Hacker News new | past | comments | ask | show | jobs | submit login

I think you will find that with many other types of technology, due to scale and interoperability effects.

For example, when the automobile breakthrough happened, manufacturers quickly converged to an "average" design that was cheap to manufacture in large numbers, captured the main strengths of the technology, and happened to suit the average small city, rural or suburban lifestyle where most of the potential customers lived.

Now, this machine was not suitable for a small mountain village. Nor was it suitable for dense urban environments. You could design automobiles specifically for those markets - for example, a much more compact closed two seater, that would save fuel, road and parking space. But you couldn't produce it at a scale that could make it profitable against the dominant design, and the clients would face difficulty, risk and even ridicule when using the dominant "big car" infrastructure.

So what we got instead was a redesign of the city to suit the dominant technology variant, with what we can confidently say today were disastrous results. Just like your Jira consultants, starting from the 60s an entire cottage industry appeared advising cities on how to correctly plan road infrastructure of sufficient "capacity", how to set minimum parking requirements, etc.

If you happen to fall right on the average case of a technology, you gain almost magical benefits and substantial competitive advantages. So your competitors will try to emulate your success even if they don't fit the same pattern. If the interoperability and scale effects are strong, and customization is difficult, we should expect a similar pattern to emerge. And with software, you have both insane scale effects (every copy sold after the first costs zero dollars to make), and strong lock-in effects (file formats and protocols, inherent complexity which leads to a limited talent pool - both your employees and suppliers aren't willing to waste time and effort into a custom technological dead-end for which you are the only potential client etc.).




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: