MayeulC

I think it should be fine with async reprojection.

Honestly, it wouldn't surprise me if this was a conspiracy to crash the "west" AI labs. Might as well pop the AI bubble and see the USA economy go down the drain. Even if not orchestrated, I am sure they see how they could benefit from that outcome.

Now, all this talk of pacing the frontier obviously means that they are afraid of the competition. It could be the open models eating their margins, but also competing frontier models forcing them to invest more and more for diminishing returns, just to keep up. They would certainly benefit from a "Moore's Law" roadmap to pace the advances, and seeing that they lobby for US laws, it would probably mean they are more worried about increasing spending. Though outlawing both open models and Chinese models would be good for their bottom line as well.

I think that "machine translator" is a much better metaphor: there is a lot more to translating a text than there is to compiling code: contextual cues, cultural settings. Various translations can be equally valid. All translations are imperfect.

Now, if you excuse my ramblings, here are a few ideas: I think we probably need new "programming" languages that are actually specification languages: reproducible default states, deterministic spec-to-code transformation.

I am not sure we actually need the determinism, but that would be a good property to have. At least with the same model/spec/temperature.

Now, prompts mutate the spec, which can also be edited by hand. You can already do this with the final code of course, but it is tedious as it contains many trivial implementation details.

But then, how do you handle bugfixes that need to persist during re-generation? Such as "Both Foo and foo can be present in the same directory if the file system is case-sensitive". If you add them to the spec, you are micro-managing implementation details again. So I think such "bug fixes" should be part of a prompt/spec that is automatically loaded when generating similar snippets (here, file I/O). It feels like I've just reinvented the concept of software libraries, though.

The four-vector contains coordinates of a point in time and space (x,y,z,t).

As a sibling said, light travels forward in time, but also from a light source towards the camera. Raytracing traces rays backwards from the camera (because there are fewer rays to trace this way: just pick the ones that land on the camera and work backwards towards light sources). So backwards in time, trying to figure where the ray originally came from.

Anyway, you need to keep track of the time dimension as light moves (at constant speed) through spacetime in a straight line, but the black hole also stretches the time dimension, affecting what a straight line looks like for an outside observer.

That's my understanding at least, I am not that well versed in relativity.

Edit to make it more interesting: (x, y, z, t) is your position in spacetime, I do not quite understand how you can take its derivative¹ (probably not time-wise), but you can derive the four-velocity, the magnitude of which is constant (c). Intuitively (and probably naïvely) I understand this as: you move trough spacetime at constant speed. Therefore, if your spatial coordinates move faster, your time coordinate moves slower. This gives you back the special relativity results with time that "slows down" for fast-moving objects.

https://en.wikipedia.org/wiki/Four-vector

https://en.wikipedia.org/wiki/Four-velocity

¹: derivated along the "proper time" (time from the perspective of the moving object) according to Wikipedia. Still unsure how to compute that.

Thanks for the feedback! A triggered power-on seauence for drives and fans makes sense, though I feel like this could also be solved by putting a couple of large capacitors on the 12V power rail (those need to charge as well, obviously, which might be an issue on startup).

In my experience, transcoding works fairly well over NFS (over fiber Internet). Mind you, I didn't try 80+ GiB media files.

Thanks for the data point, and the vote of confidence. I do not have the best setup to measure power draw (I just broke my dedicated smart plug), but it it may also increase power draw in my case (20 to 30 W, ERYING M-ITX 12500H + 2 nvme SSDs). Your numbers are impressive for 5 HDDs, are they 3.5"?

Anyway, it may be a transient that I am measuring, as the PC turns off during the boot sequence, after the CPU fan starts revving up. My 5A PSU may not be enough, I'll try limiting the TDP and the power brick.

I am most interested in that DC power supply. I bought a similar one recently (rgeek brand on aliexpress), but haven't yet dared to connect it yet.

I'm wondering what:

- the failure mode is if it can't provide enough current.

- How to properly test it under a fake load

- If these tiny power cables can really handle 150 W (this is 12 A at 12V). I don't have such a powerful 12V PSU, so I am thinking about adding capacitors to handle transients.

I imagine that it is really built to handle transient peaks to 150 W, with a more reasonable baseline.

Anyway, if anyone knows a good knowledge source for these PSU, I'd be interested.

My initial motication for going with these was:

- hopefully increase my PSU efficiency

- less cable clutter

- hopefully power it from a small DC UPS (there are plenty of "router" UPS sold on the market, though the power rating is a bit low, 18-36 W usually).

A more likely measure, which I recall seeing years ago, is to measure the impedance of the enclosure of the thing you want to protect. If someone tampers with it, you would be alerted. It works at many scales, from a protective metal mesh over your IC to a PC case.

Sure, but I would like to point out that gamenative largely exists thanks to FEX-Emu, which was started at Valve's initiative, and which Valve has stealthily funded since its inception.

