10,524 karma · joined May 25, 2019
The whole concept of it is a scam.
We know. But of course the hype of "microVMs" is just a rebranding of existing technologies with some modifications, macOS needed extra virtualization technologies just for containers for years:
Docker Desktop macOS (until 2025): QEMU + Virtualization.framework + Linux kernel (without extra drivers) = "microVM".
Now they use the native Apple virtualization libraries instead of QEMU for both containers and microVMs. [0] Linux never needed to use KVM for containers, but requires it for microVMs: Linux: KVM + Linux kernel (without extra drivers) = Firecracker "microVM".
In reality, it is different depending on the OS that you are using.[0] https://www.docker.com/blog/docker-desktop-for-mac-qemu-virt...
Now that we have a KVM 0day + VM escape vulnerability [0] right now.
You just answered your own question. Why would anyone pay for Turso if the core of the software is available for free and being licenced permissively under MIT?
The bots are now more intelligent at scraping at it is affecting everyone and it is not just Reddit.
Unfortunately it is slop, beyond the comprehension of the author unless they are experienced with DLSS internals to explain it in depth.
But the line is drawn when it involves CUDA and any part of their closed source compilers (nvcc).
There are obvious reasons why they are closed source, but it’s becoming pointless since Deepseek have open sourced their AI compiler and compute libraries with DeepGEMM and eventually they will catch up.
It is perfectly valid to have OSes that are more memory safe by default, and are also open source at the same time.
Might need a re-think about whether if Linux is still fit for purpose on sandboxing in the first place given its memory model is riddled with C-style security issues.
Seems like this hype around TUIs makes no sense at all and sounds like a fad. Maybe it is time to move on.
Why is it that iOS is crashing but not Android?
Google should know better to not have such an incident like this as the Firebase SDK runs on millions of apps.
If you did not implement a fall back mechanism, perhaps you based your testing assumptions on that “Firebase could never crash my app” which is now false.
This will be the new normal as all human “engineers” from staff to seniors are now down levelled to interns and junior engineers when using AI; unable to understand what they are doing and pushing broken updates like this into production with “agents”.
Better not blame Claude on this outage.