Hacker News new | past | comments | ask | show | jobs | submit | norenh's comments login

> 24-bit audio is obviously a noticeable improvement even on $150 headphones

Not going to question your subjective experience, but I do not think most people will hear any difference between 16 and 24 bit playbacks under normal conditions, even with fancy headphones.

"120dB is greater than the difference between a mosquito somewhere in the same room and a jackhammer a foot away.... or the difference between a deserted 'soundproof' room and a sound loud enough to cause hearing damage in seconds.

16 bits is enough to store all we can hear, and will be enough forever." [1]

[1] https://people.xiph.org/~xiphmont/demo/neil-young.html


Your source disagrees with your claim there.

" It's true that 16 bit linear PCM audio does not quite cover the entire theoretical dynamic range of the human ear in ideal conditions. Also, there are (and always will be) reasons to use more than 16 bits in recording and production.

None of that is relevant to playback; here 24 bit audio is as useless as 192kHz sampling. The good news is that at least 24 bit depth doesn't harm fidelity. It just doesn't help, and also wastes space. "



> Not going to question your subjective experience

Read enough glowing product reviews from "audiophiles" on nonsense like $100 electrical sockets and $75 Ethernet cables and you'll start to question everything.


> NAT was a necessity when it was being introduced, so I don't understand the complaints. Sure it's annoying, but what was the alternative?

IPv6


Nope. IPv6 still isn't supported everywhere.

When you're running out of addresses, you need a workable solution at the moment.

Not something that requires all networking equipment worldwide to be replaced with a new standard all at once, with is a practical impossibility.


It doesn't have to be replaced all at once. Most of it already has been replaced.


But it has to be entirely replaced in order to turn off the last NAT, essentially.

Being half-replaced or even mostly-replaced gets you nowhere in terms of NAT removal.


IPv6 would have been great if it was just bigger addresses but the actual system we got suffers heavily from the second-system effect.


Thanks! This info is always nice to have when setting up firewalls. Every application should come with a list of connections made, what the purpose is and what the consequences are if you disallow it.


I guess windows does ask if you want to let a program connect to the internet. Of course that pop-up occurs so often that everyone just hits “go for it,” right?


> Does IF Metall already represent all of Tesla's employees?

No, but Sweden has mainly unions based on industries (like IF Metall represents workers in industrial and metal-sector, you have Transport that represents workers in transport industry, Unionen of white-collar trade workers, etc). There might even be no workers in the specific union at the specific workplace and they might still go on strike since they are the union you are supposed to negotiate with. There are requirements of negotiation but no requirements of results from that, which is the cause of the current situation at Tesla. Tesla "negotiated" but flatly refused any outcome involving the union so they fulfilled the lawful requirements. Now the Union uses the tools they have in place when the negotiations ends badly so everything is going as expected.

>Does that mean there are no active Tesla services in Sweden during the strike?

Like I wrote above, Tesla might have all services running if they have workers doing it, but chances are at least some are members of IF Metall. There is also the effect of sympthy strikes from various other unions: https://www.ifmetall.se/aktuellt/tesla/darfor-tvingas-if-met... This includes transport (loading/unloading in harbors, transport of goods, etc), Electricians (yep, you can guess it, practically everything involving electricity), Builders (maintenance and construction of buildings), service and communication (This is postal workers refusing delivery of post/goods to/from Tesla), etc. The effect goes beyond Teslas services and will extend to all kinds of infrastructure and other parts of society unless Tesla manage to get their own self-sufficient (and probably totally isolated) mini-society without any needs from other parties.

So the more Tesla refuses to find a deal, the more the unions can escalate and it will get quite tricky to find laborers who want to deal with you. Other companies might refuse to deal with you with risk having the strike extend to you to if this goes far enough.

> If union membership is really the way of life in Sweden, is there anything preventing the government from a legislative solution?

It is really complicated, but the main thing preventing government going in with a legislative solution is that no-one (neither the industry or the laborers) wants it. It has been a solution that the union(s) and industry can solve this by themself and as long as they can, the government will stay out of it. This is kind of unique relaxed solution that puts a lot of responsibility on both the industry and the unions to sort things out and has worked well so far. Playing hard-ball like Elon will probably not end well for Tesla in Sweden (possibly Europe) if this continues since the system is kinda built on being mature and able to negotiate. My way or the highway attitudes are not well regarded.

>Just mandate collective agreements?

But what collective agreement? The system in Sweden is built on negotiation between the involved parties and finding some kind of common and sensible ground. There is really no easy top-down fix for that if one party refuses to play ball. Either the other party need to fight back or the whole system collapses and we risk ending up with government moving in, which neither party wants.


And what will GPT info tools learn from, once the public curated sources are gone?


Probably the great swaths of documentation out there that for most use cases people need not waste time sifting through if a computer can do it faster...


Isn't it fun, that ChatGPT's success poisoned the well for everyone else? :)


By then, AGI will be ready, right?


One reason is that tuta does not require you to have any other connection to create and account. Protonmail require a second mail, phone or possibly some kind of payment if I recall correctly (for verification?) that could be linked from your account in theory.

Without having a good anonymous starting point, protonmail does not let you get that starting point, at least the last time I tired (maybe a year ago).


ProtonMail never used to require another email to verify, and only asked for a phone if I was on an IP that had made more than one account already.


Or, AFAIK, if you're registering using TOR.


Yes, Proton is hostile to Tor even though they deceptively market themselves as anonymity friendly: https://news.ycombinator.com/item?id=37174259

It's a stark contrast to Tuta, which allows anonymous account creation with Tor Browser if you pay with cryptocurrency (Monero or Bitcoin, via their partner ProxyStore) and doesn't require a whitelisted verification email address or any other data.


