Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
nimish
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
18 ms
·
271.
▲
by
nimish
3y ago
Fermat's principle (least time) predates Maupertuis' but it's not obvious it's basically the same thing. Interestingly Maupertuis was motivated by placing time and distance/space on the same footing, predating Einst
272.
▲
by
nimish
3y ago
45k sounds low, but I suspect you aren't US based. Id like to move some of my hobby stuff to FTE but it's more favorable to work in software :)
273.
▲
by
nimish
3y ago
I am curious as to why. They should be represented like joins, right?
274.
▲
by
nimish
3y ago
Where is that from? $4k for that kind of HW is a steal.
275.
▲
by
nimish
3y ago
Glad to see avisynth style editors still developed.
276.
▲
by
nimish
3y ago
Frozen marimo?
277.
▲
by
nimish
3y ago
That's the academic price. The commercial/regular price is likely 5-10x more.
278.
▲
by
nimish
3y ago
Correlated subqueries are absolutely fine and are equivalent to (possibly lateral) joins. They aren't as easy to identify and optimize as such, so I wouldn't necessarily recommend them, but sometimes they make sense. Not an issue
279.
▲
by
nimish
3y ago
I mean embedded is very much physics based in contrast to the web stuff that pays the bills. I can only do very basic embedded but it's always fun to see something tactile and tangible. It also needs a kick in the pants with software.
280.
▲
by
nimish
3y ago
All of modern technology owes its origin to measuring light precisely. Optics is an ancient science. It's quite low tech, surprisingly so.
281.
▲
by
nimish
3y ago
It really isn't, it's basically just laziness and tradition. You can get a significantly better approximation that is coherent and causal by adding a one or two extra linear equations to basic newtonian gravity (not MOND, GEM) and
282.
▲
by
nimish
3y ago
No, it's a nice pun and nothing more. People read too much into wordplay, perhaps we need retinal cells that stabilize that
283.
▲
by
nimish
3y ago
Where will I get my crank science ads on the H-bond angle of water?
284.
▲
by
nimish
3y ago
Not that I'm aware of: https://developer.nvidia.com/gpudirect-storage is only supported on Linux. Microsoft DirectStorage isn't GPU Direct Storage, annoyingly enough. No idea what AMD's equivalent is though t
285.
▲
by
nimish
3y ago
Hm, without a consumer equivalent to GPU Direct Storage (PCIe P2P DMA), this is not as cool as it could be. Still has to bounce to and from the CPU for no good reason.
286.
▲
by
nimish
3y ago
Yep. They are 100% correct that vaccine drives can be hijacked by geopolitical enemies. It is not an unreasonable fear.
287.
▲
by
nimish
3y ago
No one knows what it means, but it's provocative. It's mainly about who is allowed to control what other people can do, i.e. power.
288.
▲
by
nimish
3y ago
If the openai board can't get humans to align, what hope do they have of "aligning" ml models?
289.
▲
by
nimish
3y ago
The complexity is already paid for since you aren't going to be able, or even want to, wholly rip and replace IPv4 since it gains very little and risks a lot. IPv6 would have been great if it didn't have to work in a brownfield en
290.
▲
by
nimish
3y ago
Thanks! The correct authors are Graham, Knuth, and Patashnik.
291.
▲
by
nimish
3y ago
It's easier to treat these as linear difference equations with constant coefficients and use the ansatz of λ^{n}. This is morally the same thing but without the need for matrix manipulation. Knuth, Oren and Patashnik's Concrete Ma
292.
▲
by
nimish
3y ago
They did and the companies whined that they then had to live up to their demands to arbitrate: https://www.perkinscoie.com/en/news-insights/another-court-r... The courts have not looked favorably on the US Chamber
293.
▲
by
nimish
3y ago
When I bought one a year ago in the US I had no issues. I guess their insistence on it being a neural network inference device has bit them in the ass.
294.
▲
by
nimish
3y ago
The elusive linear algebra book that isn't simply matrix algebra. A gem.
295.
▲
by
nimish
3y ago
Cerebras' strategy is the winner then if hardware was all that mattered
296.
▲
by
nimish
3y ago
That's degradation rate in capacity, not overall capacity. But trading off 40% of the energy for a 25% lifetime increase is not a great trade for everyone. You would then consume almost 2x of those cycles just recharging to make up for
297.
▲
by
nimish
3y ago
Which is how the infamous self-pipe trick works: https://cr.yp.to/docs/selfpipe.html Nowadays signalfd and pidfd are the better options.
298.
▲
by
nimish
3y ago
This is maybe half true. Dark matter is required to explain these under the current lambda-cdm model, but the behavior of galaxies is not evidence for it in a constructive sense. It's only evidence that the current model is untenable.
299.
▲
by
nimish
3y ago
It is? And not any esoteric analog design? That explains the secrecy.
300.
▲
by
nimish
3y ago
Lest anyone think this is limited to "soft sciences" we should recall that this "settled science bias" happens in hard ones, too: https://hsm.stackexchange.com/questions/264/timeline-of-meas...
More ›