I switched to my own modem and router recently for privacy from my ISP and it was a fantastic experience / worth it but it cost some money and time which can be hard to find.
I think many are overstating how much people are giving up. People exchange control for comfort, but most people never had any need or ability for this control in the first place. That's why cloud-services became popular, and remain popular.
> Unfortunately it’s a fair bit of work to reclaim everything as your story shows.
This work would be necessary anyway, that's the whole reason why people prefer letting other people doing this work.
> I switched to my own modem and router recently for privacy from my ISP
I'm curious, which privacy can you regain from an ISP, who is already seeing all your internet-traffic? And are we talking here about separate modem & router?
I can--and did for the better part of ~15 years--run and maintain my own self-hosted everything (hardware, DNS, SMTP, httpd, etc, etc, etc). Then I got married and had kids and went to grad school and had a demanding job where I was doing many of the same things I did at home.
I just fucking don't have the personal time nor desire to manage that shit any longer. Why? Because I have better things to do w/my free time than fuck around with my homelab (or whatever the in-term is these days). When I'm done with work, I just want to go outside or read a book.
I am VERY WELL AWARE of the risks and privacy implications; but, my actual freedom from the day-to-day is worth far more to me at this point in my life.
My self-hosting infrastructure will probably outlive me.
I personally relate to the person you're replying to. I sleep better not worrying about HDD health or if my APs can reach their controller. Tried it - not for me.
Regarding need: strong disagree. I want to be able to re-read a book, to open it in any an ebook reader on my desktop to search / copy from it, etc. I want to re-watch good movies any time. I certainly don't want to lose my photos or any media I produce because of some corporate policy or quota, or politics.
I self host everything. I only buy what can be de-DRM'd and if it can't be, I return it immediately.
Regarding ability: Sure it's a bit of a pain, but it's not that hard if you're just a bit technical. Everything is done via GUI, there is never anything to type in a console. And if you're not technical yourself, you probably know someone who is.
That's your demand, not everyone's demand. And it seems are also indirectly assuming here that this is impossible without self-hosting, which also is not necessarily true.
The problem, is, we don't know. Self-hosting is like backups, it's working for a situation which might or might not happen; it's annoying, and it can save your ass, but most of the time you will never know if it ever will save your ass, until it actually happens. And until that point, it's just annoying. So we usually don't know if we really want to re-read a specific book and whether it has been become unavailable for us. We simply don't know that, until it happens.
> I certainly don't want to lose my photos or any media I produce because of some corporate policy or quota, or politics.
True, but that's why you should have backups. You don't need to manage a whole infrastructure for all your stuff, when you can also just make regularly backups. Of course, to be fair, most people don't even make backups, or know how to manage them well. But I would say those people can't (or should?) self-host their infrastructure anyway, they would probably blow their own data up in one way or another and lose them anyway.
> I only buy what can be de-DRM'd and if it can't be, I return it immediately.
See, that's your stance, most people don't give an f** about this. They want things now, and don't care for some uncertain future.
> but it's not that hard if you're just a bit technical.
Which most people are not. But it's not about the technical ability, self-hosting is mainly a problem of time, money and habit. Yes, many people can get it done if they invest into it, but they don't, many can't. And that won't ever change.
Separate modem and router. Using my own modem kicks out my ISP from individual MAC so they can’t see as much device level info. Plus they wouldn’t let me setup a guest network. And now I can monitor the devices myself which is mostly for fun. I run a device VPN when I don’t want them to see traffic but I’ll likely set it up network wide when I have time, which I couldn’t do on their system.
Or, because they do not know and do not care what is happening. Yes, they only care about comfort, who reads TOS anyway, right?! : /
But if the same was happening to their physical not digital properties then they might be furious.
However, those ISP branded modem/router devices are completely backdoored and can be accessed by ISP employees for remote support. As they are your router they also get to see your internal network traffic. HTTPS traffic remains encrypted of course, but I personally would never let an ISP have access to my hardware.
I have some self hosted things, but because of the above I'm realizing that it is better to find someone to pay to take care of things for me. Someone large enough to get a sysadmin around 24x7, do trail upgrades, write the software/features... Unfortunately finding someone you can trust to do the above is important, and for many things there is no option.
I will likely always run jellyfix (or similar) for legal reasons. However for most things it would be better to pay someone I trust.
- You access your server using Tailscale VPN, he mentioned it.
- You can allow external access to your apps safely using cloudflare tunnel (per app). Immich works exactly like Google photos and there's even a really good app!
- Each app is in its own container sandbox, with basic hygiene and monitoring it should be fine. And you aren't a profitable target anyway.
- Update require to restart the container with the latest release, your data isn't erased. Solutions such as Umbrel have a community of open source devs doing the updates for you.
Overall, it's not about removing all of our dependency to commercial services, but to do the switch slowly and regain autonomy. Having an alternative, however how imperfect it is (Jellyfin often freezes for me!) is worth it - otherwise the future is bleak.
Everyone go checkout immich.
To the point about people not knowing how much they've given up, I think another way to phrase this is that people don't know how much has been taken away from them. This is why we need better consumer protections for internet services.
That’s not to say there aren’t issues of ownership and control to be concerned about, but they are providing real value to many users, especially those who aren’t technically minded.
But if I where to do such a thing:
1. Cloud only used to send and store locally encrypted compressed backup data
2. Open an ssh port to the public, but deny logins. Only allow logins using ssh keys.
3. Download data from my system using sftp/scp
This protects you from being chased by DRM lawyers because the system is not public. Plus it is very simple to setup.
Setting up a Cloud System like described here is very great for end users, but it could get you into court, or at the very least lots of take-down notices.
Here is an overview of how the payments work: https://qbix.com/ecosystem
And here is the software you can try for yourself over a weekend: https://github.com/Qbix
If any of you do, let me know what you think!
I have interviewed a lot of people on my channel, including founders of Freenet and MaidSAFE (now called Autonomi) which do in fact replace “the cloud” already, through entirely peer-to-peer nodes.
https://news.ycombinator.com/item?id=34179795
If anyone here knows Ted Nelson, please put us in touch! I would love to interview him about his vision for Xanadu
For my part, however, I am embracing a different model, where a “QBOX” black box would be hosted by our franchisees in the cloud, among other places. Placing the protocols inside the EC2 instances makes them untouchable by Amazon. Because AWS, Google et al legally are not allowed to go inside those boxes and mess with the software, or even read the contents of the RAM. And I don’t remember any story of them ever doing it even for the NSA. Do you?
Is this meant to be tongue in cheek?
Do you have links to stories of AWS breaking into EC2 instances to eg read RAM for data that is encrypted at rest?
And even if they do, this would present an issue for privacy, but the protocols would still enforce their own permissions (eg no custom amazon DRM for books).
From 2015, AWS asserted they were not involved in the PRISM program, but they would be under a gag order if they were, so you've gotta take it with a grain of salt: https://www.crn.com/news/cloud/300077146/aws-finally-release...
Meanwhile:
> From the start of this calendar year through May, AWS received 813 subpoenas from the U.S. government seeking access to customer accounts. In those five months, the Seattle-based cloud provider fully complied with 542 of those court orders, submitted partial information in response to 126 and didn't respond at all to 145.
> Through the same period, Amazon received 25 search warrants from federal authorities and turned over all the data sought by about half of them, partially fulfilled eight others and withheld information requested by four of the warrants.
> AWS fully responded to only four out of 13 court orders that weren't subpoenas or warrants, while refusing to turn over any data related to four of those.
> Foreign governments were more successful with their solicitations to Amazon. Of the 132 non-U.S. requests fielded by the cloud provider, more than 80 percent yielded complete data disclosures, while just 13 percent hit a dead end. Amazon also complied with the only request it received during the five months under review to actually remove a user's data from its servers.
But no one has to run their own servers. The only reason I see them doing so is to provide redundancy in case the cloud providers want to DELETE some data or take nodes offline.
Decades ago no one pumped their own gas.
Using a phone as a mini computer was possible. Downloading and using apps happened. I even used offline maps. It was still the preserve of nerds while regular people "couldn't understand why you'd use a phone to do anything other than text and call".
