In the age of LLM agents, still going for Docker/Podmam on anything besides local development box seems like a weird choice.
Docker has poor tooling for network level security between the containers, has issues with different runtimes per container, etc.
It is just a bit primitive if you want to expose multiple services to the internet on the same server. One of those apps will get compromised and then all the others will follow.
If you want a decent self hosted server, ask your frontier LLM agent of choice to configure kubernetes (on something like k3s) with mandatory userns mapping so nothing runs as root on the host, default deny firewall so inter-container communication is as locked down as possible, and if your router supports, set up VLANs so none of the containers can access your other devices on the LAN. Use something like backrest to handle backups, alertmanager and Grafana for monitoring, Keel for auto updates. Also consider separating ingress for public and internal services and use Tailscale with split DNS to acces the internal entrypoint. Set up Crowdsec as WAF and subscribe to their free blocklists to filter out bots. Ask it to set this all up using Ansible, so it can be maintained.
While this would be extremely time consuming to set up and maintain by hand, an agent can do and test it in a few hours.
(author here) I'm checking with my team to confirm these came from us, but if so I apologize for this - posts like these should come from a real account with proper disclosure, but also these sorts of requests feel premature for a project like ours that doesn't yet have a sizeable user base.
You are also characterizing another project that was open-sourced (sandstorm) and is under active development as "long abandoned". Open source is difficult and requires a fair amount of cheerleading to engage users and make them part of the community. As a result it's hard to see your project in a positive light.
Looking at the issues, do I understand it correctly that you have some way to install directly from repositories if they contain a `cloudinabottle.toml` file?
If yes, regardless of who made the issues, that seems like terrible architecture/design there...
I checked, it's an Imbue employee (or someone impersonating one, which seems a bit far fetched). I could could cite my sources but it feels like doxxing / drawing too much attention to the mistake of an individual. My read is that it was a separate github account that they let an agent control.
The person doing this doesn't understand how pull requests work. The issues all say "I have a version in my fork". Open a pull request then! That's the whole point of forks on GitHub. Creating an issue first is pointless spamming akin to the classic "asking to ask". Just seems completely out of touch with open source to me.
It's not spam if it's the first time I heard of it and it's useful. It shows they care about their project. This makes me want to try it out even more.
Meta and Google can spam their unsolicited ads to a billion users but regular people aren't allowed to promote their own projects on niche forums? This is bs. You're the problem here. Self-promoting your harmful values. You're the spammer.
I'm surprised there's no such thing as a self-hosted deployment spec. An opinionated docker-compose file (or similar) with well defined inputs, outputs, and requirements that works for the average use-case.
I think the time for this space is here. The appetite to depart from subscriptions and loaning your personal data to ad/ai companies is stronger than ever.
Currently, a lot of this space is docker compose based and simply inaccessible to so many who may otherwise be interested. Whether or not it's this project or another, I hope something gathers enough steam soon to truly break the barriers to entry.
General public is weird because everyone, yet no one, is the general public.
Fwiw I've started to see a shift in many non-tech circles I'm a part of. Especially when you take the time to explain the problems. Many do care, they just aren't made aware of the issues.
The general public will do whatever their friends do. If there was an iPhone charger sized box they could plug in anywhere and have their own personal cloud that can interconnect with other personal clouds, and a celebrity said it was cool to have one, the general public will do it without giving it a first thought.
Precisely. My uncle has been using the builtin Samsung web browser for years on his phone, just because it said "Internet" in the name, and "Chrome" just wasn't a familiar name to him. Most people are like this.
Digital rights are indeed important, but only a small portion of the entire concept of human rights. I can imagine general public places much more emphasis on wealth (re-)distribution policies.
A lumberjack may also argue GP don't care about environment laws if they had timbernews.
True, most people get the full ad driven internet experience and it drives me nuts. Everytime I have to help other people with computer stuff I almost get an aneurysm when opening their browser...
How does being docker compose based make it inaccessible?
Self hosting implies having hardware and software knowledge to set it up. I agree, having a meta package that would set the system up as systemd would be an even lower barrier. But then again, there are podman commands that do this for you today anyway.
Self hosting could also imply buying a device that just does this with an additional management interface for updates. This could be build around a nuc/nas with above package. It still will require some knowledge to keep it maintained.
Then, you could offer maintaining them as a service again. Maybe there is value in that.
