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

I'm not surprised at all. LLM responses are just probability. With 100s of millions of people using LLMs daily, 1-in-a-million responses are common, so even if you haven't experienced it personally, you should expect to hear stories about wacky left field responses from LLMs. Guaranteed every LLM has tons of examples of dialogue from sci-fi "rouge AI" in its training set, and they're often told they are AI in their system prompt.

Monkeys and typewriters seems like the least likely explanation for what happened here.

They don't tell you you can put anything with an @ and a . in that field.


Yea email fields without required verification get my maximum character length pseudo email.


Why not your SQL injection attack email?


Is there a reason why this isn't Nokia branded? It seems insane to me that HMD own the most respected phone brand in history and... just choose not to use it for some of their phones? I feel like I'm missing something about the HMD/Nokia situation here.


Nokia hasn’t been the “most respected phone brand” in at least a decade… and I’m sure it costs them money to license it. They’re slowly letting go.


This seems like a kids phone, so the "decision making unit" should be the parents in the end. I am sure they would know about Nokia, that they were happy users years ago and buy the brand again for their kids.


It's not unfair. I have a different standard for proprietary payware and proprietary freeware. With the former, you know where you stand, but in my experience, the reasons for not releasing freeware under a FOSS licence are normally user-hostile, like baiting in users with a free product and charging money once they're hooked. I've been bitten too many times, so I don't tend to install proprietary freeware anymore.


Yes, I think SetWindowDisplayAffinity is the Windows API function.


For me there's two obvious improvements:

- Support for multiple monitors with different scale factors, eg. a high-DPI internal laptop screen and a low-DPI external screen. Dragging windows between each display is seamless with Wayland, which is better than you can even say for Windows.

- Touchpad scrolling in web browsers is now smooth tear-free 60fps+. For some reason, both Firefox and Chromium-based browsers had frame pacing issues with Xorg, especially on low-powered hardware or with full desktop composition. These issues were so well known that testufo.com refuses to give results on Linux browsers. But both browsers work about as well with Wayland as they do in macOS and Windows.


For me it's exactly the same two points and I think for many more people as well.

It's crazy to imagine that "just" these two points are what's making people move to a completely different architecture. I wonder how things would be if xorg implemented multi monitor scale factors and fixed their tearing and performance.


If a game doesn't work in Wayland, you could always launch it in gamescope[1], which AFAIR doesn't expose WAYLAND_DISPLAY by default, so games should treat it the same as an X11 desktop.

[1]: https://github.com/ValveSoftware/gamescope


I haven't had anything flat out refuse to work on Wayland, but unfortunately Discord won't work right. It runs, but it can't detect that you're AFK any more, so you stop getting messages on the mobile app. It's a bummer because otherwise I would love to use Wayland.


> In any case, it changed years later when a startup using Nodemailer was acquired for half a billion dollars. I was financially not in a good place back then, and when I saw the news, I started to wonder – what did I get out of this?

This is really what you should expect when you work to improve the commons in the same world where there are entities that are hyper-optimized to make the most short-term profit out of anything they can exploit. Of course they're not going to give anything back. It could happen to any FOSS dev. It sucks, and it's definitely human to look at all the money they're making and feel like you deserve some of it. You do deserve it! Everyone deserves to make a living. But the world is still a better place with FOSS in it. It's a shame for this to happen to someone and for them to decide that improving the commons was a mistake and instead they should have been making projects that FOSS orgs can't use and individuals and small orgs are priced out of (but is still "peanuts" for big businesses.) If you make best-in-class software that's FOSS, everyone benefits, and you can feel proud that individuals have access to the same resources as big corps because of what you've done.

I'm also tentatively in favour of the idea of scaring away big corps with GPLv3 or AGPL licensed software.


This is just the thing - there needs to be a very clear reason for you to partake in FOSS, something that you want to gain from it that has a bigger value to you than the cost of allowing your time and effort to be used by others for free, and money can not be it.


