geertj

8 ms·

91.
by geertj 2y ago
The consumer side of this would be polling a memory location for new logs, correct? It would not be possible to wake up the consumer in 7ns as that would take a FUTEX_WAKE system call with is O(microseconds). I've been wondering about
92.
by geertj 2y ago
That was my instinctive reaction as well, but as others have pointed out, the team does have credibility from e.g. the Ruff linter.
93.
by geertj 2y ago
Yes, democracy is subject to a 51% attack, like blockchain stuff. Better than a 1% or 10% attack though. Some countries like the US have a constitution that can only be changed by a majority >> 50%, offering additional but still not f
94.
by geertj 2y ago
> You cannot protect a democracy against anti-democratic forces through purely democratic means. Riots and political violence are an expression of speech and arresting the perpetrators takes away their democratic freedoms. Should an idea
95.
by geertj 2y ago
> The question that needs to be answer is how far democracy is willing to go outside of democratic bounds to preserve itself. I would say it should not do that essentially ever? If so, what kinds of undemocractic behavior would be allowe
96.
by geertj 2y ago
> What other institutions (or democratic tools) should have acted to halt the extremist anti-democratic movement lead by Bolsonaro? I am not familiar with Bolsonaro's movement, but censoring people under the guise of protecting demo
97.
by geertj 2y ago
Per the description this uses busy waiting in the workers to get to nanosecond level latencies. I wonder if anyone has a perspective on how realistic busy waiting is in large applications with tens of thousands of tasks? Maybe it works if t
98.
by geertj 2y ago
I see.. You're saying that theoretically, it may be possible to compute the digits of R in a different way, without computing Σ(n). Since it's not possible to unambiguously get Σ(n) from my definition of R, you cannot use the halt
99.
by geertj 2y ago
> I have no idea what isn’t computable. The busy beaver function Σ(n) is not computable, at least, not for every n. If it were, it could be used to determine if an arbitrary Turing machine halts, solving the halting problem. To get from
100.
by geertj 2y ago
These levels are are still used in the US today for eg electricians and other licensed trades. Interesting that they survived for so long.
101.
by geertj 2y ago
You might be getting downvoted because many people know Python is among the slower dynamic languages, and there are other reasons to use it. Speaking for myself, the reasons that make me reach for Python for some projects are the speed of d
102.
by geertj 2y ago
You would still have to getdents() everything but this way you may indeed save on stat() operations, which access information that is stored separately on disk and eliminating these would likely help uncached runs. You could sample files in
103.
by geertj 2y ago
Heat
104.
by geertj 2y ago
Mine is not visible, and I did not opt-in.
105.
by geertj 2y ago
I am familiar with that number but this does not address the point the top-level comment is making. The commenter claims is that there exists a class of problems where lifetime and ownership are not knowable at compile time, and makes the c
106.
by geertj 2y ago
> Thankfully we have actual data on this, not just random numbers, and no, this is not the case in practice. Mind to share that data? I think it would be hard to get quantitative data either way.
107.
by geertj 2y ago
First came across this 2 days ago and found this extremely impressive. There's also a YouTube presentation where Sean goes over the main features of Circle's C++ memory: https://www.youtube.com/watch?v=5Q1awoAwBgQ
108.
by geertj 2y ago
Yes, also referred to as “Assured Access.”
109.
by geertj 2y ago
Because of a "wheel of cheese," usually referring to Parmezan cheese.
110.
by geertj 2y ago
I agree with the article that most data sets comfortably fit into a single traditional DB system. But I don’t think that implies that big data is dead. To me big data is about storing data in a columnar storage format with a weak schema, an
111.
by geertj 2y ago
Very hard and foundational capability need not be correlated though. I think the more likely explanation that orbital refueling hasn't been done yet is not that's it's exceptionally hard, but that there hasn't been a nee
112.
by geertj 2y ago
Yes, the article lists a few reasons, none of them convincing. Specifically: > You have to get the propellant into space. This is going to take a large number of flights (~15) at a pace that has not been done before for a vehicle of that
113.
by geertj 2y ago
Why is cryogenic propellant transfer any more difficult than other difficult things SpaceX have already done (eg landing a rocket, and building a full flow staged combustion engine)? They do this on earth every time they fuel the rocket. I
114.
by geertj 2y ago
Again I am not saying one system is better than the other. The companies you bring up are excellent global companies that are well run. But they have not caused the radical change like the ones that I mentioned.
115.
by geertj 2y ago
Those very generous working conditions, and the mind set that it comes from and creates, I am convinced, is also why Denmark does not have an Apple, Tesla, Google, Nvidia, or SpaceX, and why the AI revolution is heavily based on the US. I’m
116.
by geertj 2y ago
Thanks! A bit late to the party, but if you still see this, I presume the authentication exchange between the web server and the device is some kind of challenge response? And if so, does the challenge/response depend on the type of cr
117.
by geertj 2y ago
I would add to that list: * Has random binary test files in its repo, ideal for hiding exploit payloads.
118.
by geertj 2y ago
Nice phrasing, I lack that mental model as well. Anyone here willing to distill down the whole thing to a few sentences? Who stores what kind of secret, and is there some kind of challenge/response at auth time?
119.
by geertj 3y ago
I wonder if someone knowledgeable could comment on OneAPI vs Cuda. I feel like if Intel is going to be a serious competitor to Nvidia, both software and hardware are going to be equally important.
120.
by geertj 3y ago
Are you sure that with https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2020/p05... this is still undefined behavior?

More ›