Update: apparently there is https://freedombox.org/ which is actually an appliance for this. Although it is more privacy focused and I don’t see a productivity or photo suite on it.
> How does being docker compose based make it inaccessible?
Because it requires people to also have knowledge about the ins and outs of docker and docker compose. If we are talking about web applications it is yet another layer that has been added over the years. It used to be that you could fairly easily host most things on a old fashioned lamp host (I am talking decades ago) and all you had to know was basic file transferring.
Of course, this was on shared webhosts in a time when VPSes were not really a thing yet or affordable. But, even on a VPS setting up a LAMP stack is relatively straightforward.
Once you add modern dockerized application to that mix you are now still looking at some sort of ingress to do the reverse proxy bit. So while it likely will not be Apache or nginx there will be some sort of layer there in addition to now having to setup docker properly (rootless and all that) and then also making sure your docker compose setup is in order.
It is not difficult *once you know how to* but it is another layer of knowledge and experience people need to acquire. That does make it less accessible for novices.
At the same time buying compute is more expensive than ever, and the gap between homemade projects and what's available online in term of speed, functionality and LLM capacity is bigger than ever.
Building the equivalent of Google Photos locally wasn't that hard 10 years ago, given enough tolerance for speed and indexing lag issues. Trying to do the same today will cost two arms and leg and require a lot more overall knowledge of what to compose to get the same result.
The dust will settle and we should go back to a nearer gap in I hope a few years, until the. it might be a glacier age for self-hosting.
> Building the equivalent of Google Photos locally wasn't that hard 10 years ago, given enough tolerance for speed and indexing lag issues.
You could argue it’s much easier on the software side, because self hosting Immich is much easier thanks for AI assistance with setting it up. If you can live without the built in facial recognition feature you don’t need particularly beefy hardware.
I find myself taking on far more ambitious self hosted projects thanks to Claude. Extending home assistant in various ways, deploying a “self hosted Spotify”, making sure everything is fully accessible via Tailscale out of the range of the wifi and so on. Of course the general population isn’t really aware or interested in most of this.
for self hosted and personal use, compute lasts a good while though. and in the time you would've been waiting to time the compute market, you get to benefit by being less a part of the AI data dragnet.
The biggest difficulty is probably that from my observation, there's a sharp cutoff line somewhere between Gen Y and Z where anyone born before that (and don't work in tech) genuinely have no idea how to navigate their digital life and gets preyed on by those subscriptions. And these people are still a super large portion of the population paying for those subscriptions.
Only Gen Z and later are starting to care and willing to spend the time and effort to self host. All the Gen X and Y people that I know of, including many that are very intelligent and accomplished in their field, literally have no idea how to operate a computer outside of watching youtube, checking emails, and opening Word documents. Ads? They hate it but have no idea how to disable them. iPhone running out of storage and iCloud says I can solve that for $2.99/mo? They would be like IDK what that is but sure I'll pay that $2.99 if that means I can record more videos of my kids.
I really hate this situation, and I try my best to help out the folks around me who I see are getting ripped off (either they don't know or they know but think that's what it is). But still, it's really hard to communicate or teach those people when the skill of operating a computer has become intuition and human nature to me. I really can't put these things into words on the fly that would help them understand :(
> Only Gen Z and later are starting to care and willing to spend the time and effort to self host.
I can't quite confirm that. My younger siblings' classmates (still high school students, born >=2010, non tech average people) don't give a shit on these topics and behave identical to the Gen X and Y described.
Looking at my younger sister, I'd risk saying it's the same in all generations. I'll even say more: I think earlier generations had a general thought of "computer consists of separate parts; you can buy them separately and replace if needed / want better performance", even the less technical ones. But now everything is a black box (therefore e.g. why not overspend and always buy a black box from Apple if that's fashionable now).
But maybe that was kind of related to times and the place – where I live people needed to be a bit more savvy, so thought like "why you buy an iPhone if Xiaomi can do the same 4x cheaper; if you chose the right model you may even have a good camera too" was popular even in non-tech circles.
But yeah, now that hardware is a bit more of a commodity and such tradeoff might give you a bit more adware with a bit less convenience, maybe it's different. But the next generation's understanding of computer/phone as something else than a magical black box is gone.
> Only Gen Z and later are starting to care and willing to spend the time and effort to self host.