SUDDENLY once it became seamless and trivial to set everything and it was all brought together on a device that was aesthetically pleasing and ergonomic demand rocketed upwards. It turns out that regular people very much wanted a mini computer in their pocket.
This all took me very much by surprise coz almost everything that was revolutionary about the iPhone... I was already doing all of that while it was announced.
I think self hosting is in a similar spot right now. The apps exist (many are extremely nice!), the software exists, but the seamless, aesthetically pleasing and ergonomic experience does not. It's a pain in the ass to set up self hosting.
I don't think the "advanced users" market has shrunk much, it's just the whole pie became so much bigger that the overall ratio decreased.
Phones are amazingly powerful. Why not "self host" apps on phones?
Where it gets complicated is there's a (totally understandable) expectation these days that your data is synced across multiple devices, and you can collaborate with other users, who may also have multiple devices themselves. In practice, that necessitates some kind of always-on server that maintains state for everyone. A phone can technically do that, but you'd probably kill your battery in the process.
- Battery life. One of the main reasons your phone lasts as long as it does is because it severely restricts the ability to run always-on things. A phone of course can run an email server, but the battery life will immediately tank to the point where the device becomes largely unusable for its original purpose.
- Phones make extremely poor servers because connectivity is intermittent. This is fine for software that's 100% local, but a lot of the most useful software needs to talk to the internet - or more importantly, has to allow the internet to talk to it. Imagine losing an email because you walked into the subway and your phone was unreachable the moment an SMTP server tried to connect to it.
At that point, you're better off going with some N100 mini-PC or such. But that's not a phone.
And I'll remind folks that we've been talking about the power of people owning their own servers in their homes for decades, and yet the vast vast vast vast majority of users aren't doing it.
> Imagine losing an email because you walked into the subway and your phone was unreachable the moment an SMTP server tried to connect to it.
Dont SMTP servers already retry a few times before giving up? Plus it is not like you're using the phone to host content for the whole of the internet. It would be just for your close circle usually.
I am not saying phones make the perfect servers for all kind of applications but for certain kind of workflows... I think Phones are pretty good. Our network infrastructure (NAT, firewalls etc... limited data plans etc..) is the main headache for most of these use cases. But the network infrastructure is a problem even for our laptops, home computers etc..
I could imagine self hosting becoming more accessible but don't see how it could become mainstream when it's just an alternative to stuff that's already available in the cloud
They're already doing that on the hardware side. https://arstechnica.com/gadgets/2025/04/synology-confirms-ne...
- There were OOTB features on Mac OS X such as web page building and publishing
- There was Mac OS X, but there was also Mac OS X Server, a full-fledged, easy(-ish) to use solution to self host mail, calendaring, and so on
- There was Bonjour a.k.a Zeroconf, not just on the LAN but global as well.
- There was Back to my Mac and most importantly the technology underneath it which was essentially a "one switch Tailscale". Combined with the above you could SSH to any of your Macs from any other Mac you were logged into wherever it might be, Back to my Mac was merely VNC'ing/SMB'ing over that private overlay network.
- There was the quite budget friendly Mac Mini
- also, Airport Express/Extreme/Time Capsule, if you had one of those BtmM would magically WoL sleeping Macs.
- The Mac App Store was introduced
- Affordable residential FTTH started rolling out widely with solid downlinks+uplinks
And around that time I was god honest thinking: "these are all pieces of the same puzzle... next step they might turn each of their server features into separate server apps, and bootstrap an app store out of it for third parties to create and publish their own server apps, and everyone and their dog could have their own server of anything at home"
Instead things were dialled up to 11 towards datacenters.
I do think this is within realm of possibility if Steve Jobs is still alive. Or at least could be convinced.
Tim Coo only cares about services revenue. And iCloud it is.
https://www.macrumors.com/2011/05/07/steve-jobs-reaction-to-...
I remember my friends and my tech fiend cousin sneering at the iPhone when it was launched for this reason. I got heckled for “overpaying” for an inferior product when they learned I bought an iPhone.
Yet my actual phone computing experience was mostly better than theirs with a few notable lags (copy and paste). They had a different idea of what the iPhone was like than my actual experience and they refused to believe anything else.
It was like they lived in a world where your phone choice was your identity. They saw themselves as being at the top of the phone ecosystem and having made the right choice. They simply would not allow any other phone to be good because it was an attack on the narrative at the core of their identity.
At the time I just didn’t care. My iPhone worked well and I wasn’t interested in endless playing with all the customizations and changes they were doing on their phones. It got the job done and I liked how it worked.
I think self hosting is similar: The people drawn to it think their setup is the pinnacle of computing, but many of them have been so out of the loop on modern cloud services that they’ve forgotten what it’s like to use a cloud service that works well. They’re stuck believing it’s all useless eye candy on an inferior product.
I even see the same thing when I use Mastodon. The whole federation thing is a massive drag. Having to do the dance to follow someone on a different server gets old. I miss being able to one click follow someone and not have to pay attention to what site I’m on. Yet bring it up to fediverse fans and many will scoff at the idea that it’s a hassle at all. They might argue it’s a small price to pay. So many refuse to admit that it’s not a good experience. Situations like this run deep in every self-hosted or distributed project I’ve seen. They cater to people who enjoy fiddling with projects and debugging things.
I self-host the following:
Video: Jellyfin
Audio: Navidrome
Audiobooks: Audiobookshelf
Phone image sharing: Immich
Home automation: Homeassistant
Office suite: NextCloud
Monitoring: LibreNMS
Compute: Proxmox
AI/LLM local: open-webuijellyfin is configured to not transcode anything. The vast bulk of my library is DVD/BluRay rips of my own creation and I just ripped them in the desired format in the first place. This could probably keep up with a single DVD-quality re-encode, I dunno about Blu-Ray (depending on config, perhaps), but I just have it serve the correct files in the first place.
There's a ~$125 5TB USB drive hanging off of it for the media storage, which syncthing syncs to another 5TB drive in the house. (I don't actually "back up" my media storage in the full sense; everything else is actually backed up in the full sense to S3 via restic.) The "contention" I mentioned above is because all the big data sets are mostly on that spinning-rust drive.
The Immich AI features worked fine on this, though it did take overnight to process my initial load of ~20 years of photos. However once it chewed through that, the responsiveness is fantastic.
If you want responsive AI that uses GPUs this isn't anywhere near enough, but for any "conventional" app, $125 or $250 buys you a lot nowadays.
I ran an n100 until last week. Worked fine.
I have plex setup to transcode and it serves about 10 users just fine. My plexamp sonic analysis took like 4 days though, lol, but everyone says it takes forever.
My immich import took about 20 hours? So not bad.
I run all my home automation off it. 100+ devices, logging, etc. no issues.
I also sometimes run an OBS stream on it to transcode for YT. The n150 does fine.
Total cost for me is about $550. I saved a lot on HDD by going used server drives. $140 for the n150, $300 for drives, then a cheap UPS and router running openwrt.
As for difficulty, most of this is deployed in a few minutes using docker or install scripts. The hardest part is the choice between various solutions.
It handles all but the AI/LLM. I have a throwaway box with 32GB ram and cores with a nvidia 2080 that does the LLM side of things.
nop, but legislators should really force that anything bought without "deadline" also doesn't randomly disappear/cost extra no matter if you bought a license or not
in additions license with clear deadline should always be required to have a "be aware that this product has only a limited guaranteed availability of ... days/month/years _dialog_" which you need to agree on and which isn't allowed to be just another checkbox (which yes seems mean against companies, but their is no reason to not treat scam like, abusive business practices meanly. It's kinda the point of countries to fight against anything harming their citizens weather that is abusive business practices or violence .)
BUT, you are going to be paying a monthly sub as long as you keep using the service. And soon as the service goes down (due to financial or other reasons) - game over man.
So there is still a lot to be said for downloadable software, even if it is no longer cool or fashionable. Pay once. Keep your data secure locally. Keep using it until you can't find a computer that runs it any more.
I develop 3 commercial downloadable software products. No plans to move them to web.
Good for you. But for most people, it is an endeavor with zero gain, meaning no positive impact to their daily life, if not full of negative impact.
> And this week, I want to share with you how I did it, what I learned, and why I think self-hosting is NOT the future we should be fighting for.
You're leaving your entire digital existence up to companies who will and have ruined people's lives.
I think it says a lot about how much we've given up that control over your data and access to your data is seen as "zero gain" or "full of negative impact."
It's wild how little people care about their own rights. Capitalism and hustle culture make it so easy to give up so much while receiving so little in return. The pressure to give up more is constant and people willfully lean into it.
That's old school P2P since 25 year. this is not new and not future...
I think its not the future in its current form either, because it requires too much configuration and maintenance for typical users, although NAS devices do it quite well and easily nowadays. But I also think that the cost of having Amazon et el do the maintenance has resulted in a lot of downtime that wipes out the internet every month or so for hours at a time and with the data theft and abuse and ever increasing profit extraction.
The privileged enjoy far more privacy and autonomy and this is brought into sharp focus with wonderful hobbies like self-hosting. Perhaps it all boils down to end-stage capitalism, and perhaps there's a technical solution where selflessness overcomes end-stage capitalism. Someone else mentioned incentives and yeah, that'll help, but hopefully we'll collectively choose to do the hard thing because it's the right thing. Heck, maybe the right thing will also be the easy thing if we come up with better ideas like yours.
But I want the next thing. Which is like Tailscale2, but for people, not machines.
I want to tell Tailscale2 about all of the people in my life, and which of my self-hosted apps they're allowed to talk to. And if they're also running a self-hosted app, then I want our apps to federate together.
It feels like we're suuuuuper close to having this.
I get that you can basically do this with Tailscale. Basically. But I want the next thing to be designed from the ground-up around this kind of design. People, sharing apps with each other.
If I "Share" nodes on my tailnet with a hundred people, it's way closer to what I want.
If you can't actually download a copy of a digital content as a mere file, then you can't really host it and serve it.
You can't host your own Spotify-clone even if you are allowed to listen to songs. However, you can still download music on Bandcamp to feed your Spotify-clone.
You can't host your own your own digital Video Game Store usually because of various DRM, or because it's painful to "export" the content and painful to "import" it back.
Still on the video game side, You can't even backup your game save (at least on the Nintendo Switch, Nintendo Switch 2 and Xbox Series), it's not because of any copyright infringement or IPs misuse, it's only a way for them to get more online subscription with online game save backup.
There is still a positive side: when it will become impossible to legally own anything, I'm pretty sure some illegal system will enable you to have a massive library of whatever you want at the cost of few clicks and/or a couple of bucks. I'm saying "positive side" even though it's illegal because I mostly talk about the comfort of having your own local library.
For games, there's GOG. Good luck finding bigger releases.
For music, there's Bandcamp and CDs and vinyl. Fortunately, most albums still release on either one of these.
Audiobookshelf can be used for most podcasts (some do not have a traditional RSS feed and are in some walled garden) and some audio books are available DRM free, but tons of books are Audible exclusives. I'm relatively sure that they also stop authors from publishing e.g. on Royal Road once they're on there.
The same is true for e-books - HumbleBundle and co are great, but good luck finding certain titles. I regret buying a new Kindle, but at least had the foresight to download all my books before they stopped allowing that. Physical books are an option, but that's not an equivalent to en e-book.
I stopped caring about TV shows and movies a long time ago (largely due to the atrocious streaming fragmentation, pricing, and the sheer audacity to include ads in paid plans), but I assume 95% of all shows are exclusive to some streaming giant, too.
The website [0] is pretty clear that the content of the game can go into a SD card, but the game save resides only in the internal memory.
You can find some ways to get them with some modding but nothing official.
[0] https://www.nintendo.com/ph/support/switch/data_management/i...
The rise of better home internet connections worldwide will make this even more attainable for more people. At least on my low-level EU country that has been always lagging to progress tech-wise, we've seen great progress on fiber internet adoption, so I have hope of acceleration.
What we need is something more universal, like a more userfriendly docker, or something like flatpak+hub for server-apps.
I should create an account that posts nothing but the phrase "Stallman was right". I'd have work every day.
Anyway, I have a Pocketbook[1], recommended. Got the cheapest one, cost me something like 100 pounds. Doesn't need internet if you don't want it, and supports all the usual file formats.
Beyond that, the "how do I talk to other people if it's on my server" thing is generally solvable. Give them an account on your server. Don't want to need to make an account on every friend's server? That's why we have SSO technologies. I don't think. Self hosting and community collaboration need to be incompatible.
LOL right. I bought an ereader that works for me instead of working for Amazon. No need to run a server.
Totally agree, but there's a lot more nuance here. Giving each friend an account on my server would require it be exposed to the public internet which is difficult to manage securely. And SSO doesn't really make this very convenient because that means everyone would have to sign in and sync to everyone's servers which is a lot of work for the user. It's a UX problem.
The solution as I see it here is services that can interoperate and sync files across hosts. So, my friend's Alice and Bob can both have their photos synced to a separate server and can choose which photos to share to my server. Separate but connected.
Thanks for reading and for your comment!
This could all run on one of those $130 N150 minipcs that uses like 8W and could run 24/7. It's a lot of integration work, but there's no reason why it couldn't be a fairly off-the-shelf product.
You could also explore other service discovery patterns since buying a domain name is a pain. Like have the URL provide the initial wireguard config (including outside IP) and DNS search domain, and then the servers on each end can query (private) DNS on the other end via the tunnel for services.
You're almost there with your excellent lineup of self-hosted tech. Just throw in Headscale and some Tailscale clients and you'll be there. (Or any number of mesh VPN alternatives, like NetBird)
I do like that the author gets into alternatives, like the library storage idea (my similar concept involved the USPS giving citizens gratis space and a CDN). I think that’s a discussion we need a lot more of, including towns or states building publicly-owned datacenters and infrastructure to support more community efforts involving technology. We also need more engagement from FOSS projects in making their software as easy to deploy with security best practices as possible, by default, such that more people can get right to tinkering and building without having to understand how the proverbial sausage is made. That’s arguably the biggest gap at the moment, because solving the UX side (like Plex did) enables more people to self-host and more communities to consider offering compute services to their citizens.
I’m glad to see a stronger rejection of this notion that a handful of private corporations should control the bulk of technology and the associated industry running atop it, and I’m happy to see more folks discussing alternative futures to it.
Complexity, sure. But for most people, the cost of Netflix, Spotify and whatever will quickly add up to a 500usd server. With 1-10 users you don't need much.
For 9 out of 10 self hosted programs you can have them up in ~5 minutes with a docker compose and env file.
There are whole OSes built around it, like casaOS which gives users a neat front end/dashboard for their self hosted stuff.
Also for cost eh idk. For $300 you can have enough hardware and storage to self host everything, even a Google photos alternative. Most people spend much more than that on subscriptions for storage, streaming, etc. I guess a UPS is necessary and adds a bit of cost. There are also plenty of pre-built kits for this.
I do agree that it isn't for everyone. Its finicky to get just right and security can be very annoying. Security is already a crapshoot though so I'm not sure that's necessarily a ding for self-hosted.
That is a very small part of operating. How about keeping it update and running? Data backed up?
Until setting up a private chatroom for your family is as easy as downloading an app on your phone, people are going to keep going back to Big Tech. UX for IT folk and UX for the layman are entirely different beasts, and the UX for IT is only recently improving thanks to things like Docker and the containerization of software making it more widespread and commoditized.
Last time I checked, there are about three hundred thousand different companies offering hosting, all over the world. That's a bunch more diverse than the government doing hosting, as per your suggestion. Or having towns contracting Microsoft for it, which would be the result with kolkhoz or sovkhoz cloud hosting.
Last time I checked, AWS was estimated to have ~5% of all web sites in the world hosted in its infrastructure, while AWS+GCP+Azure combined equate to ~66% of the global cloud compute market. That doesn't even get into the "providers" who are really just reselling major providers at a markup (like Vercel).
It doesn't matter if your town has hundreds of storefronts if one subsidized Walmart is putting them all out of business. Likewise, if every business in town is dependent on the Walmart, then it's really Walmart that controls things and not individual or collective business owners.
Lets connect! Send me an email – hn@drewlyton.com!
> Self-hosting is when you have a computer in your house do those same things
Self-hosting is more about deploying self-selected software onto a server. It can be a server at home, but I for one have a lot of services running on a VPS. Self-hosting is more about control of the data and software, than the location of the hardware.
> Well...since our friends can't access our server, the only good way to do that would probably be using an app like Google Photos or iCloud
Get a domain and set up a subdomain for Immich (maybe add a tunnel if it is a home server). I have friends using my Immich instance without problems, it's just another app.
> I'm talking publicly funded, accessible, at cost cloud-services.
I can't see how one can convince people to switch to a community cloud if Apple Cloud etc. exists. Most people just won't understand the difference or benefits.
As to the "convince people to switch" angle, I think the benefits of data interoperability would be pretty significant and eventually lead people to switch to providers that have that or would likely incentivize providers like Apple to implement that into their products.
Ideally, no one would have to switch and everything would just get better.
> It's secured behind our own VPN.
> So, how do I create a shared photo album with my friends where we can all upload pictures from our latest trip? Well...since our friends can't access our server
On such an instance, one can share a folder with a friend, for instance. And I think Nextcloud is even working on federation (?).
One disadvantage is that they have access to your data, but at least you choose the cloud provider (maybe you want one that is in your country).
You can apparently encrypt your Nextcloud data at rest at Hetzner. I host my own Nextcloud, and I know it supports encryption, but apparently Hetzner also allows you to do so.
On the other hand, if you want a standard cloud provider, pCloud provides good encryption support. Also they have a nice FUSE based client, and they're interoperable with tons of tools, too.
Returning to Nextcloud, you can share files/folders directly (with expiration/password) or add more users with limited access to your folders.
BTW, keeping a Nextcloud instance is really easy, let it be container based or bare-metal install. It never let me down over the years.
Doesn't it mean that they can still access your data while the server is running? I mean, they run the server, they must have access to it, right?
> pCloud provides good encryption support
You mean e2ee? If it's about sending files to an untrusted server, I use restic. Works with pretty much everything (including pCloud) :-).
> BTW, keeping a Nextcloud instance is really easy
Sure, but what I was saying is that either you do it at home and it makes it harder (you want your home LAN to be secure :-) ) or you do it on a VPS, and someone else has access to your data.
I barely trust Google.
I trust the long bearded neighborhood nerd much less than most companies. Even if I probably am that person in my neighborhood. But nobody should trust me, and I am not going to tell them to trust me.
Even if everything is encrypted, I can almost guarantee that the community shared server will be confiscated by the police once in the next three decades.
Thanks for reading and commenting!
What if instead, you just store local copies of your data, possibly organized and synchronized? If necessary it can be done manually, just download anything important enough that you might want it later. If a service decays, then import it into another.
A big point the author makes is that many cloud providers don’t let you download the data. But any media that can’t be accessed outside bigco’s cloud can’t be uploaded to your cloud in the first place. If bigco’s cloud prevents you from downloading data that you create or upload, only then the solution is to use a (possibly self-hosted) alternative. However, in practice I rarely see this happening, for example downloading from Google Workspace and OneDrive is very easy (it can even synchronize a folder on your local machine), and if you’re worried about it happening in the future, again, you can backup important files.
I think of the centralization of content and the licensing as something that works so long as it’s a commodity market, that is, it’s hard to 2x the price of an ebook over a dead tree which I can own. Investors may wish otherwise, but they have to add tons of value to get consumers to play along.
I’m fine with commodities in my life. Power and water and gas come to mind. They cost what they cost and I don’t have problems with it.
I could build a nas and run software and admin it, or I could pay $20/mo to Adobe and another $33 to Apple for my family’s shared storage. Done. Of course, if the benefits of commoditization evaporate and it looks like the streaming market, then I’m wrong and would have to change track.
I think the point is in a delivery of commodities (storage, IP licenses, water, power) there is some benefit from the generally fungible nature of the commodity, which makes it harder to put high prices on them, which makes doing it yourself more expensive and inefficient unless you value something very specific.
It's true I don't own the water from my city nor own access to it (it's a license, effectively), and I pay a delivery fee and purchase units of water. But like most people around, I don't value the intangible of truly owning access to the water under my land and drill a well, I just use the commodity. So it goes with e-book licensing and video licensing, too, and I don't think that they're regulated utilities affects this decision whatsoever - enough people value cost and convenience sufficiently to think licenses are fine for their use case instead of ownership.
>The point of the former is to make money, the latter is to enrich our lives by taking care of basic human needs.
The former could say they make money by enriching lives in their own way.
Is this arguing basic human needs should be charity? If so, even the most humble city will charge for water. Further, companies are often created to make money by providing production and distribution of that human need. Utilities are not altruistic but can be fair enough when held in check by a state.
Thanks for reading and commenting!
It's strange to me that we never included public spaces in our growth and innovation of the internet over the past 30 years. Of course I expect companies to do their thing as they've had free reign to do, but it wouldn't have taken much cost or effort to add a couple publicly funded data-centers where everyone gets a little space for themselves.
At least in the US, I think it's because we've allowed those who run our government to get far too old. The people running the country have not really understood the public good of the internet outside of commerce. Don't get me wrong, I've benefited from said commerce for my entire career, but I think we, as a society, have lost quite a bit of ground by not collectively owning a piece of this thing as it grew.
Once upon a time the airwaves were ours, and music thrived because of it. These days the airwaves are all practically walled off with massive monopolies controlling them. It's an overall detriment to our creative progress.
I know I'm an old man barking at clouds, but I miss the radio from when I was young - there was actually new and interesting music there. The internet feels the same way for very similar reasons.
I have deployed simple UniFi setups for all my relatives, and they are very happy (though they couldn't have done it themselves). IMHO, they have the DNA to go further and offer a full self-hosted cloud, if they're willing to put in the effort to make it even easier and more integrated.
But why should a (public) library be interested in providing such services? For funding? What about costs? On for example censorship/regulations/compliance/maintenance etc?
I'm not so sure a publicly funded library would have any interest in doing that. Think about it, if libraries can/welling to do any of that, then Amazon would never have any chance to grow this big.
I think that's why only private companies is capable of doing it, at least currently. They found out a way to make a profit while operating a sustainable (all things considered) cloud service.
In fact, the at-cost service provided by the libraries will probably collapse as soon as a for-profit company comes up with a cheaper plan.
Also, host by a library still creates centralized service, which comes with all problems that a centralized service inherits. It only shifts the problem, not solving it.
> But why should a (public) library be interested in providing such services? For funding? What about costs?
Public institutions like libraries are usually funded through government mandates. We as citizens decided that having free access to books is a good thing and nations, states, and municipalities dedicate tax dollars to fund those programs. So, if we decided providing internet-based services through the library was also important, we'd enact mandates for that, too.
Not saying that's likely, but it is possible.
> At-cost service[s] provided by the libraries will probably collapse as soon as a for-profit company comes up with a cheaper plan.
At-cost actually means it couldn't be cheaper (at least if economies of scale are equal). That gets a little hairy because companies like Google can provide services like Photos and Drive for "free" because they make so much money selling search data, but generally speaking that's the deal.
> Also, host by a library still creates centralized service, which comes with all problems that a centralized service inherits. It only shifts the problem, not solving it.
Totally agreed – if there was only one library. But, there are tons! And as I mentioned, if the services are based on interoperable standards, you could easily move your data between services and have them talk to each other so there's no vendor lock-in. Think ActivityPub for files.
Thanks again for reading and engaging in the discussion!
I'm afraid that's not how things work.
For example, it is possible for everyone to self-host their own service, it's true and everyone can do that right now. However, that's not what happened.
In reality, people oped in to use these cloud services, with full understanding of it's downsides, in exchange for convenience and low cost for themselves.
And as I've pointed out in my last comment, the companies has optimized their services so well, it made it very hard for a library, which is a "outsider" in the service field to compete. The library and it's lacking of technical know-hows, political resources etc will eventually doom the service, making it a product people only buy with higher-than-market price as a show off of their goodwell. That's not a sustainable business model or any model, really.
You must know all these things if you want to change the world for the better. Dreaming on vague an idea is easy, making things actually work is hard.
Also, you need to reconsider the meaning of the term "self-hosting", because unlike what people widely believed, "self-hosting" is not equivalent to "setup and running a server by yourself". The definition is much wider.
In fact, if you ever downloaded something from the Internet and storing it on your computer for later use, that's self-hosting, you just not sharing it with remote access. And if you copied what you've downloaded to multiple computers, then you've just created redundant and distributed backups.
But if you taking that into consideration, that downloading is self-hosting, then there's even less reason to use self-hosting service provided by a library, because why should you pay extra for all of that if you have already done it with no added cost? Just for a expensive remote backup maintained by people who has other jobs?
- "if there was only one library. But, there are tons!" - This is just empty dream. Not many will be welling to do it in reality.
- "if the services are based on interoperable standards" - Designing protocol is not easy, and by simply defining the protocol you also restricts what the protocol can do. A file sharing protocol? Good luck getting it to also do bookmark or chat. So, empty dream based on empty dreams?
I do feel you have a good heart and wanted a good change, but you look inexperienced. I would recommend that you learn the industry or simply work in it for a few years, then maybe you'll come up with something that actually works.
You could stream content from it over your home network (as long as you were connecting from another Apple device)
Is this lost technology or just a figment of my imagination? I've long since switched to linux and run the typical Jellyfin setup etc
You do exactly that.
"Oh but security."
Any security you get from hiding behind a firewall is illusory at best. You still need to keep on top of updates and tech news. And I want to be able to access my stuff from wherever too.
Most of my friends don't have to, because they have me and at least 3 other friends who also self host.
There's a couple of things I won't let others in (like, my email domain. That's like my last name, so nope). But things like _sharing a video_? Yeah, I'll let them log in.
All my pictures are stored as plain files in various folders on a big networked hard drive. So is all my music, audiobooks, movies, documents, projects, etc. This is backed up 5 times over to more hard drives periodically. I give a couple to family that lives out of state when I visit.
You might laugh, but I'm not really sure what I'm missing that would have me do something else. And yes, it's work to take care of it, but that's true of any of your possessions. Just give me my files, man.
Thanks for reading and commenting!
Honestly - just make the service public. Let your wife share links to her photo albums with her friends - have them point to your domain.
Make your friends make accounts on your services if they need to - or better yet, provision accounts automatically for them (I do this).
I understand the fear here, and I get it, but I also think it's widely misplaced. Pay a small sum for backups, rotate them, and let it rip.
The suburban web is actually pretty good these days (at least in real suburbs, I have 2gbs/down 1gbs/up in mine) and it basically only gets better.
---
My experience comes from hosting several sites for my family (including extended family in several different cities and countries) and also several sites for my neighborhood. The vast majority of them are public (as in - there is a public domain that resolves to my services with no need for preshared secret [aka: tailscale or other wireguard based vpn]).
Yes, you get clearly bogus traffic scanning for the lowest of low hanging fruit (ex - php_myadmin/wp-admin/etc) but auth solutions have come a long way, and I don't even bother blacklisting/fail2banning anymore. It's a waste of time and effort for small peanuts.
It's pretty easy to configure SSO pointed at something like Keycloak/Authelia and then have your friends get a centrally managed account with 2fa required. Ex - Jellyfin, Bookstack, Gitea, Immich etc... I host all of these (and lots more) and SSO support is pretty good these days.
Personally, if all your public infrastructure is behind a keycloak login form... I don't think you're going to have many problems.
---
Side note - this is one perfectly acceptable strategy to reach the point you want (community based self-hosted solutions). I host services for my neighbors & family. Not every household needs to be an expert, and no need to get the gov involved (not that I mind the idea of a new digital services library, either).
But fear of the public web means you can't ever reach that spot.
You would put two in different locations for redundancy and it begins to be a personal “cloud”.
Another option is an app like nextcloud. You learn it and it does everything 80% as good as possible, which is often more than enough!
Certain things will be cloud-based or otherwise provider-hosted. Some things will remain self-hosted, for those who prefer it.
It's like owning a car: you take the trouble to maintain it, but it's yours and will take you where you want, without the limitations of a taxi or even a rented car. I live in NYC and don't own a car, for I have too little use for it. OTOH if I were a plumbing contractor, I most definitely would own a car, or maybe a light truck. One size does not exactly fit all.
Who is doing this anyway? Nowadays everyone has his instagram profile on private and if you need to share some pics, you do it via Airdrop. lol
During the Enlightenment, owning a physical copy of a book meant intellectual freedom. You didn’t rent ideas; you had them. Today, most digital knowledge is hosted, locked, or streamed — *leased from platforms*, not owned. We’re in fact drifting into *digital feudalism*, where access to culture, tools, and even history depends on gatekeepers.
In a perfect world this should go beyond market logic. It’s not just a question of what's sustainable or profitable. It's about *civic autonomy*. If the infrastructure of knowledge is centralized, then so is control over thought.
Self-hosting may not be for everyone, but *distributed, open systems are essential* to preserving a democratic and durable digital commons.
But if I’m being honest, I think this claim that if you don’t own the book you don’t have the knowledge and society will turn into digital feudalism is hyperbole. Knowledge is proliferating faster than ever, becoming more accessible than ever, and it’s easier than ever before to get the info that you’re searching for, even in this streaming world. The idea that I’m going to lose knowledge from a book I read 5 years ago if it disappears from my library just doesn’t track. In fact, it’s rare that I return to my physical books these days because I can find equivalent info faster from a quick search online.
Don’t get me wrong: I prefer having my own copies and so on. However, when people start throwing around concepts like “digital feudalism” and trying to draw parallels to the enlightenment it feels like this is all some abstract philosophical debate rather than a discussion of what’s really happening in the world.
Media is being deleted or locked in vaults.
Games are being shut down with no way to restore them.
The written word that has been vetted by people with domain specific knowledge is being locked behind paywalls and not being advertised, while AI machines directly lie to the curious and the seekers of knowledge.
I can throw a digital stone in any direction and hit something that is worse off thanks to the modern internet.
I would have agreed with you a few years ago. But now Google, DuckDuckGo etc. at most provide 3 pages of results, with many irrelevant or wrong. There are alternatives:
https://wiby.me/ https://clew.se/ https://kagi.com/
But that's not the majority experience and more importantly, it shows that it really can be "taken" from us.
The real problem with this is that there are vested interests at play in managing what information you see first - push something to the 2nd or 3rd page of google results and it becomes effectively invisible, especially when you have pages and pages of results that seem to push the narrative that those vested interests want you to see.
I tend to think that Huxley was right over Orwell, information is lost in the shuffle of distraction and rigged systems. The "truth" is there to find, but it's a needle in a haystack of believable lies, and those lies were crafted specifically to obfuscate that nugget of truth.
So the amount of information moving around is irrelevant if it's not useful, or it's intentionally misleading from something that might upset those who benefit from the status quo.
Information is proliferating and is more accessible, but a huge amount of that information is lies and manipulation I'm not sure that really counts as knowledge.
> The idea that I’m going to lose knowledge from a book I read 5 years ago if it disappears from my library just doesn’t track.
You might not forget what you learned from a book you read 5 years ago after it gets stolen from you, but it does mean that others are cut off from that same information. Worse is that what you saw 5 years ago might still be made avilable, but only in censored/altered forms which could easily have you questioning your memory of something you read or saw just 5 years ago.
It's not just an abstract philosophical debate that books and other forms of media are being changed, censored, or removed entirely. Or that gatekeepers want to decide what we're allowed to see and extract rent from us every time that we do. The dangers are real and understood and very much present in today's world.
I don't think that's any different to any other period of time when communication was suddenly able to expand. Gutenberg's press didn't come with an automatic lie detector that meant the printed word could only contain true facts and nothing else. Instead, it was mainly used for pamphlets and other campaigning propaganda - some of which surely had some truth to it, but much of which was partially or fully fabricated.
I think you are romanticising the past's approach to the written word here. It has always been possible to completely rewrite history, if you're willing to put the work in, and totalitarian regimes have had no issues in convincing their populations to burn their own books if necessary.
This has never been possible in the past.
Is there something materially different here with the internet? Are we now entering an era of too much free speech? Is it now too easy for us to communicate with each other? And if so, what's the cutoff? What arbitrary barrier would we need to put in place to make the internet more like the printing press and safe to use again?
Yes. There are effectively no hard copies. It is possible to change the historical record of any non-printed material to suit your particular needs like never before.
You can think of this as a world beyond Orwell's or Bradbury's wildest nightmares.
I would argue that the opposite is true: it is now harder than ever to change the historical record, which is why we now talk about hypernormality and post-truth, where even if there is evidence for something, people will still lie and claim the opposite and be believed. We live with an abundance of evidence, and yet the Orwellian ability for people in charge to tell you one thing one day, and another thing the next, has never been stronger.
And I think you're again making the mistake of thinking of the printing press as a device for printing books or other materials designed to be long-lasting and valuable. In practice, the printing press brought about a revolution of flyers and pamphlets - ephemeral documents that were distributed one day and then abandoned the next. These things should change freely, and many never entered the historical record at all.
The result is a kind of anti-literacy. Most people can read the words, comparatively few people are media-literate enough to filter truth from lies with any reliability. So the current media landscape is unusually poisonous. It's mostly vested interests lying to you and trying to manipulate you, through ads, troll farms, and mainstream media.
The fix would be AI filtering of content. Right now there's no chance whatsoever of that working accurately, but it's possible in principle to counteract the rise of AI disinformation with AI critiques of it.
Among all of the other revolutionary changes promised by AI, that possibility has flown under the radar. But it would be a political and economic showstopper if implemented, because everyone would suddenly be seeing authoritative, accurate news and analysis - like old-school fairness doctrine journalism, but better because it could presented at a level that matched the reader, while also allowing questions.
Ironically Grok was doing something like this for a while, until it... wasn't any more.
Steam library not so much, most likely they will have to re-buy the games because even if they inherit or I just leave credentials and 2FA I can imagine someone there in business thinking "hey this account is 100 years old, we should clean that up, unless guy sends us birth certificate and proof he is still alive.".
Are you serious? To judge what's happening in the world, you need a framework. And the frameworks provided by the ubiquitous commercial and political interests are all biased in comparable ways. Abstract philosophical debate is just what the doctor ordered to get you away from the incessant assault of propaganda and brain washing.
You are talking about preserving intellectual independence.
Both are nice to have, but they are sort of different problems, right? Yours seems more important. And yours could probably be solved by a local copy of Wikipedia and an FTP server full of digital textbooks.
IMO one dangerous misstep we can make with self-hosting is to assume we need to start by matching the centralized services look-and-feel and polish (which is getting worse every year anyway).
That's an interesting take. I think matching these services isn't a necessity, but getting a polished look-and-feels just helps adoption. Adoption isn't an exclusive scenario and everyone is free to choose and mix how they see fit.
My private collection won't ever compete with Netflix, Google or the like, and that's completely fine. It will stay a private selection of media with a strong personal preference - it ranges from research to entertainment, and also includes stuff that documents my own individual history. It'll shrink and grow as I want it, and if it reaches a scale that makes the jump from archival to hoarding work I'd simply need to reconsider my preferences.
Here's my take: The scaling issues of these tech giants won't ever reach my personal archive and any challenges with re-indexing, data analysis etc. should be completely approachable on SOTA hardware. Running anything that improves the searchability of my own archive can be run locally and in the timely intervals I prefer. To have this kinda quality approachable is a huge thing, and I can't wait until I can self-host some RAG enhanced vector search engine for a personal archive that grew overs years to take shape.
By this do you mean family photos and the like? I'd like to hear more about this. I'm building up a personal library like this too.
It is only a matter of time before the grid goes down, the country restricts the internet, or the service you rely on goes away.
(and many comments here didn't seem to read it)
I can also see it possible to 'self-host' things once you use a cloud where you can do 'confidential computing' stuff aka. the hosting provider does not have access to whatever it is you're running. That functionality is there on the major clouds now (EC2, Azure, GCP) all have the Intel/AMD/Arm TME/SEV/RME stuff implemented but finding it on a device that you can self-host in your little storage cupboard is impossible right now (EPYC 9004 seems to be the lowest available with that technology). At a minimum you want secure boot + attestation + memory encryption if you are not in control of the hardware space itself.
The only substantive argument I can see is that the technology is immature:
> Well...without exposing our services to the public internet and forcing our friends to signup for our weird app
Which, yeah, of course the tech is, there's only like a dozen people doing this. The exact hurdle named is hardly insurmountable: in the standards, OIDC overcomes this¹, or guest links. I don't want my family signing up for my weird app either.
One of the other big hurdles is that ISPs like to sell "Internet access", but only deliver half the deal. If you're not getting IPv6 connectivity in the year 2025, I'm sorry, that's a crippled product that your ISP was defunct and didn't properly inform you of when they sold it. (It's a lot easier to self-host on the v6 Internet. Some of my personal services are v6 only b/c of that, and that it works well enough in all but the most extreme or temporary locations.)
(¹but the half-baked OIDC implementations out there might require you to pre-register your app with them. That, rightly, might be a PITA.)
It comes off to me as the author not wanting to do the hard stuff of working towards their values. Just kind of defeatist and trying to make a splash but leaning on a pretty weak premise.
Unfair IMO. The author _did_ the hard work. And recognized that most other people, not similarly motivated, would not.
Most people do not give a rat's ass about the security of their data. They know their social media apps are tracking where they go and who they meet, and they'll say it's creepy if you ask them, but they don't actually care enough to lift a finger to do anything about it.
a) Just because humanity as a whole did hard things, doesn't mean that most humans did or were willing to. It's perfectly possible that all the hard things we did were accomplished by a handful of remarkable individuals, doing things that the majority never would have been willing to.
b) just because people in one age have been willing to do things, doesn't mean they are willing to do so in all ages. So it's not like the past necessarily proves anything here.
Self-hosting is like spending money putting a swimming pool in your backyard when you could walk to a public pool instead.
To me, the risk of backing things up in one building is too high, but the inconvenience of going even somewhere else in my own town regularly enough to rotate my backups is too high. But if my family members and I could easily back up each other's systems from our various states? Or my group of dorky college friends who are now all over the world could easily share with each other? We'd be all over it.
self-hosting has a lot of degrees. if you want your own TLD and peer with Tier1s, then it's astronomical, woo! But using dynDNS is also an option.
Especially if you compare to non-self-hosted services. You get a subdomain and that's it. (Or nothing, maybe some handle on Instagram.)
But beyond that self hosting is a hobby. It’s not nearly turnkey or cheap enough to justify unless you enjoy the process of self hosting itself.
There are other benefits outside the monetary equation of course like control of which the value is dependent on the self-hoster.
The problem is that people still believe that if they don't pay money, a service is free. But so many do not question why it is free. Hint: Not because Google just wants you to succeed and have a good life. And then, without any second thought, they literally upload their whole private digital life.
If you don't pay, usually, you're the product.
You yourself have hand-waved away an important part - security. It's not (just) about the friction of signup (though, I'll get to that later) - it's the fact that you'd be utterly insane, as an individual developer without a full-time security team, to expose a self-hosted application to the Internet.
And sure, you can give them a login to your VPN, but that doesn't negate the next part...
> and forcing our friends to signup for our weird app > in the standards, OIDC overcomes this
It's not the signup that's the hurdle. It's the fragmentation. Sure, if you implement OIDC, your friends can sign up to your photo app. And they can sign up to Sam's, and Joe's, and the app of the cute bakery on the street, and a couple others. What then? The whole value of a network is that the components are interconnected and can intercommunicate. If I have to upload my photos seventeen times to seventeen different partitioned applications for my various social groups to see them, I'm just as likely to not bother.
Fediverse-like ideas go some way towards addressing that, but they don't seem to be in any state of usability for anyone non-technical (I say that as someone who was using Mastodon as my only social media for the last couple of years)
SaaS/cloud providers propagate this FUD 24/7 and then Okta, which should be pinnacle of security gets hacked and has issues with disclosure.
Relax. Most companies has security team incapable of operating beyond checklist.
You don't have to. The article mentions Tailscale--the whole point of which is to not have any Internet-facing app exposed. Everything is done peer to peer between clients that are behind firewalls. There's nothing listening on an Internet exposed socket for random connections to come in.