Exactly, before you go to open source, take a hard look at

* why I am doing that

* plethora of burned-out maintainers and their posts

* how I am going to deal with the issues/PRs, toxic entitlement

* what's my exit strategy

The first thing before you go into open source (provided it's actually used open source) is to answer these questions honestly for yourself. Because it's massive time sink with no money and *there will never be money* (unless you go open core or your employer pays you, in that case that's just a job just like any other).


One important thing the author got from working on open source is free feedback (issues).

I don't view people taking the time to open issues as entitled people, but people offering their free time providing invaluable feedback.

Those issues are quite often different from what I expect, and they represent of how people are using the software.

The only mistake the author did was waiting too much monetizing, not doing open source software in the first place.


TBH, I get way better feedback from paying users than previously from free users. Free users like to tinker and think in terms of "what if," so they bring up all kinds of features the software should also have because it can or it would be cool. The paying users only need actual features that help their business case, and they do not care at all about these "what if" features.


I see, do you think just _starting_ with payed product would have been better? (Or starting with a product + open source tools for marketing?)

BTW good luck to scaling up to $60k / month, it will be fun


I did not plan to make the project paid at first, I would have prefered the OS / Open Core model, but it did not work out. So what I meant about the feedback was that the feedback for a free product might not help much for a paid product and vice versa. Different target groups, different priorities. On the other hand, more users, no matter if free or paid, help to detect edge case bugs better as there is a higher chance of someone stumbling on it and reporting it. In this case the first larger wave of free users did help me, yes.


Plenty of projects are source-available, but not open source, and get tons of issues, and even contributions (https://github.com/MetaMask/metamask-extension off the top of my head)


> I'm also tentatively in favour of the idea of scaring away big corps with GPLv3 or AGPL licensed software.

GPL scares freeloaders.


Only the law-abiding ones.


Those are people who would probably steal software, anyways, so for most businesses I think worrying about those folks would be a waste of time


I don't think having a great liberally-licenced microkernel-based OS is as good for the world in a practical sense compared to forcing every Android and Chromebook vendor to publish the changes they made to a copyleft monolithic kernel. I can totally understand being worried that the rise of Fuchsia would result in a dark age for running alternative open source OSes on cheap consumer smartphones, SBCs and ARM64 laptops.


Vendors don’t publish source code now lol


Some do. If the current state of things is what happens when vendors are required to publish their source by law, imagine how much worse it would be if they weren't - there'd be nothing. What little we have of FOSS OSes on cheap consumer devices would be gone. Imagine no LineageOS, no community-supported postmarketOS devices, no installing mainline Linux on old Chromebooks etc.


This is just learned helplessness. If loss of control is a concern for you, there are FOSS privacy-focused operating systems you can install to your phone. You do control your phone.


There are zero operating systems you can install on modern smartphone hardware that give you full control of your phone.

Even the most open of Android Roms still require giving a bunch of proprietary drivers, firmware, and applications root access for the most basic of hardware functionality.

A major reason I gave up on phones and got a tiny laptop to run qubes on.


There's a difference between avoiding the very-real tracking for targeted advertising that companies like Google do, and the paranoia that your camera blobs are tracking you just because they're proprietary.

> proprietary [...] firmware

Your tiny laptop has this too.

> proprietary [...] applications

There are plenty of AOSP-derived ROMs without this.


Show me even one AOSP based ROM that does not ship with proprietary software that runs on modern hardware.

The vendor partition is a black hole of poorly understood and mostly proprietary garbage that can do whatever it wants.

Also if you do not ship your ROMs without OMA-DM and similar malware, you are not even permitted to connect to some cell networks.

There are a comical number of apps and blobs copied from the vendor partition of stock roms to make most of those "open source" roms function on recent devices.

https://raw.githubusercontent.com/GrapheneOS/vendor_state/14...

At least on a laptop I can run QubesOS to hypervisor isolate the rare hardware that requires blobs, like wifi.


