Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
eklitzke
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
31.
▲
by
eklitzke
2y ago
The devices they're riding are called electric unicycles, often abbreviated to EUC.
32.
▲
by
eklitzke
3y ago
It might be worth looking into, but: 1) Probably by design protonmail doesn't keep these kinds of logs around for very long 2) Hacking groups pretty much always proxy their connection through multiple layers of machines they've ro
33.
▲
Goblins and Kobolds and Trolls, Oh My
(frank-mitchell.com)
1 points
by
eklitzke
3y ago
|
0 comments
34.
▲
by
eklitzke
3y ago
Three cities now, they recently added Los Angeles.
35.
▲
by
eklitzke
3y ago
This is a joke right? If you have password authentication disabled then it shouldn't matter how many thousands of times a day people are scanning and probing sshd. Port knockers, fail2ban, and things of that nature are just security by
36.
▲
by
eklitzke
3y ago
Obviously compiling for the target architecture is best, but for most software (things like crypto libraries excluded) 95% of the benefit of AVX2 is going to come from things like vectorized memcpy/memcmp. Building glibc using ifuncs t
37.
▲
Google's Waymo Is Finally Moving Full Throttle
(seekingalpha.com)
2 points
by
eklitzke
3y ago
|
0 comments
38.
▲
The next steps for our robotaxi in Las Vegas, Foster City, and beyond
(zoox.com)
2 points
by
eklitzke
3y ago
|
0 comments
39.
▲
by
eklitzke
3y ago
Lots of people know.
40.
▲
Why does the `reset` command include a delay? (2017)
(unix.stackexchange.com)
81 points
by
eklitzke
3y ago
|
99 comments
41.
▲
Birmingham Quran Manuscript
(en.wikipedia.org)
3 points
by
eklitzke
3y ago
|
0 comments
42.
▲
San Francisco police make arrest in Waymo Chinatown arson case
(sfstandard.com)
1 points
by
eklitzke
3y ago
|
1 comments
43.
▲
by
eklitzke
3y ago
I haven't looked at the exact schema for this dataset but for this type of query pattern to be efficient the data would need to be partitioned by date.^[1] I'm guessing that it's not partitioned this way and therefore each of
44.
▲
by
eklitzke
3y ago
If you need to interact with existing C++ code or libraries obviously C++ is going to be a much easier path than using Rust. If you don't have such a requirement, or you think that the available Rust libraries are good enough for what
45.
▲
by
eklitzke
3y ago
You think someone earning $160k before taxes in California takes home $135k?
46.
▲
What are the ethics around asking an employee not to talk about her raise?
(askamanager.org)
1 points
by
eklitzke
3y ago
|
0 comments
47.
▲
Tortured Poets' or Poets? Taylor Swift Meets the Apostrophe Police
(nytimes.com)
2 points
by
eklitzke
3y ago
|
1 comments
48.
▲
Engraved Portraits of Gauss by Acme Klein Bottle
(kleinbottle.com)
2 points
by
eklitzke
3y ago
|
0 comments
49.
▲
Joe Rogan's Spotify Deal Renewal Worth Up to $250M
(variety.com)
2 points
by
eklitzke
3y ago
|
0 comments
50.
▲
San Francisco's DC Grid (2012)
(spectrum.ieee.org)
1 points
by
eklitzke
3y ago
|
0 comments
51.
▲
California drivers are on their own with this new parking law
(sfgate.com)
1 points
by
eklitzke
3y ago
|
0 comments
52.
▲
Cruise Releases Independent Reports on Oct 2 Pedestrian Dragging Event
(forbes.com)
2 points
by
eklitzke
3y ago
|
1 comments
53.
▲
San Francisco sues California over 'unsafe,' 'disruptive' self-driving cars
(washingtonpost.com)
23 points
by
eklitzke
3y ago
|
15 comments
54.
▲
by
eklitzke
3y ago
Apple lets their CPUs and GPUs run at a higher max frequency/power draw, but you're absolutely right that in clock-to-clock performance they're only marginally better in both CPU and GPU performance.
55.
▲
'Caucus': A Curious American Word
(merriam-webster.com)
2 points
by
eklitzke
3y ago
|
0 comments
56.
▲
by
eklitzke
3y ago
What is the point you're trying to make? Waymo isn't making claims about their driving performance in snow. And they're not comparing their performance to using a basket of human performance in all weather conditions, if you
57.
▲
by
eklitzke
3y ago
I'm not an expert in this, but I don't think you'd be using similar hardware for video encoding and AI. Encoding uses a lot of compute, but it doesn't use much memory, whereas deep learning models (especially LLMs) use g
58.
▲
Roman von Ungern-Sternberg
(en.wikipedia.org)
2 points
by
eklitzke
3y ago
|
0 comments
59.
▲
by
eklitzke
3y ago
What makes you think they have an operating margin of 90%? The blog post says in the last 15 months alone they've hired 500 additional employees ("figmates"), there's no way their margin is anywhere close to what you cit
60.
▲
Code of Ethics
(sqlite.org)
26 points
by
eklitzke
3y ago
|
21 comments
More ›