This. This is what I am looking for. I remember using iz3d years ago, with their anaglyph renderer. Playing Half Life 2, Halo and others in stereoscopic 3D was a great experience. Dead Space was particularly nice, as the game uses no HUD.

I guess most of it could run as a Vulkan layer?

Usually the pixel density is higher in the center of the FOV though (more distortion due to the lenses on the sides). It's what set the original Oculus apart from existing headsets at the time: cheap lenses and correct the distorsiin in software. It's also a feature, as the actual density is better in the center :)

I do wish they used micro-LEDs for that kind of display though. I guess that they could be manufactured at the same time as a metamaterial acting as a lens sitting on top.

I do wonder if they use local upscaling after foveated rendering/streaming. The screenshot I've seen look a bit aliased in the periphery, which I hope isn't too distracting.

Just ask someone you trust to make the reservation from their account.

I'm thinking that this would be a great device to hand over to family members for when they need help with their computer. Ask them to plug them in & log-in remotely: may be easier than starting RustDesk, and can help with bios & bootloader options.

The harder part may be networking: it would be harder to ask them to wire it up to the network; but with Wi-Fi, I may be able to pre-fill their WPA key.

I'd be curious to know the compiler used by the Geekbench version they tested.

MSVC2019+ now include "volatile metadata" to the output executables to make emulation on ARM faster[0][1], though it seems to be off by default since MSVC2022, I am curious to know why they disabled it.

Microsoft now also offers an "Arm64EC" ABI, that, as I understand it, helps calling into native ARM libraries from emulated x86 code by avoiding the need to translate calling conventions, but that's mostly relevant for third-party libraries in this context (I imagine that Microsoft ships all system libraries with the "emulation compatible" ABI with Prism?).

I'd be curious to see some Prism vs FEX (vs Rosetta vs Box64) benchmarks.

[0]: https://fex-emu.com/FEX-2504/#windows-pe-volatile-metadata-s...

[1]: https://learn.microsoft.com/en-us/cpp/build/reference/volati...

Honestly, this whole KVM thread is a bit confusing to me. I've always equated KVM with keyboard/video/mouse switches: enabling multiple computers to be controlled by the same I/O devices. Nowadays I just use a multi-source monitor and USB switch for that (can even automatically control the monitor via DCC, depending on the switch state, look it up on HN).

And I usually refer to the "IP KVMs" as a "lights-out" module/engine as I was familiar with HP's "ilo (integrated lights out) [0]. More on lights out management on Wikipedia: https://en.wikipedia.org/wiki/Out-of-band_management

I was thinking of making my own with an ESP32 + serial, as that's enough in theory, though I am unsure how well-supported serial ports are in BIOS and bootloader. Also, IIRC RS232 is ±12V, so the ESP would need a level shifter.

TFA's device looks cheap enough (well, depending on shipping to EU), I may just get one. The screen is a bit superfluous though, and I think that the ATX power button support is a must.

[0] https://en.wikipedia.org/wiki/HPE_Integrated_Lights-Out

I tried that a few years ago. The result was worse than swapping. The kernel really tries its best not to kill anything, and you end up with "thrashing" again, where code pages (shared objects, executables, mmaped files) are evicted and need to be loaded again from disk. In practice the slowdown was worse than swapping to disk, and much more sudden. I'd recommend at least zram to soften the OOM landing.

Great build! I am certainly taking mental notes for when/if I get my own house one day. The biggest change compared to a regular house is putting the computers in a server room, and this can probably be retrofitted into almost any house during renovation work.

The other big thing in my opinion is the furniture that hides everything away when not in use. I am surprised it cost as much as you say. Well, I suppose that there is a lot of it. Taking a few mental notes here as well.

I have fond memories of university, where we would play UT99 (and DOOM I, believe it or not) in the computer room on Windows XP computers. Staff eventually found out that most of us were playing games over lunch break, so they tightened up the firewalls. However, one of the PC didn't have password protection on the BIOS, so I retrieved the necessary files, and cracked the admin password offline. Good times! The games went on for years.

Back then, around half of us were in a boarding school for a couple years, fully wired with Ethernet and no client isolation despite an extremely restrictive external firewall (that had us using free proxies ~15kB/s to launch Steam, I even saw some students play League of Legends with up to 13s of latency). But LAN was the thing. We had shared NAS, local bittorrent clients, and played soo much UT2004, some Sins of a Solar Empire, supreme commander, dawn of war II, minecraft & a few others.

Unfortunately this friend group is all over the place now, and I don't even allow myself to spend time playing single player games these days. Still, I hope that I'll settle down a bit at some point, and make more local friends. It's easier than people seem to think (for me at least), you don't have to be rich either. Meet a few people, meet their friends, etc.

We used to have LAN parties with my high school friends too, maybe we should try to reconvene, though much driving would be involved.

Even earlier in my life, I was figuring out the rudiments of networking (including these pesky crossover cables) playing Halo 2 with multiple Xboxes.

So yeah, this build resonates with me, thank you for sharing. I've had something similar in mind for a wile, though I think I'd do it a bit differently:

* probably use Linux as a host. On a single or a couple of servers: older games aren't too demanding, and half the fun is in the nostalgia.

* If using a single host, a single link is needed for USB, and can be split over multiple hubs in the gaming room.

* Maybe use GOG installers, they work well for LAN: no DRM, no Internet access required. People can buy a copy if they like the game.

It's tempting to multiplex displays and I/O over the network, but I am concerned about the necessary bandwidth and latency increase. It's also feasible to send all the data in a single optical fiber over multiple wavelengths, however I just looked it up: DP-SFP converters (fiber KVMs) seem to cost more than $1k, so $30 cables are a better option.

You are buying a copy though, and would be expected to have the right to resell it. So digital stores go out of their way to "license" the copyrighted work instead of selling you a copy.

I have been thinking about this lately, and figured that another difference with classic media was the "live service" model, where games receive updates.

You can't take for granted that a copy comes with updates, so I guess that as a middle ground, publishers could license updates (and online multiplayer) & sell copies of the game. Sell your copy, fine, but if you want updates or multiplayer, you have to purchase a subscription again. This would make more sense to me, but may be difficult to communicate and implement.

> We initially considered limiting transcoding to popular content

Weird, I would have compressed cold content instead, if the goal was to save on CPU time during decode.

Yes, this could have been a 5-sentence PSA. I found the article hard to read, with comparatively little content for the length.

Probably not with DIMM modules, as longer traces mean higher latency (speed of light is ~30 cm in 1ns). GDDR typically uses larger buses (more wires) for higher bandwidth, even more so for HBM, so DIMM would be hard. Maybe CAMM would be up to the task?

It certainly seems feasible from an engineering perspective (though it does make cooling harder), at least for mid-range, not H100-class HW, but it prevents market segmentation, so EOMs may not be too interested (as long as no competitor does it).

I think it's generally OK if you leave the shielding unconnected at one end (good practice anyway to avoid ground loops). There is an electrically-insulating transformer in the connector (nothing that would resist a lightning strike, of course, though some surge protectors have 8P8C connectors).

+1 for fiber though. Long fibers are cheaper, SFP connectors and modules are commonplace and have become quite cheap. Just need to learn a few new things, such as simplex/BiDi vs duplex fiber, and multimode vs mono mode.

UV can be built (or downloaded) as a single optionally statically linked with the libc executable file, with no other dependencies. That makes it very easy to distribute on weird systems.

Oh, I had no idea you could perform such a "broadcast" ping on link-local addresses. Now, that is a nice trick that will save me some typing between computers.

mDNS / Avahi is also great if the machine advertises itself: just use machinename.local after plugging it.

Though the interface needs to be configured for link-local addresses. In my experience, NetworkManager now periodically resets the link if it can't get a DHCP server to respond. I understand the rationale, but it used to be simpler to plug into a headless machine and be confident that you could just ssh in.

Just a token fee like $5 would discourage spamming claims, though it does make fighting against re-uploads of the same content harder (the power balance isn't too bad for the play store where there is an entry fee for publishers as well).

AFAIK (at least it worked like that some 10 years ago) the captive portal just intercepts the HTTP page load and inserts its own content (most often a 302). So it just has to be a http web page. Firefox uses http://detectportal.firefox.com/canonical.html

Relevant support page, though light in details: https://support.mozilla.org/en-US/kb/captive-portal

Edit: ah, yes, DNS can be hijacked too (requires intercepting outgoing traffic on port 53 therefore incompatible with DoH), that may require fewer computing resources. Still need http otherwise the server cannot use the correct cert chain.

Edit 2: Wikipedia says both methods are used: https://en.wikipedia.org/wiki/Captive_portal and also mentions RFC 8910. I suspected something like that existed, hence my initial disclaimer.

My point was: that domain is not treated any differently from other domains.

I once forgot my phone in a French airport (Lyon). Back then I had a "bulky" (hardened, like the crosscall brand) phone. I had trouble finding it in their system as they logged it as having a phone case.

I initially put up a request in their system, which did not match anything when they found it. I ended up finding it in their "public" listing (not really public, I think they ask for a ~10€ registration fee).

Anyway, when I got it back, I tried to argue with the employee that the phone did not, in fact, have a case. But they were adamant that it did. I ended up shrugging it off, happy enough to have my phone. But some people really don't care. Maybe someone filed your bag as something else, and then it wouldn't appear in the system, and weren't bothered enough to go look inside.

The kit I bought 10 months ago had the issue. Luckily I was only unable to read some chips, and did not fry anything. Definitely check it before using it.