whyever

18 ms·

91.
by whyever 7y ago
What are its shortcomings compared to Google's offering?
92.
by whyever 7y ago
I disagree. With the use of SIMD, CSPRNG are about as fast. Xorshift and friends are usually not optimized for SIMD.
93.
by whyever 7y ago
The article explicitly compares to the matplotlib colormaps and states the advantages of their contribution.
94.
by whyever 7y ago
Yes, I would recommend to use `rand_chacha::ChaCha20Rng`, which is equivalent to `rand::rngs::StdRng`, but is guaranteed to be value stable.
95.
by whyever 7y ago
Yes, that would be better! However, they seem to be using `XorShiftRng` in their tests, which is no cryptographically secure, so they can't use `CryptoRng` over `Rng` in their API without resorting to compile-time magic. I think they s
96.
by whyever 7y ago
Regarding performance, safety and privacy, locally hosted ads would be a huge improvement over the current situation.
97.
by whyever 7y ago
Depending on what you mean, it does: https://github.com/phaazon/warmy
98.
by whyever 7y ago
This is done all the time by credit score companies.
99.
by whyever 7y ago
> libc++ random is actually decent. I disagree. It does not include a contemporary RNG, and seeding the only decent one correctly is ridiculously difficult.
100.
by whyever 7y ago
I don't think this is true for modern languages. Rust seeds the whole state by default: https://docs.rs/rand/0.6.5/rand/ Julia seems to seed 128 bits: https://github.com/JuliaLang/ju
101.
by whyever 7y ago
> (rand() % (b - a)) + a; This is no longer uniform, because it introduces a bias towards small numbers.
102.
by whyever 7y ago
I don't think that C++'s memory safety issues can be solved by giving it more time.
103.
by whyever 8y ago
This thesis is from psychology, not genetics: https://en.wikipedia.org/wiki/Negativity_bias
104.
by whyever 8y ago
> It is, unless one of you has a Nobel Prize, unsolvable without one person making a change of career. This is not really true, I know a few people who managed.
105.
by whyever 8y ago
> For senior developers, they just give you more work and slow down progress. From my point of view, the static type system does work for me. Having to write tests for what a static type system can prove is tedious, and it only provides
106.
by whyever 8y ago
Firefox supports addons on mobile.
107.
by whyever 8y ago
> Use Signal if you absolutely need end to end encryption, extremely secure chat, and no way to use it outside of your phone being on and connected to the Internet. You can use Signal on Desktop without your phone being online. (This doe
108.
by whyever 8y ago
Here is a talk that looks a bit closer at the different social credit systems in China: https://media.ccc.de/v/35c3-9904-the_social_credit_system
109.
by whyever 8y ago
https://hackerfall.com/story/a-crypto-challenge-for-the-tele...
110.
by whyever 8y ago
> that your programmer always calls free at the exact moment that the memory is no longer needed I think this is not so unrealistic if you use RAII.
111.
by whyever 8y ago
It's not about personal preference. The GDPR does not talk about a "reasonable default", but about "freely given, specific, informed and unambiguous" consent from the user. There must be a "clear affirmative a
112.
by whyever 8y ago
> but most OSS projects won't get the kind of OSS contributions they want for the kind of money they need, simply because the kind of contributors they want can get 10x better pay on the job market, so why they should bother with OS
113.
by whyever 8y ago
> Accessing components by named dimension instead of numeric indices (assuming clean syntax) would reduce mental overload for both readers and writers of code. Couldn't this be solved (at least for the reader) by using descriptive v
114.
by whyever 8y ago
> Rust is only suitable for static linking This is incorrect, Rust supports dynamic linking. https://doc.rust-lang.org/reference/linkage.html
115.
by whyever 8y ago
There was also SML, which is not so different from OCaml.
116.
by whyever 8y ago
Their FAQ [1, 2] suggests that using an encrypted password still warrants an entry. [1] http://plaintextoffenders.com/faq/devs [2] http://plaintextoffenders.com/faq/non-devs
117.
by whyever 8y ago
Is this more than a reskinning of Have I Been Pwned?
118.
by whyever 8y ago
In Germany they used @googlemail.com addresses.
119.
by whyever 8y ago
> That's why Facebook chose to just fix the slow parts of hg[1], rather than fix the slow parts of git, to build a VCS which would scale with its codebase. On the other hand, Google seems to successfully use Git in the same way.
120.
by whyever 8y ago
I disagree. Shadowing can be very useful when transforming types that represent the same thing. For example, rebinding a variable to be immutable.

More ›