Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
w-m
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
23 ms
·
241.
▲
by
w-m
4y ago
Well GitHub Copilot tells me 2040, if I pose everything up to "Answer: " as a Python comment in an empty file. And if I remove the "Question" and "Answer" parts, just leaving the question itself, Copilot gives
242.
▲
by
w-m
4y ago
At least we all agree that as one week is 7 days, when we say we'll meet in two weeks, that'll be in 14 days. Right, guys? Oh no... https://en.wiktionary.org/wiki/quinze_jours
243.
▲
by
w-m
4y ago
Gonna have to wait for someone from the community to come along and profile the game to find that 10 Mb json file of in-app purchase items being parsed in O(n*2).
244.
▲
A number theory problem where pi appears surprisingly
(mathoverflow.net)
6 points
by
w-m
4y ago
|
1 comments
245.
▲
by
w-m
4y ago
They aren't using the rendered PDFs. They are convering from the LaTeX sources, that you upload to arxiv with https://github.com/brucemiller/LaTeXML
246.
▲
by
w-m
4y ago
Here's the article rendered as html: https://ar5iv.labs.arxiv.org/html/2208.14755 (Replace the x in arxiv.org with a 5 to get there. I feel like it's working well enough in most cases, when will they start li
247.
▲
by
w-m
4y ago
Personalized Spatial Audio has me really excited ( https://youtu.be/ux6zXguiqxM?t=2739 ). The idea is to map your ears / head with the depth camera of the iPhone, to create a personal profile. 10 years ago I played aroun
248.
▲
by
w-m
4y ago
Commercialization of large machine learning models might push the RAM game for the end user up in the not-too-far future. For unified memory machines like the current Apple chips that is, and otherwise the memory on consumer GPUs for the en
249.
▲
by
w-m
4y ago
There's an inactive Meetup group with 1700 members here: https://www.meetup.com/de-DE/berlin-hacker-news-meetup/ I got an email from Meetup recently that it will shut down unless a new organizer steps up.
250.
▲
by
w-m
4y ago
Not quite! The point of the article is that you don’t need to change the class of obj to override this instance check. If you have a look, ClassA always uses the hook implementation in the examples. And this can be shortened to just using i
251.
▲
by
w-m
4y ago
Looking in the implementation of the Python version of ABC[0], you'll see that `__subclasshook__` doesn't do any language magic, it just registers the method to be called in `__instancecheck__`. So if you just implement `__instanc
252.
▲
by
w-m
4y ago
This Twitter thread highlights that this is a hyperbole/misinformation: https://twitter.com/samuel_marsh/status/1550883405105168386 „I sincerely doubt that the absence of this particular paper and AB*56 from
253.
▲
by
w-m
4y ago
Ah yes, thanks for linking that. So there was a discussion about a conflict with `--verbose` at least; whether to go for the upper case `-V` or completely drop the version flag from the patch. But then: "I do not care too deeply, and i
254.
▲
by
w-m
4y ago
The full release notes ( https://github.com/git/git/blob/v2.37.0/Documentation/RelNot... ) offer this highlight: "git -v" and "git -h" are now understood as "git --version&quo
255.
▲
by
w-m
4y ago
Wouldn’t the simplest way to punish selling crap be to bump up the years of required warranty significantly, for most product categories?
256.
▲
by
w-m
4y ago
I don't think I get how that solves the issue: you would have set a fixed cutoff time, where you switch from one window/batch to the next. It doesn't matter when you arrive within the window. But statistically, even for rando
257.
▲
by
w-m
4y ago
This may be a naive question (I have very little knowledge about building languages and compilers): Would this be possible in Python by introducing a keyword like `final`? Any object, variable, method that is marked final just has to be loo
258.
▲
by
w-m
4y ago
This is a curious reply for me. I would think that there are very few parts in pandas that could be sped-up by reimplementing them with a compiled language. Pandas is plenty fast for the built-in methods, it only gets slow when you start in
259.
▲
by
w-m
4y ago
Trying to sign this, I got a 429 "too many requests" when clicking on confirm, and something about 20 per 1 hour. Did I sign, or didn't I? Reloading the page tells me there's no pending confirmation ID.
260.
▲
by
w-m
5y ago
I came across such a prompt maybe three times total in my life, all of them making GitHub repos public or private, or deleting them. Made me stop completely in my track. So it seems to be very much dependent on what you do day to day.
261.
▲
by
w-m
5y ago
...a C++ library with a CUDA backend. But these high-performance building blocks might only be saturating the GPU fully if the data is large enough. I haven't looked at implementing these things, but I imagine uf you have smaller netwo
262.
▲
by
w-m
5y ago
YouTube app on Apple TV presented me with a dialog to click to update the app because the connection had been lost. Halfway through a video, no less. Of course the button didn’t do anything at all. Not sure where it was supposed to lead.
263.
▲
by
w-m
5y ago
The license should forbid competing artists to using the artist’s work as well. In fact, no human should come in contact with the produced art, otherwise they might be accidentally inspired by it, thus stealing from the original creator.
264.
▲
by
w-m
5y ago
Here was a weird one for me, why is numba jitted initialization twice as fast as np.arange, which presumably has a C implementation? In [1]: import numpy as np In [2]: %timeit np.arange(10000, dtype=np.int64) 3.03 µs ± 94.3 ns pe
265.
▲
by
w-m
5y ago
Could you make one that reduces the opacity of the second half of the word, while keeping the same boldness for the complete text, to compare the effect?
266.
▲
by
w-m
5y ago
Neat! I'm using QWERTY International layouts myself, where you can type umlauts and ß with special keys for modifiers (e.g. alt+u on Mac for ¨), but I still think this is a cool tool. Looking through the repo I wondered why you would c
267.
▲
by
w-m
5y ago
The amount of compute paired with the shared memory and fast bandwidth would make this an awesome fit for machine learning. But for PyTorch, the framework everybody is using now at least in computer vision, there seems to be no support at a
268.
▲
by
w-m
5y ago
The video makes some good points, but overall it’s very repetitive. Just count how many times the shot of the manual with “call a service technician” is used, always with similar narration, or the one of the finances of the company who make
269.
▲
by
w-m
5y ago
I’ve guess they’ve come quite a way when this is a newsworthy event. Remember the days when the servers would just go down daily?
270.
▲
by
w-m
5y ago
I'm happy to give my 5$/month, as I'm using Signal daily. When moving away from WhatsApp it seemed like the best option for me and my circle of friends. I can only hope that some of the money would be used to hire a developer
More ›