Yeah, you’ll need a lot of supporting evidence for this claim. From my personal and professional experience, this doesn’t track. And given how online the younger generations are, it tracks even less. I think there might be a bias towards tech bubbles in that statement.
Or, alternatively, I was a lot more willing to put in the time to self host before I had more money and less time. If it's truly important it's not in public. If it's public, I don't have time to keep up with security patches if I'm not being paid.
> 53 is required because instance runs its own authoritative DNS server.
Are you planning to offer alternative solutions for this, e.g. pointing a wildcard A record at the server running the instance? I'd rather not run a publicly accessible DNS server.
"We (Imbue, the company I work for) also offer a managed version, which I think is really important to making this widely accessible - and it gives us a straightforward business model to support the project."
Looking at your hosted page, I see no reference to backups. I see in your docs reference to backup, but you should definitely offer some sort of turnkey, appropriately-marked-up backup solution. Configuration difficulty is what kills self-hosting on the front end, updates make it hard over time, but when it all dies and then I have no backup is when I give up.
I can't find any reference to disk size limits anywhere. Can I host Immich on this? Can I host my media server? What will it cost? Surely not $10/month for my media collection. I'm not a hoarder and mostly have stuff ripped from my personally-owned media, which limits the size, and you're still not hosting that for $10/month. I see you have Jellyfin in the set of apps but I don't know what it will cost. The CPU & RAM limits I understand from the deploy page.
Oh, and I'm specifically referring to your hosted service here.
(author here) Thanks for flagging, I should make this more clear. Our managed instances have pretty small disks (~50GB), but there's a way to link a S3 bucket to your instance to use for bulk data storage. Apps that store bulk data (like immich/jellyfin) should automatically put it in the "archive data tier" which can be configured to go into this S3 bucket instead of local storage. We're planning to make this (and backups) an auto-configured part of our managed instances - billed separately (so you pay for just the additional storage you use; cloudflare R2 (S3 compatible) at-cost is like $15/TB/mo so it should be in that neighborhood).
There seems to be a lot of variance between stated and revealed preferences around self-hosting.
It often seems like presenting the image of being autonomous to our peers seems more important than actually achieving it.
It takes a lot of energy to replicate what AWS and friends have done once we factor in concerns like the passage of time and entropy. I've never been able to keep a media/NAS appliance alive for much longer than 3-4 years. Inevitably, there is some kind of catastrophic event and I have to start all over with new vendors, etc. Even without any data loss this gets old. I promise you get tired of this after a while. It might seem like that would never happen and then one day it does.
You can get so much more done if you can actually afford to use the cloud. I have a really hard time believing that many people are genuinely winning on total cost of ownership with self hosting.
How are you backing up 100TB of data? Have you factored the consequences of catastrophic data loss into the economics? How much value would you lose if all your data was gone?
There's no way you are consuming this much content actively. You could store all of it in S3 deep archive for $200/m (100TB) and maintain a tiny fraction in the standard resource pool.
This is a great project! Yes, the personal cloud is coming and it's going to be for everyone.
I'm working on a project with similar goals but some different design decisions - instead of the typical containers I'm going for thin-client apps that hit a common data layer built on a fully serverless architecture. Not ready for production quite yet, but for those interested: https://starkeep.app/
This pitch is somewhat interesting, but, much like Sandstorm, browsing the library of pre-built apps kind of leaves me cold. It's too much like browsing a rather tiny mobile app store. There are maybe three apps that I might be interested in.
Since we have AI now, I think exe.dev's pitch is better: they provide Linux VMs, a web proxy, and integrations. You can ask the AI to build whatever web app you like. I'd love to have exe.dev, but running on my Mac Mini.
If this can largely automate the backup and update side of things, then it seems like a great middle ground for people like me:
I’m trying to self-host as much of my stuff as possible, but don’t have time to tinker like I used to. Hence updates, backups (etc) fall by the wayside in favour of “set and forget”.
selfprivacy appears to be limited to a handful of services. e.g. It is not clear to me if I can host postgresql with it. i.e. It is not a general purpose container hosting product.
Happens. Had an idea (simple and stupid one), built it (with LLMs) went to register the domain. Found out somebody had the same idea and already put it there :)
I'm trying to understand the difference between this project and FreedomBox. I mean it's great to have options, but have any of them looked over that way? Heck, FreedomBox is even built into Debian now.
