Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
sliken
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
49 ms
·
721.
▲
by
sliken
4y ago
Sure, that helps, for looking up a single file. Doesn't however help with ls or du. Even things like what are the 10 biggest files in this directory are painful. I've seen numerous efforts (Microsoft and BeOS spring to mind) to
722.
▲
by
sliken
4y ago
Well the curvature of the orbit at the earth <-> sun distance is minimal. So you have to hit a decent fraction of the speed of light to create that centrifugal force to replace the earth normal 1G. Those incredible speeds mean that
723.
▲
by
sliken
4y ago
"Experimental data shows that the observable universe is flat", agreed. However the research put a maximum bound on it's non-flatness. So it's low, but it's impossible to prove it's perfectly flat.
724.
▲
by
sliken
4y ago
Sure, it might stay centered on the star if it started centered. Any mild disturbances would cause the closer parts of the ring to receive more solar wind and light pressure while the more distant parts of the ring would receive less. How
725.
▲
by
sliken
4y ago
I think you are missing that the gravity comes from the centrifugal forces. Said gravity is for humans to be able to walk around, as well as keeping air and water from being lost to space. So sure you could make a non-spinning ring that mi
726.
▲
by
sliken
4y ago
So maybe it would work with a red dwarf with much less gravity and brightness. As a bonus the star would last 10-1000 times longer.
727.
▲
by
sliken
4y ago
Be warned that even 20 years ago when I was doing research on this stuff, that it was standard procedure to exploit the kernel. This included things like read or stat showing a valid file, but exec would run an exploited file. It also all
728.
▲
by
sliken
4y ago
So sure, it feels great and provide predictable braking. But is harder to maximize range and minimize brake wear and brake dust. I generally drive only using regen, because I can tell. If you remove the feedback it's going to be hard
729.
▲
by
sliken
4y ago
Not so sure. On other cars say you want to maximize your range and minimize brake wear/dust. How do you drive? On a Tesla you just lift the throttle, and if that's not enough you are aware of it, and change your driving patterns
730.
▲
by
sliken
4y ago
Be fair, those aren't comparable cars. BMW Model 3 and the Tesla model 3 are pretty similar in size, weight, and cost. The fastest of both is AWD and the Tesla is 4072 pounds and the M3 Competition and M3 Touring (the 2 variants with
731.
▲
by
sliken
4y ago
If you use both pedals at once it will reduce acceleration and warn you.
732.
▲
by
sliken
4y ago
Well regenerative braking is limited and even more limited when cool. So Tesla's trying to train you to use the brake pedal when you want stronger braking. Makes sense to me, last thing I want is to get half as much braking as I expe
733.
▲
by
sliken
4y ago
Indeed, but the man hours involved was quite surprising to me. Some work was per platform, some per cellular provider, some per country, etc. Also it's less confusing for users if they can say "if you use signal, it's encryp
734.
▲
by
sliken
4y ago
Maybe it's the long rumored locally produced System76 laptop.
735.
▲
by
sliken
4y ago
Years ago the GCC FAQ had an entry for a strange compiler error, that pretty much was only triggered by overclocking or memory errors. I just looked and couldn't find it.
736.
▲
by
sliken
4y ago
I agree with the "if a machine passes a memtest scrub for 24+ hours then it's probably fine" in theory. Sadly in practice I regularly find dimms that generate errors plenty of errors, and test find with memtest. Not sure why
737.
▲
by
sliken
4y ago
Heh, well there's 2 things I wanted from ECC. Sure I wanted the very occasional random bit flip to be caught and fixed. Making the system more reliable, maybe save a few process crashes or reboots over 10 years. Although I was much m
738.
▲
by
sliken
4y ago
In my experience computers last longer when cool, and even more important is a minimum number of heat cycles. Things like capacitors on motherboards use chemicals that break down over time, and more quickly if they are hotter. Sure you ca
739.
▲
by
sliken
4y ago
Not sure if I'm lucky, or if things have improved. But I've probably ran it at all cores busy compiling for 100 hours so far, and had uptime in the weeks, currently at 5 days. I did use a beefy usb-c power brick, rated for 40 wat
740.
▲
by
sliken
4y ago
Last I heard they were making 7M a year, with very few reserved for end users. To improve the situation they promised 100k reserved for end users. Seems much like the GPU fiasco where a company reserves all their product, leaving none for
741.
▲
by
sliken
4y ago
I didn't realize how big a deal the natural gas related pollution was. As in the post I found out my rental wasn't doing what it should. Turns out a hail storm had crushed the hat above the water heater exhaust. Turns out the r
742.
▲
by
sliken
4y ago
Electric stoves used to suck, low powered, annoying, and they made temperature control difficult. New induction units are more powerful than gas, respond more quickly, and have temperature based controls. So you can blast and boil a quart
743.
▲
by
sliken
4y ago
My house has a large air volume, the 2nd floor is half as big as the first, so half the first floor has 20 foot ceilings. I don't monitor NO2, but do monitor PM2.5 and CO2. I see no difference in PM2.5, but do see my baseline of 550-
744.
▲
by
sliken
4y ago
Not yet, all NanoPi images use the same not upstreamed 5.10.110 kernel. However progress has been made, various reports of it working, some progress, some accepted patches, some rejected. More info at: https://forum.radxa.com&#x
745.
▲
by
sliken
4y ago
For what? As a router or small server for running DNS, PiHole, and related services I'd recommend one of the Rk3588 systems, I bought a NanoPi R6S. The 8gb flavor is $120, $140 with a nice metal case. It's pretty fast, I compare
746.
▲
by
sliken
4y ago
After a week of storms the reservoirs had moved from 33% full to 34% full. It's not that they aren't large enough, it's just the watersheds that lead into them aren't particularly large relative to the state. I lived n
747.
▲
by
sliken
4y ago
https://www.friendlyelec.com/index.php?route=product/product... Pretty happy with it, was easy to get ubuntu 22.04 going, but with their image. From what I can tell they use the same not upstreamed kernel on all their
748.
▲
by
sliken
4y ago
> Ok, then surely you won't take issue with me describing my x86 desktop as having 1.2TB/s of memory bandwidth? Unless it's a rather odd config that's counting the GPU memory bandwidth which is decidedly not NUMA sinc
749.
▲
by
sliken
4y ago
> "the M1 Max isn’t able to fully saturate the SoC bandwidth from just the CPU side Sure, wouldn't want to starve the GPU when the CPU is busy. Not sure if the AMX has it's own connection or shares the CPU complex bandwid
750.
▲
by
sliken
4y ago
A comment here on benchmarking the M1 managed to get 360GB/sec observed out of a 400GB/sec memory system, so I'd consider that pretty good and certainly a better percentage of peak than I see out of the x86-64 systems I bench
More ›