dist1ll

10 ms·

61.
by dist1ll 2y ago
I mean, the fiddly-ness comes from wanting to use inline assembly, which is often considered to be niche and second-class in most modern languages. Even writing the asm in the first place is full of footguns [0]. There are ways to reduce th
62.
by dist1ll 2y ago
Writing inline assembly in most languages is a pain. I think there's some real opportunity to make things more ergonomic. Basically taking the idea of HLASM and other high-level assembler projects, but actually bringing them into the m
63.
by dist1ll 2y ago
> BFT-style consensus such as Paxos and Raft Raft cannot deal with byzantine failures, so it's not a BFT protocol. Neither is the original Paxos, unless extended to Byzantine Paxos. Both are CFT by default. > So, we would not exp
64.
by dist1ll 2y ago
Well, clearly I don't need 99% of whatever the bank is doing to send a transaction between two accounts I own. Add to that all the things that many banks continuously fail to provide, like proper OpsSec, transparent fees, full progra
65.
by dist1ll 2y ago
To give one more counter-point to your "fraud costs are baked in": Self-custody as is common in crypto isn't really a thing in traditional finance (for most people anyways) - having your money backed 1:1 by cash is an extreme
66.
by dist1ll 2y ago
Many of the things I mentioned are absolutely not solved problems. > Payments require CVV Many companies regularly violate regulations and store CVVs against your will. How do you think millions of CC CVVs get leaked in many of these maj
67.
by dist1ll 2y ago
Assuming we're talking about low-cost chains that can actually be used for high-speed payment (which basically rules out L1 ETH/BTC) * Sending money without paying an obscene amount of settlement fees and having to wait in some ca
68.
by dist1ll 2y ago
What latency range are your strategies aimed at? Are you executing strategies on CPU, or lowering the DSL to some kind of HDL or ASIC config?
69.
What Ownership Is About: A Mental Model Approach(modular.com)
1 points by dist1ll 2y ago|0 comments
70.
by dist1ll 2y ago
Sipeed's Tang FPGAs are really interesting. I've been eyeing the Tang Mega 138k because of its 10GbE ports, especially considering how reasonable the price is.
71.
by dist1ll 2y ago
I would say the compiler explorer[0] fits the definition perfectly. It may seem like a straightforward piece of software, but it has immensely changed the way people discuss and share knowledge around compilers and performance optimization.
72.
by dist1ll 2y ago
Aren't big players colocating their CDNs at internet exchanges? Bandwidth should be essentially free for content delivery.
73.
by dist1ll 2y ago
That's my thinking as well. Any library that exposes the such shared memory IPC via shared references as a safe API seems fundamentally unsound.
74.
by dist1ll 2y ago
> You can have perfectly safe Rust code with data races No, you can't. Data races are explicitly one type of race condition that Rust protects you from. Anything else would indicate an unsound library or compiler bug. For reference,
75.
by dist1ll 2y ago
I wonder how much of that is due to the kernel history, and the influence of C idioms, and not because of some inherent design superiority. I'd be convinced once I see pure Rust kernels geared towards modern machines suddenly using lin
76.
by dist1ll 2y ago
There's definitely a bunch of accidental complexity in systems programming. Even in domains with extreme performance constraints, large portion of the code base could be expressed with simpler and safer semantics like MVS.
77.
by dist1ll 2y ago
One of the main contributors gave an excellent talk [0] on Verus at the Rust meetup in Zürich. I was really impressed how clean this "ghost" code fits into programs (reminded me a bit of Ada). [0] https://www.youtube.co
78.
by dist1ll 2y ago
Discussed this in a comment on lobste.rs https://lobste.rs/s/ljlmid/life_times_abstract_syntax_tree#c... tl:dr; use indices in Rust, they work pretty well, and proper use of the type-system helps with reducing ove
79.
by dist1ll 2y ago
That's a good example. I actually think dot syntax is really under-utilized sometimes. Although personally, I'd prefer that if v was a reference/pointer to a struct, that v.b simply performs a pointer offset, instead of auto-
80.
by dist1ll 2y ago
Wow, I had no idea! OpenBMC support even? That sounds like a decent platform to hack on... I got curious about the BF3 because of the 256-thread RISC-V data path.
81.
by dist1ll 2y ago
The BF3 is a DPU, not an SBC right? Or can it be operated without a host? I'm curious about your setup.
82.
by dist1ll 2y ago
No PCIe connectivity is very unfortunate. I think these ARM/RISC-V SBCs are very attractive in combination with high-throughput NICs (or if you're into AI, things like the Tenstorrent cards). That'd be very nice for building
83.
by dist1ll 3y ago
That's definitely a trend we're moving towards for extremely high-performance software. General-purpose operating systems often don't sit at the right level of abstraction, and lack the flexibility for certain demanding workl
84.
by dist1ll 3y ago
That's what makes me wary of modifying my NixOS config. A single typo and you get an error dump comparable to C++03 templates.
85.
by dist1ll 3y ago
RPS meaning reads or writes? What's the distribution of message sizes, and how large is your total dataset? What specs (core count, NIC) did each node have? I'm asking because without this info, RPS is not a particularly useful me
86.
by dist1ll 3y ago
Note: unaligned doesn't always mean cache-line splitting. Unaligned 64-bit loads & stores within a cache line incur no performance penalty on modern Intel architectures IIRC. Also, reading a cache-line in shared MESI state should
87.
by dist1ll 3y ago
Great use of bpftrace! eBPF-based tracing is really useful for understanding multithreaded performance in the kernel. E.g. if you're running into kernel-level lock contention, the preemption time won't show up in the perf report (
88.
by dist1ll 3y ago
Wouldn't it be better to expose 128-bit vector registers for this purpose? Like how the aarch64 module exposes int16x8_t. That seems much better than relying on a generic 128 bit type, because the use-case is clearly specified.
89.
by dist1ll 3y ago
> embedding [..] never feel quite right. funnily enough, I would argue that struct embedding is one of the things that Go actually got right . It's simple and elegant, and I think it could've been a useful addition to Rust.
90.
by dist1ll 3y ago
I'm curious, what is your frame of reference with regards to maximum speed of building inverted indices? Like, what is the maximum throughput you'd expect for this type of task, and what is your reasoning for it?

More ›