Now the key question, can the hardware be made accessible to everyone. Its definitely an improvement to make self hosting software easier but the process of setting even a raspberry pi for an average consumer interested in self hosting is still a barrier. They probably will be able to figure it out via youtube and some reddit posts but there should be an easier process or way
(author here) I am really excited to make pre-configured Raspberry Pis that people can just plug in at home, and have it work out of the box. It'll take a bit more work to make it seamless but we're definitely aiming in that direction. It's a shame how expensive Pis have gotten recently though with the memory shortage.
That is cool to hear. I know lots of folks who would be genuinely interested in this who don't have the technological chops to set up a self hosted system. One angle i have heard is degoogling and having more privacy in their data. I am really genuinely interested in this venture. I had the same idea about a year ago but was not in a place in life to act on it. I think even with the PIs being expensive because of the people will still buy them. Especially if they are seamless or near seamless.
Making self-hosting really accessible isn't that easy, even if it's easy to setup the server machine itself.
One needs to make their server machine accessible to everyone, which isn't generally possible due to how networks work, one needs a white IP address and making sure no incoming connections are rejected by the internet provider.
Also one needs a domain. Obtaining it requires some bureaucratic burden, costs money and pricing isn't always transparent.
These barriers are high enough, so that common people can't create/host their own sites. Only skilled enthusiasts or professionals can afford having their own site.
People choose not to because they see no need to, but if they put some time, effort and not a lot of money then those barriers are not so high.
I took the trouble to get a driving license because I saw value in that. I also took the trouble to learn to set up my own stuff on a VPS because I saw value in that too.
What's a "white IP address"? It is, sadly, more difficult than it should be, though. When everyone got online in the 90s they actually became bottom feeders and not true internet citizens. Not much has changed in 30 years. IPv6 is necessary, but the world also sucks now so people need to know what they're doing with firewalls etc. Domains are not too bad now, there are some decent registrars now like Porkbun.
"white IP address" is an address that your internet provider gives you and which is accessible from the whole internet, not just some address local for the provider's network and hidden behind NAT. Since IPv4 addresses are scarce, most providers don't allocate them directly to their customers.
An unique IPv6 may be theoretically assigned to each end device and thus solve the accessibility problem for locally-hosted servers, but I doubt internet providers do this in practice and I am afraid that many providers don't support IPv6 at all.
I really like the idea, and I would like to see Tailscale evolve Aperture in this direction. Tailscale is already an easy way for anyone to connect their devices, which is essential for any sort of "personal cloud", and Aperture, while initially something of an AI gateway, is also facilitating connections to other services and apps.
If you're building a personal cloud, you also want your AI connected to it, but probably with some degree of management. Aperture already provides a fair amount in this area, and probably with a connector for Cloud in a Bottle, could give your AI safe access.
I did a few pitches related to this at Nasscom Startup warehouse and IITM RTBI. Though this was welcomed by the panel members but it never took off due to lack of capital for such a startup during those times in India. I had to abandon and focus on my day job.
This happens with making many small to medium level resource hungary apps, for personal usage, run on one VPS which is not a heady duty machine costing heavy duty money. I have tried this on similar machines 2GB RAM and 1-2 VCPU and after 2-3 apps it starts swaying like a drunkard in slow motion even in that SSH session. But at least this one seems to be better than something like PikaPods, in combo pricing regard (I am not saying one is generally better over the other), where the pricing is rather per app.
Interesting project! I've found self hosting a lot more approachable with the advent of AI to help with inevitable errors. I run every new app in its own service account and as a rootless podman quadlet. There are three main difficulties that I come up against with every app. First is backups, or more accurately restoring data. My understanding is that just taking a snapshot doesn't always give you a working restore point because an app could be mid-database migration or something like that. Second is updates, and third is accessibility outside the network. I am slowly working on my own solutions, interested to know how you tackle those though?
>At the core, it's just an Ubuntu machine with a web server that hosts a dashboard and routes HTTP(s) requests to containerized apps.
So no fail over/redundancy? I think at minimum it should be two machines, so one can break. That is one of the important features of cloud applications as far as I'm concerned.
You don't have to deal with the problem of a single computer breaking.
This also adds a lot of moving parts and failure modes. For self-hosting I’d rather have downtime on a simple system that’s easy to fix rather than a complex setup that stays up but needs more maintenance when some component fails.
Like the real ones, digital clouds can take many shapes.
