India will certainly begin to try it's hand at this. China may never bother, as their own market is probably going to be big enough to consume any supply for a long time, and the very different cultures make the risk greater than a lot of companies may want to take.
It's debatable wether China will ever catch India, the difference in style of education and culture may be to great. China may end up with the widget API market, and may end up serving the Indian markets need for this, oddly enough, but wether they can break the language barrier enough to work directly with english commented and documented code is something I can't predict. It's one of thoise moments where paradigm shift actually means something.
No, mundane jobs usually lead to more mundane jobs.
I have been doing development on the east coast, also for 20 years, 10 of them as a consultant. I can think of no reason to continue developing here. I've always done development as a business, and as a business, the margins are no longer compelling.
I agree that it seems the gov't is hostile to individual software developers here. I suspect that as a group, individual developers are under-represented. We have no lobbyists. The body shops do, and chapter 1706 of the tax code of 1986 is an example of the results.
If you're happy and you know it, think again!
According to this article over at CNN.com, India wants to become the back office of the world.
This is basically IT-enabled services for tele-marketing, helpdesk support, medical transcription, back-office accounting, payroll management, maintaining legal databases, insurance claim and credit card processing, animation, and higher-end engineering design -- all of which can be delivered by phones, computers and the Internet.
India is aiming to become "the world's back-office."
Now this is something the Chinese cannot grab in a hurry at all!
McKinsey estimates the market for this to be half a trillion dollars by 2008. Even if that sounds far-fetched and we discount it by a factor of 10, it would still mean massive amout of money.
Finally India is coming into its own. Just as oil transformed the Persian Gulf, India has found its oil - her people which are her greatest strength.
The poster who suggests breaking out of the hourly rate game is on target. Most cost overruns occur on hourly rate contracts; in most cases the overrun is not noticed. Therefore corporations who only want to hire the cheapest hourly rate, won't negotiate, have turned a blind eye to opportunity. Any moderately complex project based on hourly rate must use deep bean-counting to succeed on cost targets over a fixed price project; Software success primarily requires getting rid of 'noise' in the development process and excessive management is a form of 'noise'. Distributed projects add extra management 'noise' to the development process; Mgt is betting reduced hourly rates will save the day. This can work, but requires more discipline than most companies are capable of (most people seem to use project management for presentation purposes, can't even set useful milestones - hah!).
Don't be afraid to 'game' the negotiation process. Learn all you can. If you can help a business win, you will be scoring reputation points which will lead to future opportunities. Keep your overhead low so as to minimize cost advantages to non-us competitors. Don't work for free. Open-source contributors are not working for free (most sensible ones have their finances in order), they are working for reputation points and breaking up entrenched markets, leading to more opportunity for really useful and productive people over drones.