Let me count the ways:
- The Internet itself is the basic enabler. The more people connected globally, the bigger the talent pool, the better the chance to get critical mass on an OS project from somewhere out there. The more connected and trained people, the greater the competition for gigs, the lower the compensation.
- The commoditization of computing. Very little of OS is new computer science. Most is plowing old ground, recreating functionality long available. Linux and mySQL are not about innovation, they are about consolidation and the collateral disruption of margins in erstwhile commercial software categories. The routine and understood are also easier to send outside.
- Modularity and open standards. The same methodology that enables breaking an OS project into pieces to be reassembled in Sourceforge, and run on well-defined downward APIs and data standards, likewise enables the commercial user to decompose a private project into pieces that are completed elsewhere. The stability of architecture and requirements necessary for this modularity is also a sign of the lack of fundamental new work.
- Low capital costs for developer class equipment. The same scale economies that let the college kid run Linux enable the Indian outsourcer. Used to be software costs for developer tools were a bit of a barrier, but OS fixed that issue. More competition, see above.
And some of the feedback loops:
- If you're competing with free, you've got to be - well - cheap. And that's not available in Silicon Valley. Even if you are building on free, or adding value through services, you get to cope with customer expectations for cost set by deployment of things like the LAMP stack in enterprise environments. Guess where you get cheap?
- The more offshoring, the more local infrastructure to support it, the greater ease and more incentives to set smart people on the path to learning code. The more raw talent looking to show its stuff through OS (or malcode for that matter).
- The more commodization, the less barrier to entry. The less attractiveness to investor equity capital. The more need to get the jobs done with absolutely minimal expense. (The more A round pitches I see with already established offshore teams.)
The Internet routes around choke points. It turns out the high price of implementing general purpose functionality was such a blockage. OS and offshoring are just different paths on the route around. Find something novel and defensible, or get used to it.
Source: Due Diligence