An additional email address is required only in cases when our system detects something suspicious about your network, so if you are coming across this, we recommend changing nodes. If you keep coming across the same issue, please contact us at: https://proton.me/support/contact, so we can take a closer look.

The email addresses, however, are not tied to your account - we only save a cryptographic hash of your email address. Due to the hash functions being one-way, we cannot derive your data back from the hash: https://proton.me/support/human-verification.


Who cares if you hash it, cracking a hash of an e-mail is easy AF compared to passwords. Especially on agency scale... How do you hash it? Argon2 or rather some extremely fast to crack hash?


It's a difficult issue. If they allow unlimited signups via Tor, people will bulk sign up for accounts and use them for spamming, scamming, threats, phishing and other crap. I can imagine why they don't tbh.


Proton forbidding anons from opening free accounts might be necessary for anti-spam/deliverability. But even paid accounts?

"They accept cryptocurrency, but only for existing accounts - after you've already doxxed yourself" (during the initial signup flow, where this payment option has been removed)

This looks very bad to me.


Good point. I didn't think about that.


You don't doxx yourself by creating a Free account. In most cases, no human verification method is required or it's captcha only. As explained above, an additional email address would be required only in cases when our system detects something suspicious about your network. It takes a while for the Bitcoin transaction to come through, which is why we the process is the way it is. The same process applies to users who wish to pay with cash or bank transfer.


Uh huh.

And what suspicious thing about the network would you be detecting for Tor Browser users arriving on the .onion? Their network is uniform as far as you can tell, and you are blocking them from opening either a free account without an invasive verification method (non-disposable email or phone) if it works at all, or a paid account without an invasive payment method.

For Tor users arriving on proton.me, what sense is there in saying "There's a surprise in every 100th exit node! If you cycle through enough of them maybe you too will be allowed to open an account anonymously!" Not treating them as equivalent to .onion visitors is a you problem.

> It takes a while for the Bitcoin transaction to come through, which is why we the process is the way it is.

By not allowing this payment option at all in the signup flow? Removing what would be the only way for Tor users to sign up to your service anonymously without beating lottery odds. Just use any normal off-the-shelf checkout page that waits for however many transaction confirmations you want! (Let's not even get into the lack of privacy coin support, e.g. Monero. For a privacy focused service, Bitcoin L1 only is substandard in 2023.)

Meanwhile, whenever people are concerned about user data being handed over to the authorities again, you counter by pointing out the supposed Tor support: https://web.archive.org/web/20210906132309/https://protonmai...

I'm not saying you are a honeypot. I'm saying you've cultivated such a careless indifference to data minimization that you've become indistinguishable from one.


Regarding TOR, it's based largely on volumes. Spikes on a TOR IP for example would trigger additional anti-abuse measures.


So fix your backend to exempt Tor visitors from those measures, if it's really all due to hallucinating clusters of abuse from a network where abuse categorically does not appear in clusters of the kind that your backend is attempting to detect.

To add an exemption for proton.me: The list of Tor exit IPs is public. For the .onion: That's loopback traffic from the tor daemon running on your own load balancer or wherever you've put it.


Thank you for your feedback, we've passed it on to our anti-abuse team.


I have an email that I've created and only accessed through their Tor hidden service hostname

   https://protonmailrmez3lotccipshtkleegetolb73fuirgj7r4o4vfu7ozyd.onion/
and I was never asked to provide any personally identifying information.


As always when people post this, I just tried it with Tor Browser, and, as always: "No verification method available"

Did you access the .onion with something more fingerprintable like Brave?


No, just the plain Tor Browser.


100% this. As a consumer that tries to do active choices the current model breaks any kind of active choices to support or not support an artist, except nudging it somewhere in the margins. If your pot is only going to work you interact with, you can suddenly make an impact for smaller artists.

The counterargument I see is that subscribers who listen to lots of different artists will have a smaller share of the pot for each artist than someone who listen to just one artist. But I see the inequality here is not as bad as the other option where almost all my subscriber fee now goes to a few mainstream artists I probably never listened to. I also bet a lot of smaller artists would benefit more with a per-user pot, instead of always getting a minuscule share from the global pot.


I see only issues abandoning copyright (for example it is the protection that makes Open Source Licenses work), but we should definitely discuss the length of it. The current trend of longer and longer periods of copyright protection can be questioned and I wonder if society would not benefit with a severely cut copyright period. Berne convention and other international treaties makes it hard to change below Life + 50 years, but I see no reason all the right incentives for individuals and companies would be there if you cut it to just 20-30 years after publication. The benefits of new derivative works, access and sharing knowledge and other areas would be great for society as a whole.

The moral right of claiming ownership to a work should still be protected forever.


While technically true, that DDR5 comes with "on-die ECC", it is only because the memory is so unreliable it will not work properly without it. However, even then, it is not the same as true ECC that have a extra data correction chip on the memory module and also protects against send errors to the CPU.


Depends on your needs, but it is fairly simple to buy a domain-name and manage the dns-servers and hosting of web-server (and I guess e-mail) yourself. If you have a A/AAAA-record in your top TLD pointing directly to your nameservers you do not depend on anyone but the root servers and TLD-servers to be available (in regards to DNS). Now, you just need to find a TLD that is not managed by one of the big ones, but that should be fairly easy.


I believe there is quite a bit of evidence that the major domain companies like godaddy will delete your domain if they receive complaints.


Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: