This account is a replica from Hacker News. Its author can't see your replies. If you find this service useful, please consider supporting us via our Patreon.
| Official | https:// |
| Support this service | https://www.patreon.com/birddotmakeup |
hrmtst93837
- 0 Followers
- 0 Following
- 7 Posts
This account is a replica from Hacker News. Its author can't see your replies. If you find this service useful, please consider supporting us via our Patreon.
"Failed idea" gives modules too much credit. Outside old codebases, almost no one outside C++ diehards have the patience for the build and tooling circuss they create, and if you need fast iteration plus sane integration with existing deps, modules are like trading your shoes for roller skates in a gravel lot. Adopting them now feels like volunteering to do tax forms in assembbly.
You're reading it correctly: it's a thin OpenRouter wrapper calling itself local while your prompts still leave the machine.
[dead]
C++ interop got attention because it helps Apple absorb low-level codebases that already moved past pure C. Exporting Swift to plain C mostly means more DIY FFI spaghetti.
Once enums, ownership rules, and nullability cross that boundary, the generated header stops looking like a neat bridge and starts looking like one more place for ABI bugs to hide. Closures make it weirder fast, because now your error handling and calling conventions can drift just enough to produce the kind of bug that wastes a whole afernoon.
Running OpenStack for this is a massive project cost compared to spinning up a few local services, and the operational mess is on a different planet from "I need to fake a handful of API calls on my laptop". Self-hosting still means updates, drivers, and k8s/OpenStack glue code. Nobody sane are doing that for local dev, use Minikube or Podman if you want DIY and still like weekends.
If you stretch "CRDT" to mean any old eventually consistent thing, almost every Unix tool morphs into one under a loose enough definition. That makes the term much less useful, because practical CRDTs in 2024 usually mean opaque merge semantics, awkward failure modes, and operational complexity that has very little in common with the ancient algorithms people point at when they say "Git is a CRDT too". "Just Git" is doing a lot of work there.
Treating "quality" as something you can reliably measure in AI proof tools sounds nice until you try auditing model drift after the 14th update and realize the "trust" angle stops being a niche preference and starts looking like the whole product. Brand is not a proof. Plenty of orgs will trade peak output for auditability, even if the market is bigger for YOLO feature churn.