This is the best setup I've tried as well. Syncthing works so well I just often forget about it / take it for granted. I used to just deal with plugging our Kobo devices in, but now I can just distribute the relevant media by dropping a file somewhere.
The Kobo devices are truly worth every penny and we've got 4 of them in our household at this point. These are some of the best devices to put in the hands of kids.
If the assertion here is that Kobo devices are fragile: they're not.
Our first two Kobo were purchased in 2018 and both have been on every business / personal trip since. I don't particularly go out of my way to protect mine, I have the stock magnetic cover. Other than the edges of it wearing, that's the only visible "damage". My kids have had Kobo since 2019 and they take them everywhere. The Kobo devices are not fragile in the least. I worry more about them being left behind than breaking them.
What's hilarious is that, for years, the enterprise shied away from open source due to the legal considerations they were concerned about. But now... With AI, even though everyone knows that copyright material was stolen by every frontier provider, the enterprise is now like: stolen copyright that can potentially allow me to get rid of some pesky employees? Sign us up!
Yup, there's this angle that's been a 180, but I'm referring to the fact that the US Copyright Office determined that AI output isn't anyone's IP.
Which in itself is an absurdity, where the culmination of the world's copyrighted content is compiled and used to then spit out content that somehow belongs to no one.
But it is a solution. Apple being a poor stuard of their customers is indicative that people buying their hardware and software are not their priority. Apple support used to be stellar, they used to care about customers, they no longer do.
Apple's ToS should be readily indicative of anyone using any of their products that Apple's perspective is that you don't own anything and they can do whatever they want with anything you do with their products. As the author points out you clearly don't own free access to what you've purchased.
The last thing I'll say is that it is fantastic advice to not purchase Apple in 2025. You can only be certain that this won't happen if you avoid them. I actually own a MPB, with receipts from purchase, that I had to purchase a bypass for when the device was enrolled in MDM by a family member that Apple has MDM locked and refuses to remove from iCloud.
Avoid Apple, that's the best advice. If you can't avoid Apple, minimize your footprint and make sure you're a good boy or girl else Tim Cook will steal from you and hide behind some bullshit first line support tar pit and an army of lawyers if you do happen to decide to threaten them.
But, at least with Google you can use hardware without the binding software requirement. You can use an Android device with GrapheneOS and have the phone entirely de-Googled, yet still use Android apps.
If the implication was that there's no other option outside of Apple and Google then that is unfortunate.
> But, at least with Google you can use hardware without the binding software requirement.
For now, but they are tightening things up.
And at least with Apple they provide convenient end-to-end cloud syncing. Google doesn't.
(And this back and forth can go on for a long time...)
You are just picking what is important to you and then ignoring other issues. That isn't how to craft advice that helps people you don't personally know, with needs you are completely unaware of, in a complex domain.
Ok so now we’re not only boycotting Apple, we’re boycotting banks as well! Seriously, Apple can and should fix this issue without having to retort to misery for everyone.
Apple could release a statement reassuring people that no one will be locked out of their account for redeeming any gift card going forward. We have collectively forgotten that companies have stopped talking this way. That’s what we need to change.
I mean, yes, absolutely. I don't have a count limit on my boycott list. I won't be holding my breath for empty promises from corporate. We need to build systems that assert user sovereignty and prevent abuse, not wait around for evil people to do good things.
Sure. And this is not that. This says: before we begin to think about our policy let's make sure to remove any barriers for Mr. Altman and friends so that they don't get sucked down with their Oracle branded boat anchor.
If this had any whiff of actually shedding light on these needed regulations the root OP wouldn't have said what they did. But for now I'm going to head over to Polymarket and see if there are any bets I can place on Trump's kids being appointed to the OpenAI board.
So, just an advertisement for what you built? As for the apps, what's so great about them? I'm genuinely curious.
With respect to the microphone test site I don't need it as my OS provides everything I need for this and I also don't trust your site (that's just by default for what you're asking to have access or my machine).
As for the speed test, OK? There are far better options that already exist and are fully open source.
Building things that are trivial, or already exist aren't exciting. It's great that you feel you went from MVP to "full feature". But IMO both of these are MVPs as they stand. They're not worth much to anyone but you, most likely.