That's one reason we're building https://github.com/purpleidea/mgmt/
And to quote Gabe Newell (founder and owner of Valve, the company that operates Steam):
> "We think there is a fundamental misconception about piracy. Piracy is almost always a service problem and not a pricing problem," he said. "If a pirate offers a product anywhere in the world, 24 x 7, purchasable from the convenience of your personal computer, and the legal provider says the product is region-locked, will come to your country 3 months after the US release, and can only be purchased at a brick and mortar store, then the pirate's service is more valuable."
> The proof is in the proverbial pudding. “Prior to entering the Russian market, we were told that Russia was a waste of time because everyone would pirate our products. Russia is now about to become [Steam’s] largest market in Europe,” Newell said.
from https://www.escapistmagazine.com/Valves-Gabe-Newell-Says-Pir...
---------
There are certainly cases where people will pirate to avoid paying but in the event that the option to pirate is not available, they will generally just go without instead. The only situations where piracy really becomes a matter of pricing is in the openly exploitative services like Academic Journals.
You are just pessimistic.
Afaik more or less the same thing happened with spotify and music pirating
Sure, you can own your server and have it at home. It must be nice to have enough space at home to dedicate some to a server room (servers are noisy btw). But many people live in small apartments in a city and so don't have that luxury.
You can own your server but rent some rack space from a data center to put it into. That would still be self-hosting.
You can rent a virtual or dedicated server from a hosting company, and even that would be self-hosting.
The author seems to not consider the fact that this is a spectrum but also, from a practical standpoint, mostly the same thing.
That's what I do. I use Linode/Akamai, which now has encrypted VPS instances.
Ideally, I'd have my own hardware but I don't want to deal with the maintenance and failure cases (house fire, etc). I think a VPS is a solid tradeoff.
Absolute bullocks.
For most people running a home server, a Raspberry Pi is plenty and is about the size of a deck of cards, maybe two decks if you want extra storage and use an external storage device.
If you need something beefier, you can probably just use an old laptop, or maybe a full second PC under your desk if you need more. You could easily fit a Threadripper or Xeon system with 128 GB of RAM, multiple drives, and a GPU or even two in a single ATX PC case.
If you need a full server rack, you're an extreme outlier beyond even 99% of homelab creators.
That said, the discussion seems stuck in a false binary between the control of self-hosting and the convenience of corporate services, but I think what the market wants is a third way that provides both control and convenience.
And to be honest, public libraries already do this, y'all. GO GET A LIBRARY CARD. You can stream from Kanopy at home.
Either you own and control something, or you do not, there's no third option. A best, you can outsource your stuff piecewise: run your own software on a cloud VM, or bring your own furniture into a rented apartment, or give a valet the keys to the car you own for parking, etc. But there's always some relinquishing of control in exchange to some other aspect of efficiency / comfort.
It's also easy to mistake what most people want for what everyone wants, and miss an important market.
I think there's a full spectrum you're missing. You can own something with other people, and your level of control can be continuous, not discrete & binary. For example, my public library is funded by my local government, which I can influence with lobbying and voting. I can join the board of the library, and I can just go and talk to the librarians in charge to influence their decisions.
In an individualist consumerist mindset things are pretty stark : full self-hosting or full submission. If you reject that mindset there are many more options.
Let me remind you of the open source credo about free as in freedom not free beer. You are right that there may be exchanges or compromises at play, but it was a bit shocking to me when talking about what is essentially the digital commons that no one mentioned a library, which exists.
I'm also saying from a practical perspective if you want to stream movies without giving money to big tech, you can literally do that tonight with a library card. The infrastructure already exists.
Nominally, yes. In terms of that meaning anything, no. The benefit of ownership is not exclusivity, but control. If the library doesn't have a book (or other piece of media, of course), I have no power to influence them to get it despite that theoretical ownership. If the librarian decides a book is offensive and removes it from the collection, I have no power to influence them to keep it. I have to live with someone else's decisions about what the library does and does not contain, just like with a commercial service. So my nominal ownership really means nothing at all.
But purely outcome wise, many people want the benefits of hosting their own servers
If I were to run my own version of Google Photos and the like, I'd probably go with the hybrid option:
Run all the software I'd run if I was self-hosting, but in the cloud, possibly with a backup in a second cloud. ie, put my photos in Backblaze B2, with second copies in S3 or something.
Personally, half the reason I use Google Photos is so that if my house burns down, I don't lose my pictures. A self-hosted server running under my desk doesn't carry that guarantee. Backups are off-site for a reason.
Though maybe self-hosted at home with a single cloud backup would be good enough.
I was skeptical about this scenario until one day Gmail lost 1 year worth of my emails. It's just gone. All other emails are there, but not this particular year. And there is no person who you can call to talk about that.
The things the author set up are technologically mature enough that, as long as you have the media, or as long as you can get your friends to use it, the self-hosted versions are largely better than the commercial ones. The last decade or so of innovation has really been about figuring out how to monetize these technologies, at the expense of UX.
This is in contrast to LLMs, where the commercial ones kind of wipe the floor with the self-hosted options.
On the other hand, LLMs essentially give average people superpowers for self-hosting mature technologies. My wife used Claude Code to vibe-code an educational game for our five-year-old, tailored to his preferences and the skills he needs to work on (she's a UX designer and now, a couple weeks in, reads enough Javascript to understand when Claude is doing something stupid).
If we want to buy a computer to use a server, write, and host a bespoke family to-do-list / photo store / knowledge base / calendar that syncs my wife's Google Calendar with my .org files ... we are so much more able to do that than we were even two years ago.
I think the biggest pain point is that Microsoft, Amazon, Apple and Google all wants services revenue. And they will go out of their way to force everything on their platform to become subscription based and you dont own anything.
Name a more iconic duo
https://en.wikipedia.org/wiki/Censorship_of_images_in_the_So...
I'm tired of hearing the Norwegian government talk about AI and modernization. Before we chase the next big trend, we need to solve fundamental problems. We should have one public, centralized provider for digital identity and authentication. We also need a single, secure messaging service for healthcare personnel and residents.
This same principle of focusing on the basics should apply to other services in the domain of selfhosters: secure data storage. Instead of building a complex, all-in-one platform, a community project could offer just a "digital locker" for files.
Users would connect to this storage via open protocols (like WebDAV), allowing it to work with many different apps. This gives users the freedom to choose their own tools for photos, documents, and media. This approach has three main benefits: * Lower Cost: It is cheaper to manage only file servers instead of a full software suite. * Simpler Maintenance: The limited scope makes the service easier to secure and sustain. * Predictability: The service is stable for users, and the workload is predictable for maintainers. It treats data storage as a public utility—providing the essential infrastructure and letting people build on top of it.
And if a community can’t get this basic and manageable thing up and running, a thing that has immediate and obvious utility, then maybe it’s unrealistic to expect more complex community or public utility-like services.
There are two "futures" to disambiguate here. The future for for-profit and institutional entities, which is not self-hosted. And the future for human persons, which is. The former will probably be HTTP/3 (quic over UDP) exclusively with CA TLS required while the future for humans remains on HTTP+HTTPS HTTP/1.1.
I won't be too many more years before the corporate future completely divorces itself from the actual web and goes full HTTP-IS-JUST-A-TRANSPORT-FOR-JS-APPS and becomes unable to even visit normal websites. For "security" reasons, of course.
that's a huge leap! i think most of us gloss over it, but the rest of the article is predicated on that leap.
the tv you're streaming video to probably runs Android by now. it has a stable internet connection, CPU, RAM, and probably a couple USB ports. why not install the Jellyfin server software on it, attach a USB hard drive, and let it be the machine that hosts all your media? why, actually, do you need to go out of your way to buy a completely new machine for this?
similar argument applies to Immich. you're wanting to co-edit an album among several contacts. you're probably all uploading your photos from a phone. why not just have one of your always-on phones host that album? i shouldn't expect the drain on your battery to serve an album to a few friends is that much more than it took to take those photos in the first place.
to a certain degree, you're "self-hosting" things on a physical server because that's the only platform on which we all still have the ability to run arbitrary workloads on. solve that problem and everything becomes a _lot_ simpler.
