I agree with this so hard.. https://twitter.com/scanlime/status/1299746124350152710

Say why. Because on its face I’m not grokking it. If you're writing an application for a device, who is writing it for scaling beyond that device? Are people designing apps like microservices? And if you're writing a SaaS, shouldn't scalability be a principle? Here are some reasons I can think of: • Code interviews around O(n) for products-that-aren't-SaaS is wrong. And even for those who write services, O(n) isn't table stakes. The whole reason we have horizontal scalability is for services that aren't optimized. Everything still works great! So knock off the performance interrogations. • It usually takes multiple people to write software. People take ownership of areas. Areas, even within monoliths, eventually act sort of like distributed software. So let's be good at Conway's Law. I'm a cloud guy for pay, and a completely solitary-developer-for-single-user apps on my own time, so I'd appreciate hearing insights.

I don't follow your comment or what you're disagreeing with, and that makes me wonder if I understand the tweet. But IMO: * SaaS just going by the term doesn't have to encompass the entire world, and yet the examples of it we see around us are all trying to do that. Perhaps the problem is more the culture we're surrounded by, to maximize adoption and profit. * Selling native apps for single users can also fall into the same trap, but they don't have to. The scaling here isn't in the runtime behavior but the deployment and life-cycle management. * Open source also falls into this trap, with everyone trying to gain market-share by accreting features, shipping binaries, discouraging others from forking, etc. Not sure if this clarifies or muddies things further..

I was just trying to grasp at the tweet’s meaning, and was taking a guess. I still don’t know. Apps are thinking too big? You seem to get the author’s meaning of “encompass the world” - so I’ll just stand aside for others’ conversation.

@Kartik Agaram What you are describing sounds more like a problem of today's attention economy than of scaling in the technical sense that I attribute to this term in computing. In a global attention economy, there is space only for a few well-known players in anything, be it tech, art, or whatever else.

I'm also in the "not sure what the original tweet means" situation, but if I take it to be "having one giant Facebook for the whole world is maybe bad", then 💯. But is that a "computing" problem or more of a human nature/network effect problem?

Yeah, we can probably improve on calling it 'scalability'. I see it as an aesthetics problem, and the urge for one single service provider feels like it stems from the same source as wanting to optimize the big-O and so on.