The final thing I'll say is both of these examples have the vibe coded look. It's just like text, images and audio now: AI content is easy to pick out. I'd gather things will get better, but for now there's low likelihood I'm interacting with these in any meaningful way and zero chance I'm buying anything from sites like these.
These offer nothing but free services. Even if they have a vibe coding feel. The ctr is dismal anyways from HN. Its simply astonishing the rate these allow of development, yet it seems the vast amount of people don't see it. Crazy
Building your own utopia is now ever more achievable than I think any other time in human history. Perhaps my examples aren't as stellar as what's possible to achieve. But I only use those as simple examples, there are far better, made by many others whom don't comment on this forum.
I've hosted Immich since it came out and all my photos have been migrated to it at this point. I would never host Immich on NixOS (and I do use it for certain things). The reason? It's not simpler than a container option and creates a single point of issue. The container option is tested and supported by Immich, they recommend it. So everything I need is part of that. I moved servers midway through the year and the storage for my Immich implementation is NAS hosted and the mount is simply exposed to the Immich container. It took me less than 15 minutes to move Immich. And while that would have likely been the same with NixOS it's actually more of a chore to roll back with Nix. My Compose file is locked to major/minor and I choose when to do upgrades. But rollbacks are actually simpler IMO. I just stop the container, tar the operational directory, flip the bits in the Compose file and restart. I've not actually had an issue with Immich ever while doing it this way and I manage about 10TB of photos and videos currently in Immich.
I actually thought about doing this with NixOS last year, but it seemed counterproductive compared to how I self-host, I don't want to manage configurations in multiple places. If I switched everything it would likely be just as much work and then I'm reliant on Nix. Over the years I've gone from the OS being a mix of Arch and Ubuntu to mostly just Debian for my self hosting LXC or VMs. I already have the deployments templated so there's nothing for me to do other than map an IP, give it a hostname and start it.
To each their own, but I don't want to be beholden to NixOS for everything. I like the container abstraction on LXC and VMs and it's been very good to minimize the work of self-hosting over 40+ services both in my home lab and in the bare metal server I lease from Hetzner.
> It's not simpler than a container option and creates a single point of issue. The container option is tested and supported by Immich, they recommend it. I don't want to be beholden to NixOS for everything.
I think there's a misunderstanding here. You aren't beholden to NixOS here. You don't have to use nixpkgs nor home-manager modules. You can make your own flakes and you can use containers, but the benefit is still that you set it up declaratively in config.
It's not incompatible with anything you've said, it's just cool that it has default configurations for things if you aren't opinionated.
> I don't want to manage configurations in multiple places.
I've accumulated one big Nix config that configures across all my machines. It's kind of insane that this is possible.
Of course, it would seem complicated looking at the end result, but I iterated there over time.
> I think there's a misunderstanding here. You aren't beholden to NixOS here. You don't have to use nixpkgs nor home-manager modules. You can make your own flakes and you can use containers, but the benefit is still that you set it up declaratively in config.
There's actually no misunderstanding and this is exactly my point. With any Nix config you are beholden to that specific platform. What I'm saying is any other Linux distro can be dropped in with almost no changes in my existing implementation. I've already experienced breaking changes pre-Flakes with Nix and so I don't actually view it as stable as other options. Beyond that there's some politics surrounding Nix that I don't care to follow. So when you say "all you have to do is write your own Flake"... Why? I already have something that's able to be reliably reinstalled in minutes if need be. I don't need a specific declarative set of tooling to get there.
I like the idea of a declarative setup, but I don't think Nix is mature enough nor does it bring enough differentiation to the plate to be worthwhile as of yet.
The unfortunate reality is that HBO may have less content but there's also less garbage. I'm constantly blown away by how mediocre everything on Netflix is. I only have it because it's bundled into myobile bill at a legacy discount which makes it only a few dollars a month. I wouldn't pay full price for Netflix now and I will likely remove it altogether if they do another price hike that adds a few more dollars beyond my current discount (~70%).
