HN Explorer

Ask HN: What are new Platforms worth developing for?

3 points by geff82 ↗ HN
The PC 1982, the Web 1995, iPhone in 2008... All those platforms, at least in retrospect, became huge ecosystems. If you got around its technology, you could make huge profits and be an expert early on. What are promising Platforms and technologies like that in 2016?

3 comments

[ 4.2 ms ] story [ 14.7 ms ] thread
keithwhor
On the back-end, microservices. Starting with AWS Lambda. Google Cloud Functions. The space is going to move very quickly in the next couple of years. I'm super biased - as the founder of stdlib [1] I have a vested interest in making sure the space thrives - but basically, managing infrastructure is going to tend towards non-existence for the majority of general applications. As software developers, the way we write backend code and build apps will change significantly.

[1] https://stdlib.com

brudgers
Betting on a new platform can pay off, but only with luck. A person who bet on the VIC-20 in 1981 and was an expert in 1982 missed the first mover advantage for the PC. Probably would have switched to the C-64 in 1982 instead...and then been perfectly placed when it was discontinued in 1994 to jump onto the Web in 1995.

The first movers on the iPhone were making flatulence simulation apps and flashlights and IamRich. The bottom was empty and there was a bit of money there. But even in 2008 a developer would probably have been better off building something for the web, just as in 1995. HTTP hasn't changed and neither have the rules for publishing to 'the internet app store'. To a first approximation, all the huge profits from the iPhone have been had by Apple: it's hard to think of any other billion dollar business built on the iPhone. Even the B2B iPad loving Plangrid and DrChrono don't appear to have hit unicorn velocity.

So if I had to place an optimum bet, B2B on the internet. The 'new' platform that I think has legs is Windows 10. Between developer tooling and a roadmap that suggests a goal of running on just about anything with a microprocessor and 'integration with the Cloud' resistance may turn out to be futile.

Good luck.