sanxiyn

33 ms·

601.
by sanxiyn 4y ago
As I understand, the only ones' complement architecture in active use is UNIVAC (yes, believe or not UNIVAC is in active use and Unisys provides commercial support). Non-8-bit char is a bit more common, probably the most common is TMS3
602.
by sanxiyn 4y ago
I provided the concrete specification. What is hard about it? Get it done already.
603.
by sanxiyn 4y ago
This already exists. Don't write standard C, avoid it like the plague. Compile with -fno-strict-overflow -fno-strict-aliasing -fno-delete-null-pointer-checks, like I do, like Linux kernel does, and like everyone sane does.
604.
Falsehoods programmers believe about undefined behavior(predr.ag)
151 points by sanxiyn 4y ago|226 comments
605.
by sanxiyn 4y ago
To disappear: ICE car mechanics. Sure, we will need a few to service museum pieces, but it won't be a big occupation like it is now.
606.
by sanxiyn 4y ago
I compile all my C code with -fno-strict-overflow and -fno-strict-aliasing, and I recommend you to do so as well. C standard committee and GCC and Clang are being stupid, but that does not mean you should suffer their stupidity.
607.
by sanxiyn 4y ago
No. Just compile with -fno-strict-overflow, like I do.
608.
by sanxiyn 4y ago
The lesson is you shouldn't use standard C, until the standard is fixed. Until then, compile with -fno-strict-overflow, like I do.
609.
by sanxiyn 4y ago
Yes there is. Compile with -fno-strict-overflow.
610.
by sanxiyn 4y ago
These also work on Clang.
611.
by sanxiyn 4y ago
> just let it overflow I agree with this. Making integer overflow UB is ridiculous. > and don't fuck up any bounds checks But I disagree with this. That's just normal UB-based optimization, one of the point of having UB at a
612.
by sanxiyn 4y ago
macOS includes launchd, because it is in fact necessary.
613.
by sanxiyn 4y ago
Fedora made nano default in 2020: https://fedoraproject.org/wiki/Changes/UseNanoByDefault
614.
by sanxiyn 4y ago
systemd was necessary. In fact, Solaris did it first in the form of SMF.
615.
by sanxiyn 4y ago
For Linux kernel's viewpoint where they are trying something new (io_uring_spawn) due to fork's problems, see https://lwn.net/Articles/908268/
616.
by sanxiyn 4y ago
I think GCC had an edge up to 2020 or so, mainly thanks to its better inliner. I should check this again now LLVM's inliner rewrite landed.
617.
by sanxiyn 4y ago
Microsoft people explained it better than I could, so I link to their explanation: https://www.microsoft.com/en-us/research/publication/a-fork-...
618.
by sanxiyn 4y ago
I respect Windows for saying no to a hack that is fork. Fork is just bad design and no software should use it, and thanks to existence of Windows, lots of software avoid use of fork, which is a big plus to software ecosystem as a whole. Tha
619.
by sanxiyn 4y ago
Struct should be compared field by field, using memcmp is just wrong. What if there is padding?
620.
by sanxiyn 4y ago
"Linux is not Unix" is useless pedantry, but sure, Solaris SMF did it earlier than systemd, both dependency based boot and parallel boot.
621.
by sanxiyn 4y ago
Many important things happened to Unix in 1990s and 2000s. dpkg and apt come to mind for 1990s, and udev and systemd for 2000s.
622.
by sanxiyn 4y ago
OCaml supports mutability well. Basically, you can write OCaml as if you are writing Nim.
623.
by sanxiyn 4y ago
I agree Humans Only Clause does not prevent Microsoft from training Copilot from codes on GitHub due to GitHub's terms of service, but I think it does prevent, say, Salesforce from training its CodeGen model. So if the clause is widely
624.
by sanxiyn 4y ago
Why shouldn't it matter? Many open source licenses require attribution, so it is reasonable to think one point of making software open source is to get attribution. Generative AI prevents getting attribution, so it matters whether it i
625.
by sanxiyn 4y ago
Yes.
626.
by sanxiyn 4y ago
What lessons did it learn?
627.
by sanxiyn 4y ago
It doesn't make much sense because if you want a universal runtime, JVM is a better choice.
628.
by sanxiyn 4y ago
JVM also can be targeted by C/C++ and Rust. In fact NestedVM was published in 2004... I concede WebAssembly does sandboxing better.
629.
by sanxiyn 4y ago
No, intent labeling wasn't manual, it was done automatically from raw data.
630.
by sanxiyn 4y ago
They do have structured game model although it doesn't capture everything in chat. Language model still had lots of problems with consistency even with structured game model input.

More ›