simonpang

As a software engineer, it's very important to learn about Gall’s Law, which states that complex systems cannot be created successfully from scratch. In reality, even large systems, such as Netflix, Google, or Facebook, have started small and built incrementally over the course of decades.

@ChShersh You get what you do. Writing incrementally gets constant feedbacks. Iterate much faster. Gain insight. Low risk of technical mistakes. Design upfront gets you a better design, better model, less tech debt but holes in the system when you’re wrong.

I noticed two completely different ways of programming: 1. Write some small code, run, see the result, adjust, repeat for hours. 2. Sit for hours writing code, modelling with types, writing tests, then spend minutes fixing what you’ve got wrong. Turns out, this deeply influences the technology, programming languages and frameworks you prefer.

@dumimo @AutismCapital @DOGE_DOJ These are not classified but good materials for prosecution. No one will go to jail. They were allowed to do this to keep DOGE safe.

@AutismCapital @DOGE_DOJ why would USAID have classified documents, let alone be shredding and burning them in the middle of what appears to be closing down of fraudulent activities? Shouldn’t there be a preserve order of some sort in place and enforcement?

🚨BREAKING: USAID is telling people to prepare for an ALL DAY SHREDDING EVENT to burn and shred all USAID documents. This isn't even parody. They're deleting history in real time. SPEECHLESS.

@flyosity I think you are on point. It could be also less dramatic. People don’t get promotion for maintaining updates, so it didn’t worth the time to do it the right way.

I have no inside information but I'm gonna tell you exactly what happened At Apple there is an A+++ 10x design-engineer-prototyping team that works on all core iOS features and built this perfectly Then it got handed to a B player for updates and they broke it

@steipete 4K 24-inch screen is incredibly rare in 2024.

“The entire concept is increasingly archaic and niche” Got one for my mum years back and yeap. true. theverge.com/24303351/apple…

@elonmusk "We don’t really patent things. Patents are for the weak" - Elon

@ChShersh FP only work when there’s no side effects. But this is NOT what computers do fundamentally. This affects how software is written in ALL layers. A huge abstraction cost of an extra FP layer isn’t worth it for most programs.

My main explanation why FP isn’t popular: Tooling > Language In Java, I can jump into any codebase with a debugger and find my way out. In Rust, I have pretty and helpful error messages. In Go, I can depend on just a GitHub repository, no need to deal with package manners at all. In C++ In JavaScript, I can start writing code by having just a browser. It doesn’t matter if the syntax is elegant or if the type system is powerful for most devs. Devs care about DX, not aesthetics.

@BallisticBooger @jacobtechtavern @yasirmturk Unfair lock most likely is a spin lock. This is the correct type of lock to use in this case.

Advanced Swift Concurrency: don't use actors!! You can protect access to shared mutable state AND keep your API synchronous by using OSAllocatedUnfairLock.

@b_nnett @rjonesy App Store is highly competitive, low growth is pretty normal.

Ballpark Pixelmator acquisition math. $50 * 51k ratings / 15% rated (?) = $10M lifetime * 4 for web purchases (?) = $40M lifetime over 7 years = ~$6M/year avg run rate 3-4x multiple = ~$20M acquisition iOS ≈ $0 Photomator $600k ARR * 7-8 for subs = ~$4M So ~$25M

@azamsharp Count me in.

Would you consider starting a new project in UIKit in 2024 or beyond?

@mutagen_afk PlayStation自己想走Nintendo路線，就唔好賴微軟逼佢⋯

索尼要靠astrobot翻身是咪代表佢好快收皮😂😂😂😂😂😂

@twostraws almost 20 years 🫠

This definitely makes me feel old: I signed up to Gmail 20 years ago today 🫠

SwiftUI is great if your app targets the latest iOS.

@alexqgb @awilkinson New forms can be introduced to the market anytime. But the underlying tech took a decade to build. iPhoneOS can’t be built without OS X. iPhone had full web browser which was unprecedented.

