the8472

11 ms·

181.
by the8472 1y ago
rodent -> homo sapiens brain scales just fine? It's tenuous evidence, but not zero.
182.
by the8472 1y ago
> As of very recently, you can also force EC2 instances to stop or terminate without waiting for a clean shutdown or a ridiculous timeout Not true for GPU instances, they're stuck 5 minutes in a stopping state because they run some
183.
by the8472 1y ago
io_uring alleviates a lot of those problems by essentially providing a batched, asynchronous syscall API, including futex calls, as mentioned in other comments.
184.
by the8472 1y ago
But still no futex_swap[0][1] [0] https://lore.kernel.org/all/414e292195d720c780fab2781c749df3... [1] https://blog.linuxplumbersconf.org/2013/ocw/system/presentat...
185.
by the8472 1y ago
Yes, and? Human eyes also have limited instantaneous dynamic range much smaller than their total dynamic range. Part of the mechanism is the same (pupil vs. camera iris). They can't see starlight during the day and tunnels need adaptio
186.
by the8472 1y ago
Ah yeah, that's making even more assumptions. Not only does it assume the cameras are powerful enough but that there already is enough compute. There's a sensing-power/compute/latency tradeoff. That is you can get away w
187.
by the8472 1y ago
Human eyes are unlikely the only thing in parameter-space that's sufficient for driving. Cameras can do IR, 360° coverage, higher frame rates, wider stereo separation... but of course nothing says Teslas sit at a good point in that spa
188.
by the8472 1y ago
It could be emulated in userspace with F_DUPFD. But that's costly because the kernel table is optimized for dense data, not sparse. The rust standard library aborts the program in debug builds when it detects a double-close; at that po
189.
by the8472 1y ago
> In a way it's not a fair comparison, Indeed. And the comparison is unnecessarily unfair. You're comparing the dynamic range of a single exposure on a camera vs. the adaptive dynamic range in multiple environments for human ey
190.
by the8472 1y ago
lowest-number-reuse is also a robustness issue. If multi-threaded programs UAF or double-free a file descriptor they likely end up touching FDs owned by other parts of the program which can result in various kinds of corruption and memory-u
191.
by the8472 1y ago
The mozilla devs were warned that mandatory addon signing and code verification turns them into the main arbiter which extensions may exist and thus a legal target. They insisted that for the sake of security it must be done anyway. The spi
192.
by the8472 1y ago
If we're counting all the phases then european homes get 400V 3-phase, not 240V split-phase. Not that typical residential connections matter to highend servers.
193.
by the8472 1y ago
Byzantine Failure seems more appropriate.
194.
by the8472 1y ago
Commoditize your complement.
195.
by the8472 1y ago
During peaks they can kick out background jobs like model training or API users doing batch jobs.
196.
by the8472 1y ago
When I was on an ISP with DS-Lite the IPv4 functionality regularly failed because the AFTR's port mapping saturated (equivalent to reaching ip_conntrack_max on linux). IPv6 wasn't affected since it doesn't involve a stateful
197.
by the8472 1y ago
Well, then stop doing those transactions in <country>, not globally. Why am I not allowed to buy something just because some organization in <country> threatens lawsuits there or whatever? In other cases multi-nationals (e.g. AW
198.
by the8472 1y ago
It's easier since you don't don't have to deal with symmetric nat, external IP address discovery and port mapping.
199.
by the8472 1y ago
Like I said, that smell like anti-competitive behavior. It basically means the payment processor denies competitors a business opportunity for cases they don't want to service.
200.
by the8472 1y ago
> I don't think it's realistically viable to compete with Steam (or Itch) without access to Mastercard and Visa. They could not allow those games to be sold through those particular payment processors and require wire transfers
201.
by the8472 1y ago
It was an optimization bug in llvm. https://godbolt.org/z/GqdPGarsn We filed an issue, this one is fixed now. https://github.com/llvm/llvm-project/issues/65763
202.
by the8472 1y ago
> Adding the separation of lightyears seems to make people talk as if they do know what will happen in the future, which we don’t. Perhaps a distinction is that we like to think that at least in principle we can influence almost any futu
203.
by the8472 1y ago
> So it would still be 500 years between message and Death Ray. Yes of course, but you'd still act as if the death-ray is on its way, i.e. firing it already happened 500 years before you received the message. The systems may be caus
204.
by the8472 1y ago
You receive a message from an an alien civilization 1000ly away that they have started a deterministic process that will shoot a death ray at earth in 500 years. Should you act on the information that there's doom heading for earth rig
205.
by the8472 1y ago
The tractor wheel, the race bicycle wheel or the 50000RPM flywheel?
206.
by the8472 1y ago
I think steamos would be a better example, also mentioned in the article. Focus on a single application by default, immutable A/B OS filesystems, but can be opened if desired.
207.
by the8472 1y ago
Good at autovectorizing? Eh, if you massage your code in godbolt until it kicks in. But just naively writing code often falls off the happy path. And even when it autovectorizes it can sometimes produce such gems as vmovdqu xmm1,
208.
by the8472 1y ago
NP
209.
by the8472 1y ago
but then someone wants statx...
210.
by the8472 1y ago
io_uring doesn't support getdents though. so the primary benefit is bulk statting (ls -l). It'd be nice if we could have a getdents in flight while processing the results of the previous one.

More ›