Not Android but I think PinePhone and Librem 5 meet your standards. AFAIK they have proprietary WiFi/BT firmware, but they use virtualization (IOMMU) to protect themselves from it. I don't see how that's fundamentally different to your Qubes laptop.

Also I think you're ignoring my first point. I don't think any of this paranoia is reasonable. Unless you're being targeted by state actors, none of it is necessary to fill a reasonable definition of "controlling your device." This is all in response to an article about a child who doesn't like phones because he doesn't like social media. The adtech machine that powers social media is pervasive and it has its fingers in proprietary smartphone OSes, but it doesn't have its fingers in AOSP ROMs like GrapheneOS, so installing one of those is an easy and achievable way of avoiding it. Whether it's possible to remove every last drop of proprietary software is a side-issue here.


For context, I run a security consulting firm focusing on high risk clients like fintech firms and my job is quite literally defending my clients from sophisticated adversaries which includes state actors. Your threat model is not my threat model.

Degoogling and solving for privacy from corporations makes CalyxOS a great option for many, but privacy and security against actors willing to more blatantly violate laws is not as easily an attainable goal on Android sadly. I spent months building the aosp-build project and compiling and signing my own AOSP roms from scratch but even then there are just an insane number of blobs with way too much power to ever be able to trust.

Binary blobs with root access to my devices is a non starter for my privacy goals, free software preference, and threat model. QubesOS is about the only OS that suits my requirements, and even then it only runs on hardware that is hard to trust. Ideally it can be ported to my ppc64le Talos workstation soon.

All of that is only half the reason though. I was addicted to connectivity and over reliant on tech to live my life. Giving up my phone means not having constant bombardments of information and notifications and distraction making it difficult to be present in and enjoy the real world. When I walk away from my desktop computers, or leave my home, I am offline.

The ability to confidently navigate the real world and be comfortable in my own head with minimal tech took a while to build, but it feels like a superpower now. I have no interest in going back. Giving a phone to a kid feels like child abuse to me now given how much control I feel I have over my own brain again. I want that for others, especially kids.

I would maybe carry a Precusor when it gets a matrix messenger as a single purpose wifi only device I use to coordinate with people at events, but in no rush.


> I was addicted to connectivity and over reliant on tech to live my life. Giving up my phone means not having constant bombardments of information and notifications and distraction making it difficult to be present in and enjoy the real world. When I walk away from my desktop computers, or leave my home, I am offline.

Why didn't you (want to) do this with your phone? Couldn't you turn off notifications, leave it on your desk, maybe leave it at home? Was it too much of a temptation at all?


If I am only using my phone at my desk, I might as well use my actual desktop on the same desk that has a proper screen and keyboard.


"If you have nothing to hide, you have nothing to fear"


That's... not really it. Do you really think there are backdoors hiding in firmware blobs from reputable vendors?


100%. I have plenty of insider knowledge here I sadly cannot share, but search for supply chain attacks and the sheer number of public headlines of known attacks can keep you busy for days. State actors are constantly trying to get footholds in software supply chains, and often succeed.

Consider Intel ME firmware is a literal well documented backdoor we do not allow on US government systems... only civilians.

Most of the time our adversaries do not need to be covert enough to mess with firmware. Consider OMA-DM apps that run on most phones with insane permissions taking orders from cell towers.

https://gist.github.com/thestinger/171b5ffdc54a50ee44497028a...

We cannot even keep public open source repos like NPM free of supply chain attacks. Proprietary blobs make it that much easier to hide things.

Also all you need to backdoor every encrypted messenger is a kernel module that ensures /dev/urandom is a bit less random on the devices of targeted dissidents and journalists. Now look at how many proprietary blobs from piles of random vendors we load into modern phone operating systems, even "open" android roms, and think about SolarWinds for a second.


As another commenter pointed you, you don't have the ability to fully control your phone hardware, but also even allowing yourself to be tracked by the cellular network is a trade off. Your OS will probably not save you from stingrays and dirtboxes


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

Search: