HNHacker News
TopNewBestAskShowJobs

tliltocatl

1,072 karma · joined June 23, 2024

submissionscomments
tliltocatl··on I was recently diagnosed with anti-NMDA receptor encephalitis
A lot of gender disparity is cardiovascular disease being 50% more likely to be diagnosed in men than in women of same age, so not directly related to risky behavior.
tliltocatl··on Project Is Dead. Completely
I wonder, what did they expect? The whole thing was basically about "fuck regulations", which one might or might not consider a worthy goal. But how could one expect the regulators to be complicit?
tliltocatl··on ESP32-S31
Is there any reason you want wireless? Bluetooth audio is a disaster, AFAIK. You don't want to use it for music. Just go wired, the ether is too cramped already.
tliltocatl··on The dead economy theory
But that's core product. That's how they make money. Whatever it's a good idea to make money that way is a different story.
tliltocatl··on Adding Reflection to C
Been there… Pendantic nitpick: it's more akin to introspection, reflection would be self-modifying code.
tliltocatl··on AMD pulls a bait-and-switch on Linux users with Vivado licensing changes
Xilinx has the best silicon. Everyone else is behind. Altera is basically dead thanks Intel. Lattice is nice for low power but performance-wise they are behind. Don't know much about Microchip, but from the little I've heard their tooling is a disaster even by the standards of FPGA tooling. Then there are Gowin (not bad, but Chinglish docs and everything), Gatemate (pretty innovative and vendor-backed nextpnr support - but only one low-mid FPGA with a promise to release chiplet assemblies of it latter). And Effinix - don't know much about them, do anyone have experience?
tliltocatl··on AMD pulls a bait-and-switch on Linux users with Vivado licensing changes
The software is useless without their chips and the chips cost a fortune. It's not "business model no longer supports giving away free stuff". It's just bean-counters cutting corners.
tliltocatl··on YouTube to automatically label AI-generated videos
> - AI upscaling of low res footage

I'd say this is far more problematic than plain slop.

tliltocatl··on Stack Overflow’s forum is dead but the company’s still kicking
Without continuous feedback from real world, lower-probability token (and soon high-probability ones as well) will be complete garbage.
tliltocatl··on Show HN: GuideOS – A radar-first, off-grid navigation kernel for edge robotics
Looks like a typo (extra H at the end of URL).
tliltocatl··on Why is Vivado 2026.1 dropping Linux support for free tier?
I guess FPGA division (nee Xilinx, which was always a bit sketchy, even if they had best silicon) doesn't learn much from the GPU division.
tliltocatl··on Bun's unreleased Rust port has 13,365 unsafe blocks
Just write C then.
tliltocatl··on "Don't Be Evil" Was a Lie from the Start (2024)
> What could be more anthropomorphic than "social structures"?

Group of humans doesn't behave like individual humans do. And we can't have same ethical expectations towards groups as we do towards individuals. As simple as that.

tliltocatl··on Why does the arrow (->) operator in C exist?
That's the most annoying part of C. And yes, "I hate implicit" people, I can hear you, but this hampers refactoring A LOT. The number of times I've found myself in "need to move this piece into this function, but then this struct becomes a pointer and I've fix the dots and arrows"…
tliltocatl··on "Don't Be Evil" Was a Lie from the Start (2024)
Don't anthropomorphize the grass cutter. All social structures, be it for-profit corporation or something else, are ultimately like this. Use the good thing while it is still there, ditch it once they pivot to being assholes. No hard feelings.
tliltocatl··on GitHub confirms breach of 3,800 repos via malicious VSCode extension
> full project code search

grep

> go to definition

/definition

Not even using vim, I'm more of a micro person.