I run my own email, DNS and the rest. Yes: email. I run several domains including my own company and my own vanity domain and several more for friends. I have been doing it for decades.
Goog, MS and co do follow standards and if you do too, they will be largely merciful if you keep your nose clean. I have even managed to run an email system from my home connection as a test IPv4 and 6. I'm UK based. It does seem that IP denylists do seem to be a bit brutal in the USofA, so that might explain the downer meme on self hosting email.
The future is and always will be self hosted if you give a shit.
I can read the books and acquire the knowledge from my kindle. If Amazon removes it, I can just pirate it?
I get the theoretical argument but as a very pragmatic person it just seems like tilting against the windmill.
No you never owned them, only a really permissive (for amazon) license that permitted access to them.
It's "skin in the game." Right now, cloud services fail horribly at much of what they promise or merely imply; safety, security, long term availability, etc.
And so, to make them not fail at this, they must be punished when they fail at this. The other side of this coin is probably "you have to pay them," but that's not so bad either.
Hardware: The Cloud
Applications, Data, Control: Self
I make use of google drive, apps and google photos and therefore suffer the knowledge that my stuff is probably being used to train AI (I’m personally comfortable with this) - but then I have local backups and sync for everything. InsycHQ gives me a locally (and NAS) backed up version of all my documents, I also backup all photos shot by my wife and I to my Synology, and then daily backup in 2x places that I own from there. My films and media are locally stored, and I buy a fair bit from BandCamp and I use Plex to serve but I also have Spotify for the convenience factor. …and so on
In other words - it’s a sort of multi-tiered approach. I’m not subject to the whims of cloud providers because if they change their pricing or terms radically I’ve still got all my stuff locally; ditto backups on infrastructure that I own and control in-house.
It’s always seemed to me like this is a pretty good setup, combining the utility of cloud with the reassurance of self hosting.
To me it seems very reasonable to have these things hosted in central repositories, with large corporate stewards of the creative works, which I can access on any device for a monthly fee. The creators and owners of the works could then upgrade them over time, e.g. to newer formats or to fix errors.
But in practice, this isn't how it turns out:
* Tons of Kindle books have minor typos or OCR errors. These will never get fixed. If I had local copies, I could fix them... But nobody else would benefit from my fixes.
* Disney+ has misconfigured four episodes in Daredevil season 1 to show forced English subtitles for when English is on the screen---e.g. when there's an exit sign, there will be forced subtitles for "EXIT". I can only imagine if I submit some sort of ticket for this it'll just disappear into the ether.
* The Marvel Unlimited comic app, to their credit, is doing a great job digitizing their giant backlog. But they haven't paid a lot of attention to the flow of series, so e.g. "annual" issues are not slotted into the series they're part of. Back in the days when I collected cbz files, I painstakingly placed them all in sequence.
* Spotify's music metadata is pretty bad, and its collection is missing things like game soundtracks. (Although it has plenty of indy remixes of game soundtracks, clogging up the search results.)
* I worry that the "original quality" of all this media is getting lost over time. Certainly watching shows on Netflix is going to be lower quality on an absolute scale than Blu-ray rips, right? Similarly, comics are transmitted to my device as JPGs---I hope someone has the original, uncompressed pages stored somewhere.
If I had local copies of all this media, I could organize it beautifully, fix typos, set up perfect metadata/subtitles/etc. I used to do that, with pirated media, back in college. But it doesn't feel like a great use of time these days, mainly because nobody else will benefit from my obsessive work.
I wish the custodians of this media would care more about it, or put in place systems for community contributions to improve it. But the incentives are not there in terms of $$$, sadly.
It is definitely true that piracy is primarily a service problem, not an ethical one.
Even if I had to (and this is what I actually do actively) push the photos I want to share to Google Photos, ALL the benefits remain. I see it, it's a slight inconvenience of having to do like 2 more taps and wait for the upload, but that's it. You get so much for this small inconvenience. You own your data, your infrastructure, you're not locked in, and your data is private. But having to share the photos via another app is the dealbreaker?
Instead we should have ideas of AUTH, AUTH fraud (distributing content with wrong authorship), the right to generate revenue by distribution.
Restricting distribution of something that is essentially free to distribute feels wrong. If you tried to explain to aliens that you had solved "food" with a replicator but The Gov actually banned this and people had to starve ... you would likely be met with some questions.
jqpabc123•19h ago
What we need first is incentive for smart, dedicated, part-time sys-admins to devote time and effort to community hosting.
Without this, it will work --- in the same way that open source works --- without any guarantees or commitments whatsoever.
In other words, you're on your own for the most part. So it really is just a variation on self hosting. By the way, we've already been there, seen that and done that --- it was called "co-location".
When you need something more with service and reliability, well --- you're right back to paying corporate overlords.
But thanks for the round trip thought experiment.
drew_lytle•19h ago
HPsquared•18h ago
__MatrixMan__•18h ago
It's easy to trust a corporate overlord with your pictures or your email, because the immediate damage doable by somebody who has compromised those things is relatively low. Privacy is important I guess, but not when compared to things like whether your car or your insulin pump does what it needs to to keep you alive.
Eventually, the bad guys will be sophisticated enough, and the tech will be integrated enough, that it's no longer safe to trust economic incentives alone. You're going to want your sysadmin to share your interests (in a more specific way than you get from they-also-like-money).
dylnuge•18h ago
Really, it's easy to get sysadmin types interested in this; the problem is that most people aren't sysadmins and don't know any. If you really wanted a business model out of this, it'd probably be a managed service that lets non-tech-savvy users spin up their own versions of this without learning the details.
> Without this, it will work --- in the same way that open source works --- without any guarantees or commitments whatsoever.
There are plenty of successful economic models around open source, and plenty of open source software is used in high-reliability contexts. What comparison are you trying to make?
cmilton•18h ago
Co-location is still readily available. Which service and reliability improvements are you looking for that competent sys admins couldn't provide with multiple co-lo's? Not everyone made the cloud jump.
jqpabc123•16h ago
I moved to AWS and haven't had that problem since.
sgarland•18h ago
I’d do it for free. I’ve long been frustrated that I have more reliable infrastructure in my homelab than most companies I’ve worked for, and that none of them have any interest in shifting out of the cloud.
I don’t see a market for it, though. Most people are generally happy with Google, Apple, etc. to host their stuff, and I get it - it’s quite reliable, integrates with the rest of their respective products nicely, and Just Works. Add to that the economies of scale, and it’s a non-starter unless you find a niche group of people.
Google One is $99/year for 2 TB of storage. For me to have confidence in uptime to offer public storage, I’d need at least 4U of colo rack space, and ideally 6U (2x 2U for HDD servers, 2x 1U for hosting applications in HA-ish). That would cost a few hundred USD/month, not to mention an initial outlay of tens of thousands of dollars for servers and drives (mostly the drives… high capacity enterprise-rated HDDs aren’t cheap). And that’s only for one site - ideally, of course, there are at least two, or at the very least, off-site backup like rsync.net.
jqpabc123•17h ago
And if you get hit by a car? Or worse --- maybe you get married and have kids<g>?
One big reason people *buy* service is sustainability/longevity/redundancy.
There are no absolute guarantees but I think most commercial endeavors nowadays would bet on AWS/Google/MS/Apple over "Hosting by Joe and Friends".
esseph•16h ago
jqpabc123•15h ago
Personally, I'm betting on those who are highly incentivized and have the resources and structure needed to sustain reliable service.
ryandrake•12h ago
jqpabc123•5h ago
udev4096•2h ago
sgarland•10h ago
Also, FWIW I am married and have kids. Hasn’t stopped me from homelabbing.
esseph•16h ago
fragmede•16h ago
Not all corporate overloads are equal. Or rather, if you and your buddies get together and pay the $350+fees to legalzoom to start a corporation, you too, can be a corporate overload. There's still miles to go before you're Facebook, but congratulations, you're now... still the same person you were before you clicked that button on legalzoom's webpage and spent $500 or whatever.
Where is the problem of people turning into corporate overloads for you? Is it at 10 employees? 100? 1,000? 10,000? If we're too stupid to differentiate specific corporations because their legal structure means they're all exactly the same, then yeah, I guess there's no hope and we're all doomed.