[throwaway0asd]

1. Avoid frameworks and tools performing heavy lifting that you could do yourself. Software frameworks are architectures in a box designed to eliminate innovation and performance in exchange for safety and convenience.

2. Set ambitious product goals. For example you want to drastically increase speed of transmission handling in your product. This means exploring various different transmission options, such as streamed sockets versus HTTP. You have to measure the difference and pick a winner.

3. Be continuously aware of your learnings. Very quickly you will find that the things you learn, from evidence and measures, differs drastically from popular approaches. These differences build on each other. After applying a few of these learnings in your product it will become both vastly superior to the alternatives on the market and completely unrecognizable to popular approaches for most developers.

Those three steps will impose a critical education path based upon evidence that will provide direction, originality, and constructive feedback. It’s not important that one option is superior to another. What is important is just how superior the better option is, which is something learned from experience and will dictate future learnings.

[adityam582y]

That's a new insights about frameworks. Thanks to share.