Forgot to reply, dunno about what is the focus in different markets but in Sweden the HBO-Discovery merger meant that 3 of the 12 or so old linear channels got sucked into it, what those channels lived off was cheap reality-TV of drunk/famous/naked/whatever.
I got HBO to watch good shows and suddenly the frontpage was filled with promotions of this inane reality-TV crap that I didn't have any interest in really paying for.
Yes, Netflix has added some reality crap like those fixed Tyson boxing fights and similar, but the ratio of crap to good content was still in heavy favor of content (and most importantly, they didn't try to fill my front-page with the crap).
> HBO may have less content but there's also less garbage
If you leave the featured areas and venture into any of the categories, you will see that HBO is also full of junk. HBO -> Browse by Genre -> A-Z -> any of them are full of junk.
The Netflix featured pages are more geared to showing you stuff you haven't seen yet, while HBO is geared toward showing you popular stuff, even if you have watched it on HBO.
And that buys you what, exactly? Your point is 100% correct and why LLMs are no where near able to manage / build complete simple systems and surely not complex ones.
Why? Context. LLMs, today, go off the rails fairly easily. As I've mentioned in prior comments I've been working a lot with different models and agentic coding systems. When a code base starts to approach 5k lines (building the entire codebase with an agent) things start to get very rough. First of all, the agent cannot wrap it's context (it has no brain) around the code in a complete way. Even when everything is very well documented as part of the build and outlined so the LLM has indicators of where to pull in code - it almost always cannot keep schemas, requirements, or patterns in line. I've had instances where APIs that were being developed were to follow a specific schema, should require specific tests and should abide by specific constraints for integration. Almost always, in that relatively small codebase, the agentic system gets something wrong - but because of sycophancy - it gleefully informs me all the work is done and everything is A-OK! The kicker here is that when you show it why / where it's wrong you're continuously in a loop of burning tokens trying to put that train back on the track. LLMs can't be efficient with new(ish) code bases because they're always having to go lookup new documentation and burning through more context beyond what it's targeting to build / update / refactor / etc.
So, sure. You can "call an LLM multiple times". But this is hugely missing the point with how these systems work. Because when you actually start to use them you'll find these issues almost immediately.
To add onto this, it is a characteristic of their design to statistically pick things that would be bad choices, because humans do too. It’s not more reliable than just taking a random person off the street of SF and giving them instructions on what to copy paste without any context. They might also change unrelated things or get sidetracked when they encounter friction. My point is that when you try to compensate by prompting repeatedly, you are just adding more chances for entropy to leak in — so I am agreeing with you.
> To add onto this, it is a characteristic of their design to statistically pick things that would be bad choices, because humans do too.
Spot on. If we look at, historically, "AI" (pre-LLM) the data sets were much more curated, cleaned and labeled. Look at CV, for example. Computer Vision is a prime example of how AI can easily go off the rails with respect to 1) garbage input data 2) biased input data. LLMs have these two as inputs in spades and in vast quantities. Has everyone forgotten about Google's classification of African American people in images [0]? Or, more hilariously - the fix [1]? Most people I talk to who are using LLMs think that the data being strung into these models has been fine tuned, hand picked, etc. In some cases for small models that were explicitly curated, sure. But in the context (no pun) of all the popular frontier models: no way in hell.
The one thing I'm really surprised nobody is talking about is the system prompt. Not in the manner of jailbreaking it or even extracting it. But I can't imagine that these system prompts aren't collecting mass tech debt at this point. I'm sure there's band aid after band aid of simple fixes to nudge the model in ever so different directions based on things that are, ultimately, out of the control of such a large culmination of random data. I can't wait to see how these long term issues crop and and duct taped for the quick fixes these tech behemoths are becoming known for.
Talking about the debt of a system prompt feels really weird. A system prompt tied to an LLM is the equivalent of crafting a new model in the pre-LLM era. You measure their success using various quality metrics. And you improve the system prompt progressively to raise these metrics. So it feels like bandaid but that's actually how it's supposed to work and totally equivalent to "fixing" a machine learning model by improving the dataset.
The Kobo devices are truly worth every penny and we've got 4 of them in our household at this point. These are some of the best devices to put in the hands of kids.
reply