Simplicity Is an Advantage but Sadly Complexity Sells Better
eugeneyan.comIMO, this article is too vague to be useful. The focus isn't clearly defined, and the examples are few. Like most philosophical texts, it's not neither right not wrong, and its impact depends on the reader's state of mind.
There's a particular dubious point: equivocating simple systems with stable systems. The paragraph about Instagram looks plainly wrong to me, about sticking to "battle-proven and easy-to-understand PostgreSQL and Redis". PostgreSQL is certainly not simple nor easy to understand. And Redis was not much "battle-proven" in 2012, just 3 years after its first release. It's even funny to oppose them to "NoSQL data-stores", since Redis is a NoSQL key-value store.
I've seen many articles like this. While I don't disagree, more prescriptive advice is needed. Perhaps getting a bit more concrete regarding what is meant by simplicity and complexity would be helpful. I'm afraid that "simplicity" will become a mantra used to justify a lot of status quo inertia and wheel re-inventing.
I would add one more thing to why simplicity is good: - it is easier to reach consensus on.
I have started to realize that so much in our society is how it is because we need some kind of consensus to move forward. And so much of added complexity exists because there are no good consensus.
Think about OS portability for example. If we had consensus on the OS APIs then there would be no cost to portability at all almost. But it is a complex thing and hard to get consensus on so we can't have things that is compiled towards an ISA and then runs everywhere. If something is super simple then it will be easier to make a consensus among people that it should be adopted.
But this also connects to the topic of this article, if it is easy to create a consensus among others about it that might be bad for business. That will make it super easy to compete with you. And make it harder for you to sell it (it is hard to figure out how to monetize open source).
This is partly why browsers and javascript got so much traction. Also why I expect webassembly to eventually be the default compilation target. Who knows WASI might evolve to be that de-facto OS API layer
From wikipedia:
Solomon Hykes, a co-founder of Docker, wrote in 2019, "If WASM+WASI existed in 2008, we wouldn't have needed to create Docker. That's how important it is. WebAssembly on the server is the future of computing." Wasmer, out in version 1.0, provides "software containerization, we create universal binaries that work anywhere without modification, including operating systems like Linux, macOS, Windows, and web browsers. Wasm automatically sandboxes applications by default for secure execution".
“Instead of having a complex, catch-all solution, consider multiple focused solutions”
Aren’t most complex solutions just a recursive version of this strategy? I don’t hate this idea, in fact I often love this idea but it can more often than not pathologize into not invented here or the xkcd standards meme.
Is this one more instance of confusing complicated with complex? I am a complex system scholar and I think that educated audiences should use complexity only when there is an emergent aspect.