Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ethelward
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
1.
▲
by
ethelward
5y ago
Indeed, you're right.
2.
▲
by
ethelward
5y ago
Fromthe mouth of CockraochDB's CTO: ‶So if we were starting at this point in time, I would take a hard look at Rust, and I imagine that we would pick it instead of C++.″
3.
▲
by
ethelward
5y ago
I'd say that life at startups paying peanuts and expecting rockstars is considered slavery, not startup life in general.
4.
▲
by
ethelward
5y ago
> However, modern HN just feels so deeply cynical, angry, and negative with much less of the entrepreneurial tech optimism of the older posts Isn't that a direct reflection of the current state of the tech world? Since the 00's
5.
▲
by
ethelward
5y ago
> When they say like a family they mean treating each other with kindness, friendship and compassion We have a word for that: politeness.
6.
▲
by
ethelward
5y ago
At the same time, you can't expect Wiki editors to be expert in all cutting-edge domain of knowledge to be able to determine which academics are or aren't notable.
7.
▲
by
ethelward
5y ago
> civilization ought to have collapsed several times over. Which it did, in some ways.
8.
▲
by
ethelward
5y ago
Yeah, it has left the 4chan community and entered the mainstream quite some time ago; e.g. here ( https://old.reddit.com/r/virginvschad/ ), and it even has derivatives now ( https://knowyourmeme.com/m
9.
▲
by
ethelward
5y ago
For an overwhelming majority of Zoom users, scaling over 1k users is not a concern.
10.
▲
by
ethelward
5y ago
> Zoom was the very first to let people join meetings without a) signing up; and b) downloading anything E.g. BigBlueButton and Jitsi were doing it for much longer.
11.
▲
by
ethelward
6y ago
That's ~4MB per thousand emails; while I agree it's conceptually wasteful, I doubt it would have any real consequences for anyone.
12.
▲
by
ethelward
6y ago
I agree that these names are kind of meh, but they stand for things that will mostly be chosen on their technical merit, and that you don't have to sell to your friends/mom/gf/brother/etc; so it does not really matt
13.
▲
by
ethelward
6y ago
> Any other reasons to go beyond 64 (or 80) bits? Algorithms stability. As precision decreases when magnitude rises, many tricks are used in sci. comp. to scale everything into ranges where computations won't diverge due to a lack o
14.
▲
by
ethelward
6y ago
Because the academic ML community does not care about shipping product to end users not equipped in nVidia.
15.
▲
by
ethelward
6y ago
Although AVX instructions, those are scalar. The SS pattern at the end is for Scalar Single-precision, so only one lane of the whole SIMD register is used.
16.
▲
by
ethelward
6y ago
Thanks!
17.
▲
by
ethelward
6y ago
> [1] Except empirical Bayes, which is a dark art. As a not-an-expert-in-stats, why would you say that? Is not empirical Bayes basically the same thing, but with priors stemming from the known data?
18.
▲
by
ethelward
6y ago
> there are some minimum legal hours of seminars you have to attend True, but that's like a few dozens hours over 3+ years. > and classes you have to teach during your phd. Nope. Might depend on the “École doctorale” or the lab y
19.
▲
by
ethelward
6y ago
> The RBF kernel nakes any dataset linearly separable, as long as the bandwidth is small enough That's very interesting, would you happen to have the paper proving that somewhere?
20.
▲
by
ethelward
6y ago
BeamVM is much worse than the JVM in the raw computing performances department.
21.
▲
by
ethelward
6y ago
> web-based And that's already not a decent concurrent for org-mode. Not that I have a grudge against web tech, but no self-hosting of my most personal data and no offline abilities is a deal-breaker for me.
22.
▲
by
ethelward
6y ago
> 80-bit precision is actively harmful How comes? Unexpected clipping when converting back and forth to 64bits?
23.
▲
by
ethelward
6y ago
If I understand correctly: > we still make a small profit on low oil prices. and > when prices are low everybody expects to rise back to what we're used to again soon, because we're naturally optimistic. But that's not
24.
▲
by
ethelward
6y ago
I don't understand what it has to do with reentrancy?
25.
▲
by
ethelward
6y ago
Why would you want to port Phoenix? As long as the underlying VM follow the spec, Phoenix should be oblivious of on which it is running.
26.
▲
by
ethelward
6y ago
How is brew different from apt or dnf?
27.
▲
by
ethelward
6y ago
Until you want to write code that works on platforms not supported by these compilers.
28.
▲
by
ethelward
6y ago
> For a bad transition, look at Perl 6 At first glance, you're right. But if you take a closer look, from a purely pragmatic perspective, there are now at least two coexisting ecosystems in two different languages, peacefully sailin
29.
▲
by
ethelward
6y ago
Except for _e.g._ memory alignment, or types of a given width (until C99).
30.
▲
by
ethelward
6y ago
There is a middle ground between having enough cash reserves to survive for months, and not having cash at all because everything has been bled out in dividends and buybacks; namely, having enough cash reserves to survive for some time, but
More ›