Sometimes it means "rented virtual resources", sometimes it means "stack of intertwined services", sometimes it means "SaaS", sometimes it means "alternatives to SaaS"...
This is neat! I've been building something very similar for personal use. I may lift a little bit of the app manifest part.
I'm a little surprised that there's just one backup app, instead of each app coming with its own backup sidecar. To backup Immich, for instance, requires coordinating between its four different programs. I guess I'll read more about how that backup app works too ha
Has anyone else found NextCloud to be slow and unreliable? I've been using it for the last three years (switching over from OwnCloud before that), and it's never missed a beat. Straightforward to set up, decent linux desktop clients, has mobile support that's also been rock-solid, and basically does what it says on the tin. Doesn't feel remotely enterprise targeted to me either -- it feels very much targeted towards home users.
Thanks for bringing this up, I was going to ask the same.
I haven't heard any complaints about NextCloud, they've added some enterprise features, but all in a positive way, from what I can tell, not detracting from home and personal use.
We need gmail, docs and sheets replacements. With AI, maybe people can develop solid alternatives very quickly. And Linux phones.
Trillion dollar companies try to extract more and more from us, the only response is for us to develop alternative solutions and continuously chip away at their strengths- what made them monopolies in the first place.
I had rather good experience using agents to setup and deploy own cloud, Jellyfin and co. I experimented also getting a VPS, just giving the LLM the ssh keys and having at it. It’s pretty good.
If the AI stays afloat, then I think the future lies somewhere there. Not too sure about the security side of the things …
> just giving the LLM the ssh keys and having at it. It’s pretty good.
I'm mostly fine letting AI write code because its easily auditable and you can follow changes through the Git history.
Whereas, it you let the AI go free on a server, you mostly have no idea what it did and where, so it's quite hard to audit and even harder to try and fix things.
So, unless relying on a declarative paradigm like NixOS does, I'm not ready to give AI the keys to my castles
This is not me saying security doesn't matter (because it does) but having everything not publicly accessible and only accessible over Tailscale or Wireguard makes me slightly happier.
Running the authoritative dns for the zone seems elegant, although wouldn’t that imply you absolutely can’t use cloudflare or similar services to avoid ddos/bots?
I recently bought a motherboard for my first self made pc. Six to eight sata slots is already enormous and the fact that you pay once and use your hardware freely opened my eyes. I was paying a subscription for cloud 1 tb for a trivial service. I am now hoarding my torrents worry free and plan to convert it to NAS. However i do still use my onedrive plan shared with three friends for 10 dollars for sharing large files till i set up my own nas with a good permanent net connection. The other issue is that I’m renting currently and if I’m changing countries I’ll have to rely on the onedrive again
accessible self hosting is already here, its called 'claude' or 'codex' you basically start it and tell it what you need, dont even have to know any software names.
I think what this needs to be truly accessible by general public is a mobile app (most people don’t have computers) and to be able to deploy it (including setting up domains) from there. Payments would definitely be a problem because of the various app stores’ rules.
Yeah, it was called Openhost for a while before changing to Cloud in a Bottle. Either filename works.
It proxies requests from the user to installed apps, and adds a header 'is-owner'. Apps have to be built to respect that header - some apps do this by running their own little proxy/sidecar that will detect that header and inject saved credentials.
An aside: Fantastico was such a wonderful piece of software for me to learn and toy and tinker with web stuff. I was so impressed by the simplicity of it. (I phased out of the shared scene a almost two decades ago; I guess everyone moved to Softaculous, https://softaculous.com/fantastico/)
How is this different from say Docker? I procrastinated getting acquainted with the latter for a long time because I thought it was more for industrial-scale infrastructure, but when I finally wound up using it I was surprised at how easy and low-maintenance it was.
I picked it a few years ago for much the same reason as anyone else (sick of this shit), and it's main attraction to me was it being a turnkey solution that focus security so i wasn't as likely to completely fuck everything important up.
I think a "simple" installation package would be very attractive to most people, and while cosmos has a lot of growing to do in order to be that, I'm still pretty happy with it.
I'm even more happy to see more in this space, because I'm just at the point of "do i stick with cosmos or do i start rolling it all myself" and god the leap is terrifying if you're thinking of exposing to the internet. I get it, but at the same time it feels like you're always one missed config away from a potential nightmare, and that's after the "welp guess this is my second job" problem.
