Starling: The first real desktop written by AI

I think this is quite interesting…

  - another proof point that AI upends project economics 
  - AI will soon be able to make custom desktops and apps (msoffice clone anyone) that will undermine historic monopolies
  - if you are not doing to them they are doing it to you
  - makes language & framework a “don’t care”

AI prose is great at celebrating whatever you told it to celebrate, but it always misses the most important detail: why should I care?

I don't understand why this exists or why anyone would use it. It just looks like Cinnamon with a b00merang macOS skin. So if there's a critical bug or security vulnerability, are you prepared to support it? If not, why bother making this? Everyone will rather opt for one of the proven distros/desktops.

This post makes me feel grief. I mean no offence to you OP, this looks like a beautiful desktop and I am sure there was a lot of time put into this. But just sharing my grief.

3 years ago if I saw a post on the HN front page about a desktop environment written in Flutter (or a Swift port of Flutter), I'd be absolutely blown away.

Now, I see this, and I feel nothing. I'm not impressed. I don't want to know how it works. I'm not going to read 300k lines of AI slop code. I don't care about architectural decisions made because they're not made under any true constraints. I'm not even excited about the AI part of it, because I know AI can do this. We could all build this if we just prompt hard enough.

I'm tired, boss.

Weird, the swift port of flutter seems waaaay more interesting than the desktop itself. Was it just the dart code of flutter ported or the runtime too?

/edit Nevermind 'just' dart: https://github.com/starling-build/starling-engine#what-starl...

Still that part is kind of cool, 273K LOCs of the 335K

Who would use this? There is no known author for starling. We don't know if a single person plagiarized another Window Manager, if it is a company or FSB effort.

Right now it is just another marketing stunt.

No, thanks. I'll give this a pass.

If you want to say you're not like other girls and make a point of differentiating yourself from other people claiming to have made an LLM-generated desktop, the least you could do is take ten fucking minutes out of your supposed 6 month effort and write the description of what you've done yourself. When I see an obviously AI-generated website with obviously AI-generated copy, I have negative infinity reasons to believe you've put 6 months of human effort into this project. If you actually did, you're really not doing yourself any favors by not being willing to put a few minutes into presentation.

So this is yet another Linux desktop environment (DE) But with “AI”.

Nicely done! This is an example of how engineering discipline mixed with good agent guidance can avoid the slop.

A couple of things:

- The test pyramid (static -> unit -> functional -> VM release gate) works but is dependent on someone remembering to run it locally. Some kind of CI would help here.

- The C++ Swift interop layer has some huge files. Understandable given that you're porting the framework rather than reimplementing it, but right now it's the least documented, most god-object part of the codebase.