I’ve spent 24 hours with the Apple Vision Pro. I’ll start by saying this: I am the ultimate Apple fanboy. Going back decades, no matter what Apple releases, I’ve been all in, to the point where my friends make fun of me for it. Even when Apple ships something objectively bad, like the original Apple Maps, I’ve always defended them and been confident they would figure it out. In the past, I’ve felt like I can see where they’re headed and that the first wobbly step is a step in the right direction. But after using my Apple Vision Pro for the last day, I keep asking myself the same question: “Would Steve Jobs have shipped isolating digital ski goggles that leave a red mark on your face when you take them off?” Sure, there’s all the rough edges. The slightly blurry field of view, the wobbly video passthrough in low light, the weight. Those will all get polished and fixed iteratively, as Apple always does masterfully. But I struggle with the actual form factor and idea of this device. And the fact that Apple has never made “quantum leaps” but is highly iterative. Today’s iPhone is just an iteration of the original iPhone. The form factor is the same, just evolved. Bigger. Lighter. Faster. Better software. Same with the Apple Watch. The one from 10 years ago looks basically the same as the newest one, its just been refined a ton. 
Yes, it’s a massive improvement. But the device’s core form is pretty much the same. But Apple didn’t make the Apple Watch round. They didn’t make it 70% thinner. The rough size and dimensions are the same. It’s the same product, just grown up. But in the case of Vision Pro, I just can’t see this form factor—even a highly evolved 5-10 years our version of it—as a breakthrough product. Unless Apple changes their typical approach (which I can’t imagine it will) and produces a radically different Apple Vision that resembles a pair of glasses—without the light shield/ski mask/isolation, I struggle to see this being a device that most people use in the longterm. Let me ask you this: Can you imagine sitting at your kitchen island with your kids running around wearing an Apple Vision Pro? Or donning your digi-goggles in a coffee shop full of people? Of course, there will be people who will. But there’s something inherently isolating and weird about this device. To me, the one killer app is watching movies, but even that has its problems. Yes, it’s incredible to be able to transport yourself to an IMAX-sized theater with Spatial Audio. Until you realize that you can only do it alone. What about when you want to sit as a family and watch a movie? Do you and your kids all throw on your Vision Pros and join a SharePlay? I doubt it. Is it neat to turn down the volume on the world and work from the side of a crystalline lake? Yes, but it’s a vitamin not a pain killer. Who asked for this? It feels like a novelty Can I skip ahead a few decades and see this as the future? Of course. If I can simply wear special contact lenses or glasses and get the same experience, that will be magical and incredibly useful. But I believe that technology is decades away. I’m with @zoink — this feels like the Apple Newton. Correct in direction, just a decade too early in the wrong form factor. An impressive tech demo that should have stayed in the lab. I hope I’m wrong.

@ianzelbo @ChristianSelig @Casshibra The speaker is bad. It also has the worst line out. Average PC motherboard has better sound qualities.

@ChristianSelig @Casshibra Horrific. I had it

Okay scratch that. Give me a desktop monitor recommendation that is at least 4K, 27”, ideally not grossly matte, has HDMI out, and bonus points for built in speakers

@azamsharp Approved.

Working on a big update to an already detailed article on SwiftUI architecture. You can read the current version here: azamsharp.com/2023/02/28/bui… #SwiftUI #iosdev

@krzyzanowskim Maybe you already knew the cause. The tool is not perfect. Pushing the Swift compiler too hard and using too much generics?

@simonpang that's not something to be on the agenda to address on the language tooling so far. and so far the only workaround is to link prebuild optimized binary. that itself has its own issues.

Monthly reminder that Swift code in "debug" builds result in up to 10000x slower execution than release builds

@krzyzanowskim Probably it should be called “defect”. Performance is a defect even in debug build because it slow down development.

@simonpang Some can't be "fixed" if not bugs. That's something easier to state than execute.

@ios_dev_alb Why import Foundation?

Tell me you are an iOS developer, without telling me you are an iOS developer👇