The unified auth + inter-app permissions are the real differentiator vs Coolify/CapRover-style container hosts. Data tiers (local DBs vs S3/R2 archive) also look like the practical way to make Immich/Jellyfin viable without giant managed disks.
Meh. Maybe I'm a grinch, but I don't see this as easier. I have a little NUC running Ubuntu. Putting services on it is easy: PiHole, OwnCloud, Wireguard, etc.
a similar project which i think is a lot better than this (in fact it's so good i'm jealous of its genius) is https://exe.dev. It feels like something that hasn't really hit the mainstream yet but I suspect everyone will be talking about soon.
How does it deal with storage? Storage is always the weakest link in self hosting. Having a bunch of stateless containers is ultimately just a fancy docker compose file. Still valuable, but you need a proper story for storage. Ideally all storage would be on individual volumes per app and all backed up etc.
> coolify: hosts containerized apps, but every app is an island with its own login and little integration with the host or other apps.
That’s not an issue tho, I want each application to be a separate login, so I can have one pointed to a different subdomain or domain not all in one login and location.
I remember back when I first got into self-hosting (over a decade ago), I had a Netgear READYNAS. People made a small amount of plugins you could install. But the CPU was ARM before that was cool (I think mine was actually SPARC!) so it was a lot of effort to get things to compile, so plugins were usually quite outdated.
Then I decided to painfully break out of vendor lock-in, built my own NAS with FreeNAS which was all the rage (still attached to a vendor, but at least I could control the hardware). This was before Docker really took off - the plugins were installed in FreeBSD jails. It worked for a while, but again, plugins could be woefully out of date and broken, upgrading the OS was a pain and would break all your plugins.
Through the 10+ years of this... I now follow two simple rules:
- separate your network storage and application hosting. Yes, it complicates things a little bit with NFS setups but it is a forcing function for a more resilient setup
- use the docker container provided by the maintainers or LinuxServer.io. Vendor maintained wrappers can be initially helpful but end up as a form of tech debt when they're abandoned or neutered - especially when they're freemium and the company starts looking to squeeze.
The quiet part of self-hosting is that it shifts the responsibility of security to the user. I'm sure cloudinabottle has more sensible defaults than provided docker containers, or builds in reverse proxies etc, but I doubt they're offering to take on liability for data loss or breaches.
The problem I see you'll face are two-fold:
1. Projects like this have the power users self-select away from it and those who need a lot of hand-holding to self-select into it. If I know how to configure and deploy a docker container already I have no incentive to contribute. ReadyNAS got around this by having folks charge for their plugins (that were just OSS projects packaged up for their OS) but then it ruins the appeal, they likely only got away with it because they were only a couple of bucks and you were already locked in to the ecosystem via the hardware.
2. If the hosted aspect actually makes you money, required for the project to be continued long term, capitalism will ensure that incumbent cloud providers take that away from you by offering ready-made instances at your cost-price, like they did with redis, elasticsearch etc.
Note that these are only issues if you're motivated by making a profit, or even a cost-neutral project. But if this is a charitable exercise backed by an entity that doesn't need the money, I wish you all the best - more things that let people dip their toes in are good! The cloud is all rent-seeking subscriptions and even with the recent supply shortages: fast, small, low-power compute is incredibly cheap these days, especially old business/mini machines.
IMO, one way to retain the power user crowd is to do the hard parts that they might have skipped themselves. For a couple examples:
- after selecting your observability apps, every _other_ app could automatically configure centralized logging, its own dashboard of metrics, etc
- the host OS could ask the user simply for the backup destination (S3 url creds, attached local disk, SMB share, etc) and every app is again automatically configured with backups
- the host could have a control panel to roll back apps between versions (maybe it's FS snapshots or something)
- everyone can use containers, but if the project went through the extra effort to make each app its own VM instead, that could be worth something
- if there was a companion Android/iOS app for management and (importantly) alerting?
---
That doesn't really solve the monetization issue you described, but I do think you can still get powerusers on board if you handle even harder problems than they'd want to take on themselves. Or...harder problems than they'd risk taking on themselves if their family uses what's being hosted, for example. Less tolerance for downtime
No offense, but this is not accessible to everyone. Seriously, take a look at the setup instructions and tell me with a straight face that a 4 year old could set this up.