tliltocatl··on Kioxia and Dell cram 10 PB into slim 2RU server
Data retention is probably unusable for archival purposes.
tliltocatl··on Ask HN: We just had an actual UUID v4 collision...
Meh. The problem is that it might start receiving you local radio station and end up deterministic enough to screw you. So you need to shield the dirt properly.
tliltocatl··on Inventing Cyrillic (2024)
Х is ch in Bach rather than h, and apparently, anglophones can distinguish those. I can't.
tliltocatl··on Inventing Cyrillic (2024)
It absolutely doesn't as pronounced now, yet Thomas is Фома, Theodor is Феодор, etc. Just like Hertz is Герц, even through Г and H are as far from each other as one can get.
tliltocatl··on The USB Situation
Besides what retatop and crote said - USB4 uses either binary or PAM-3, 10GBASE-T uses PAM-16. Higher modulation means lower frequency bandwidth (so longer and/or crappier cable), but also more current-hungry line driver and more current required to keep noise manageable.
tliltocatl··on The USB Situation
I'd rather prefer Ethernet over USB 2.0. Full duplex, galvanically isolated and cheap fiber converters, (relatively) lightweight host implementation possible, no "transaction translator" madness (at the cost of slightly more expensive hubs). Too bad it is such a power hog.
tliltocatl··on The USB Situation
TRRS might work for power, but it sucks for signal integrity - ergo no high-speed for you. And 4 wires is nowhere enough. You need two for each differential pair. No, half-duplex is absolutely not okay, it's the worst design decision in pre-SS USB.
tliltocatl··on Open source ballistic simulator with NASA SRTM terrain masking (Python/C#)
It is as cool as the fact that it is relevant is sad and scary.
tliltocatl··on Your phone is about to stop being yours
Before IBM PC computer's weren't particularly commonplace outside of ivory towers either.
tliltocatl··on XOR'ing a register with itself is the idiom for zeroing it out. Why not sub?
- Even each instruction does a lot of work, it is supposed to do it in parallel, so time available to fetch the next instruction is (supposed to be) the same.

- Not everything is parallelisable so most of instructions words end up full of NOPs.

- The real problem are data reads. Instruction fetches are fairly predictable (and when they aren't OOO suck just as much), data reads aren't. An OOO can do something else until the data comes in. VLIV, or any in-order architecture, must stall as soon as a new instruction depends on the result of the read.

tliltocatl··on 5x5 Pixel font for tiny screens
At that point, just go Braille! Amazing work anyway.
tliltocatl··on XOR'ing a register with itself is the idiom for zeroing it out. Why not sub?
Not really. Itanium was a result of some people at Intel being obsessed by LINPACK benchmarks and forgetting everything else. It sucked for random memory access, and hence everything that's not floating-point number-crunching. Compiler can't hide memory access latency because it's fundamentally unpredictable. VLIW does magic for floating-point latency (which is predictable), but

- As transistors got smaller, FP performance increased, memory latency stayed the same (or even increased).

- If you are doing a lot of floating point, you are probably doing array processing, so might as well go for a GPU or at least SIMD).

- Low instruction density is bad for I-cache. Yes, RISC fans, density matters! And VLIW is an absolute disaster in that regard. Again, this is less visible in number-crunching loads where the processor executes relatively small loops many times over.

tliltocatl··on XOR'ing a register with itself is the idiom for zeroing it out. Why not sub?
On superscalars running xor trick as is would be significantly slower because it implies a data dependency where there isn't one. But all OOO x86's optimize it away internally.
tliltocatl··on XOR'ing a register with itself is the idiom for zeroing it out. Why not sub?
I don't think that's how it works.

- Normally ALU implements all "light" operations (i. e. add/sub/and/or/xor) in a single block, separating them would result in far more interconnect overhead. Often, CPUs have specialized adder-only units for address generation, but never a xor-specialized block.

- All CPUs that implement hyper-threading also optimize a XOR EAX,EAX into MOV EAX,ZERO/SET FLAGS (where ZERO is an invisible zero register just like on Itanium and RISCs). This helps register renaming and eliminates a spurious dependency.

- The XOR trick is about as old as 8086 if not older.

← PreviousPage 4 of 16Next →