Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
lima
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
22 ms
·
241.
▲
by
lima
3y ago
A distributed consensus mechanism provides byzantine fault tolerance - which is helpful even with trusted actors, as this event demonstrates.
242.
▲
by
lima
3y ago
> Your bank and soon every other service will see how 'cheap' it is to fight spam and abuse simply by restricting accounts attested by [apple,google,microsoft] and nobody else, specially 'None' like this one. If tha
243.
▲
by
lima
3y ago
Beautiful example of the limits of AI - it simply reiterates the comment in a seemingly informative way, but it doesn't actually understand it, so it really just states the same thing but more verbosely. Knowing nothing about Passkey
244.
▲
by
lima
3y ago
> and BTW you cannot get the photos out at the same quality as you put them in You can get the original photos out, but they're either making it annoying on purpose, or just don't care: - When you download a photo through the
245.
▲
by
lima
3y ago
WebAuthn is a standard. From what I understood, can use any third-party authenticator like a YubiKey instead of Google's implementation.
246.
▲
by
lima
3y ago
Thanks for debunking this unfounded conspiracy theory. There are many things to worry about when it comes to firmware backdoors, the TPM ain't one of them.
247.
▲
by
lima
3y ago
Same thing, different implementation language :) Starlark is used in a number of production systems so there's good support.
248.
▲
by
lima
3y ago
Well-done parallax scrolling libraries do just fine with scroll wheels and no smoothing. This one just isn't very well done - I bet accessibility features don't work either.
249.
▲
by
lima
3y ago
You might also want to take look at Starlark, which has a Go implementation: https://github.com/google/starlark-go
250.
▲
by
lima
4y ago
There is no widespread use of end-to-end encryption beyond 1:1 chats and small groups. Discord, Telegram, and similar platforms do not use end to end encryption. So I would read this as "we need to get even better at monitoring literal
251.
▲
by
lima
4y ago
Fun fact: It appears that Yandex uploads thousands of placeholders for every internal package name in order to prevent this from happening: https://pypi.org/search/?q=yandex+AND+%22dependency+confusio... Surely that
252.
▲
by
lima
4y ago
Whoa. They didn't even block the installs from continuing to happen. Google must be really confident in their ability to contain threats like this one. In other orgs, this would be a "hair on fire" sev 1 security incident.
253.
▲
by
lima
4y ago
Who is "they"? The head of the agency? Their advisors? Individual employees? Politicians? Failure of this magnitude is almost always a systemic failure. Actually figuring out who, if anyone, committed a crime (through gross neglig
254.
▲
by
lima
4y ago
The head of BaFin and their deputy were forced out over it and the agency is undergoing significant restructuring[1], and a new agency was formed specifically for investigating financial crimes[2]. In terms of actual liability - no, at leas
255.
▲
by
lima
4y ago
Yep. Amateur radio operators couldn't care less about a pirate FM station, but will fiercely defend their own bands. Our local ham community once got a foreign country to move their over-horizon radar off the ham bands by escalating
256.
▲
by
lima
4y ago
Right. Something like lit-html simply re-renders the DOM. This is essentially as efficient as it gets.
257.
▲
by
lima
4y ago
VDOMs were created precisely because the native DOM was too slow. This is no longer the case and a tiny web component framework like Lit significantly outperforms[1] React relying entirely on the browser DOM and template literals for re-r
258.
▲
by
lima
4y ago
> Just think how many JS kBs and CPU cycles would be saved globally if browsers could do data binding and mutate the dom (eg: morhpdom, vdom, etc) natively. And the emissions that come with it. Browsers are solving these real-world p
259.
▲
by
lima
4y ago
They will revolutionize the ability to find and fix vulnerabilities. Threat actors can already afford expensive security researchers today. It removes some of the asymmetry.
260.
▲
by
lima
4y ago
Should've said "is spreading vaccine misinformation".
261.
▲
by
lima
4y ago
Haven't tried this one but Reviewable and Graphite. They're all very nice and yours looks nice also. The problem with all these third party SaaS-es on top of GitHub: - You have to trust a random (no offense meant!) SaaS company wi
262.
▲
by
lima
4y ago
Probably not a coincidence - changes are the #1 reason of outages. Moving faster does mean breaking more things.
263.
▲
by
lima
4y ago
My company uses a self-hosted Gerrit instance. A bit of a learning curve, but the code review experience is SO much better than PRs (one commit == one unit of review, clean intra-diffs between different revisions of a patch, stacking of rev
264.
▲
by
lima
4y ago
Michael Yeadon - former chief scientist at Pfizer - is now an anti-vaccine activist[1]. Similar story with Robert W. Malone[2]. Former accomplishments or positions are not a good appeal to authority. [1]: https://en.wikipedia.org
265.
▲
by
lima
4y ago
I think it's fair to call that a conspiracy theory.
266.
▲
by
lima
4y ago
> Stupidly simple and easy to host. At scale, serving tens of thousands of users? Unfortunately not.
267.
▲
by
lima
4y ago
It's allowed in Europe and Denmark as well. Some µA of leakage current are acceptable. Happens because of Y capacitor leakage in switching power supplies.
268.
▲
by
lima
4y ago
Big benefit of Advanced Protection: you can go tell less technical users to set it up and it will enforce all these best practices (no SMS, two keys, no giving random apps access to GMail...).
269.
▲
Rosenpass – formally verified post-quantum WireGuard
(github.com)
285 points
by
lima
4y ago
|
80 comments
270.
▲
by
lima
4y ago
Still much less common to have that vs. a Paypal account :(
More ›