If anything, it's the opposite. With a proper harness you stop having to spend so much energy reviewing every little intermediate step, and can focus on the higher level.
I'm actually working on a project now where the biggest problem I need to solve is that the verifier that reviews the test harness is too strict.
I think your view is lazy. The article explains some of the actual reasons, none of which have to do with laziness.
I’ve built for Electron and did a course on Swift for macOS apps. Not out of laziness, but I don’t think I’d ever build native for Macs. And the Windows folk have been complaining for a long time about the native APIs.
Now, native on mobile, that’s something else. I’ve been stuck on RN/Expo because that’s what the resources the business had allowed for, but native Kotlin is much more enjoyable (AFAIK). Swift… dunno, still icky.
"Looks could be good, but they also can be bad, and then you are stuck with platform-consistent, but generally bad UI (Liquid Glass ahem)."
Since the discussion was specifically about platform-consistency, odd that the author would decide that personal taste might take priority over platform-consistency.
"It changes too often, too: the app you made today will look out of place next year, when Apple decides to change look and feel yet again."
Seems to be arguing the exact opposite? If you adopt a native API for controls, windows, etc., your app will change next year to look completely in place (perhaps for better or worse according to the author).
Going DIY on UI and your app might still, in 2026, have brushed aluminum and "lickable" buttons.
I’m firmly anti-Tahoe and haven’t updated, and I’ve started to find Apple despicable. But…
Anyone saying they had no trouble migrating away is either lying or delusional.
Just off the top of my head: external accessory issues when Linux wakes up from sleep; trackpad quality; battery life; full-disk encryption is still spotty if you, for example, want to use ZFS; boot-level security can be a nightmare to setup (although Evil Maids aren’t a concern for me); systemd things that don’t work and don’t report they’re not working; inconsistent shortcuts for basic things.
Man, I like the ideals behind BSD and Linux, but we gotta stop pretending that basic UX stuff is done and fixed, when we know perfectly well that it’s been broken for a decade (running Linux servers or desktop-on-the-side since 2015).
I don't remember ever having external accessory issues, I don't use a track pad, battery life is fine (although I will admit macs seem to do better) and the ZFS one is just odd. Literally first time I hear this because while it sure would be nice... everyone uses btrfs or xfs or even ext4 with FDE without any problems for the last.. 15 years? Just one FS apparently has some issues?
Maybe your definition of broken is just as subjective as mine (aka I don't remember ever seeing window management so bad as OSX since fvwm)
I believe they could “afford” it, given their staggering valuation. And, by being the ones with sense, they might even attract the kind of customer that wants to do business with companies with principles! The audacity, eh?
Here’s the problem I have with your take (even if I agree): LinkedIn has a product to sell. You’re not supposed to be the product, because companies pay to advertise job postings, they sell career tools, sales tools, etc.
At what point is that not enough for them to stop doing data brokerage or sharing?
My experience with Antigravity was that 3 Pro can reason itself out of Gemini’s typical loops, but won’t actually achieve it (it gets stuck).
3 Flash usually doesn't get into any loops, but then again, it’s also not really following prompts properly. I’ve tried all manner of harnesses around what it shouldn’t do, but it often ignores some instructions. It also doesn’t follow design specs at all, it will output React code that is 70% like what it was asked to do.
My experience with Stitch is the same. Gemini has nice free-use tiers, but it wastes a lot of my time with reprompting it.
I don't use Stitch it doesn't have the context of my codebase, I just tell Gemini to make the UI directly and its able to do it. The only time it failed is when my prompt and goal was bad. I told it to swap expo-audio with react-native-track-player and it was able to do it in one-shot. Implement Revenue Cat and it did it in one shot. I do task by task like all the other agent tools recommended. The harness I made doesn't install packages, it just provides code. I don't use Anitgravity or any Electron-based coding agent, mine has a Rust core and different prompt engineering, not sure why it works so well but it does.
I need to implement a better free trial plan, it's reached enough maturity where its my only and primary way I write code, I also use web chats to help me craft prompts. Reach out to test.
https://slidebits.com/support
‘Technology is neutral’ is one of the most cynical excuses, as it permits developers to remain in a child-like fantasyland where consequences do not exist, and shifts the blame to… nowhere I guess, since we also don’t like regulation or moral takes on what gets built.
reply