why program efficiency [and usability] matters

in 2016 I wrote a small rant about the current downward trend of software and web development, entitled Why program efficiency matters:

Computer hardware has become faster, more efficient, and more powerful in recent years, which means programmers are not constrained as much by memory and CPU cycles. But does that mean programmers should just give up trying to make their code more efficient?

It doesn't matter if our programs are bigger! : I don't know about you, but I enjoy extra disk space for movies and music. Just because disk space is affordable doesn't mean programmers can excuse themselves for adding unnecessary fluff to their projects.

It doesn't matter if our code takes up more memory! : Multitasking computers have been a thing for a while now. With that said, I would like my computer to actually multitask. I shouldn't have to constantly worry about how many programs I have running in the background and how much memory they consume. Also, there are plenty of older systems running in corporate and educational environments that simply cannot handle modern (and memory-hungry) software without constantly locking up.

It doesn't matter if our code is slower! : Speed is always a value to strive for. Any sensible person would choose "faster" if presented with two programs that perform the exact same tasks but at different speeds.

That said, if you have to sacrifice any of the above for security, please do so. Otherwise, if there is any way to make a program smaller or faster or more efficient, without changing the core functionality of the program, then take the time to improve in those aspects. Laziness is no excuse for a slow, fat program. At the same time, don't let yourself be consumed by trying to make your code perform better before you have even finished writing the program.

this applies to desktop, server, mobile, and Web software all alike:

I'm making this post today because someone sent me a link to a post Casper Beyer made regarding Electron, entitled Electron is Cancer. I'll quote some notable passages from the post:

Well, it works fine on my machine, and I only have 32 gigabytes of ram. - Silicon Valley Developer, 2017

If that’s you, well then that’s good for you, but just because something performs well enough on your machine doesn’t mean there are not any performance problems. You are not your end-users, and you if you are a developer most likely do not run average hardware.

^ I made this point in my 2016 rant -- people have different hardware and developers need to keep this in mind, lest they want their programs only to run on a small set of machines in the world.

Electron is so great, we did not have to hire new people we can just use your web designers that we already have in-house and it is so easy! - Someone Actually Said That

Okay, sure having a plumber cut out a square wheel from a plank is also a lot easier to do than having a woodworker carve a perfectly round wooden wheel, but it is gonna be one hell of a bumpy ride, and square wheels are actually fine, right?

^ I've seen this a lot too; people have derived from do one thing and do it right philosophy, both in software and in expertise (although on the expertise side of things, it helps to be well-versed in several areas so you're more valuable in a job, but usually those areas are close enough together that they complement each other. you wouldn't want that plumber performing heart surgery on you, would you?)

if you have time, read Beyer's full post because it covers a lot of good points about Electron and about modern software developers as a whole. it's a rarity to find a decent dev nowadays who cares about efficiency, usability, and accessibility; and that certainly affects where technology is going as a whole. as we depend more on technology in our everyday lives (mobile, IoT, business) there is really no room for sloppy code to run in banks, hospitals, vehicles, and other mission-critical devices.