destroycom

7 ms·

1.
by destroycom 7mo ago
> I think you should remove Claude as a contributor to your repo I actually really appreciate it when people do not hide their use of Claude code in their repo like that. It's usually the first thing I check on Show HN posts these d
2.
by destroycom 7mo ago
> and not some Electron thing Ironically, most of the app is a webview. The comments just have some additional CSS styling slapped on top of the hackernews website. So you still have an entire HackerNews site loaded at all times when rea
3.
by destroycom 8mo ago
Sadly, their "native" client is a web browser. Besides the usual complaints about electron and CEF applications, another pain point is they work horrendously in emulation. GoG Galaxy is only available as an x86 application on Wind
4.
by destroycom 1y ago
This doesn't seem like an article that was made with proper research or proper sincerity. The claim is that there is one million times more data to feed to LLMs, citing a few articles. The articles estimate that there is 180-200 zettab
5.
by destroycom 1y ago
> Written in Zig for optimal performance and memory safety Ironic to see this, as even a cursory glance over the code shows a lot of memory bugs, such as dangling pointers, data races, some potential bugs like double free, etc. It is goo
6.
by destroycom 1y ago
> "Which standard library should I use?" is not a question most languages have: > https://stackoverflow.com/q/693672/265521 There is no such question when using D2 either. It was only an issue with
7.
by destroycom 1y ago
This isn't mentioned anywhere on the page, but fork is generally not a great API for these kinds of things. In a multi-threaded application, any code between the fork and exec syscalls should be async-signal-safe. Since the memory is r