How do I reach out to you? We're building something that you might want to be a part of. (The "Safebox" comes with all the software you need, and you can just clone its AMI on AWS etc.) Would love to have a conversation.
LLMs can be self hosted though. Using Claude to build a self hosting platform is a bit like telling people they should be vegan while you're eating a bacon sandwich.
> In the pre-cloud days, open source worked - authors distributed their software, and everyone "hosted" it for themselves (i.e. ran it on their own PC)
This is just rose tinted vision of the past. None of the software of the past had the capability, reliability or robustness of modern sass.
Yea. When people talk about how simple things were in the past, I think they oblivious to the fact we were all writing out X11 configs by hand and having them break in all sort of strange and interesting ways. Open source has always been very "I gave it to you for free, I'm not going to hold your hand".
I've written X configs by hand, but only to get a few extra pixels of overscan. I've never needed to do this. My experience has always been that things just work. The UI has never been great, in that I need more explanation than the built-in manuals provide – unlike, say, Windows 95, where you can learn everything you need to know by clicking around – but it's not hard to avoid breaking things, and it's not that hard to learn to do new stuff if you have a good book (or, lately, blog post) to consult.
In the age of LLM agents, still going for Docker/Podmam on anything besides local development box seems like a weird choice.
Docker has poor tooling for network level security between the containers, has issues with different runtimes per container, etc.
It is just a bit primitive if you want to expose multiple services to the internet on the same server. One of those apps will get compromised and then all the others will follow.
If you want a decent self hosted server, ask your frontier LLM agent of choice to configure kubernetes (on something like k3s) with mandatory userns mapping so nothing runs as root on the host, default deny firewall so inter-container communication is as locked down as possible, and if your router supports, set up VLANs so none of the containers can access your other devices on the LAN. Use something like backrest to handle backups, alertmanager and Grafana for monitoring, Keel for auto updates. Also consider separating ingress for public and internal services and use Tailscale with split DNS to acces the internal entrypoint. Set up Crowdsec as WAF and subscribe to their free blocklists to filter out bots. Ask it to set this all up using Ansible, so it can be maintained.
While this would be extremely time consuming to set up and maintain by hand, an agent can do and test it in a few hours.
Skip kubernetes and deploy to bare server
These guys have been spamming issues in repos trying to promote this project and no disclosure whatsoever that they are associated with the project.
https://github.com/search?q=%22cloudinabottle.toml%22&type=i...
It's evidence of bad abstraction if you need everyone else to make changes and couple themselves directly to you.
If for some reason your services can't speak an already common language, I think it's on you to add those shims into your own project.
(author here) I'm checking with my team to confirm these came from us, but if so I apologize for this - posts like these should come from a real account with proper disclosure, but also these sorts of requests feel premature for a project like ours that doesn't yet have a sizeable user base.
The user was called adl-collab, when it forked repos to add implementations it made them under a branch called "andrew".
The biggest contributor to the project in the last 3mo is andrewlaack... and we can also just take a guess that "adl" in adl-collab is AnDrewLaak.
5 replies →
You are also characterizing another project that was open-sourced (sandstorm) and is under active development as "long abandoned". Open source is difficult and requires a fair amount of cheerleading to engage users and make them part of the community. As a result it's hard to see your project in a positive light.
1 reply →
Looking at the issues, do I understand it correctly that you have some way to install directly from repositories if they contain a `cloudinabottle.toml` file?
If yes, regardless of who made the issues, that seems like terrible architecture/design there...
> posts like these should come from a real account with proper disclosure
This line makes me think you are running a spam campaign and andrew just isn't part of it
1 reply →
Why you apologise for something which is perfectly legal. Give an inch, they'll take a mile. F the authoritarians and monopolists.
6 replies →
Hardly seems worth the pitchforks.
shouldn't we hold back with allegations while there's a non-0 chance this is just a user wishing their other services were supported too?
I checked, it's an Imbue employee (or someone impersonating one, which seems a bit far fetched). I could could cite my sources but it feels like doxxing / drawing too much attention to the mistake of an individual. My read is that it was a separate github account that they let an agent control.
project literally launched yesterday... come on
Looks like the account in question is deleted now - do you have the original username?
adl-collab
It's only 11 in the last 48h, on lesser known projects.
11 in 48 hours is way too many and I'd call that spam.
2 replies →
Disclosure would be nice but 11-12 issues in entire GitHub is hardly spam.
Tell you what though, thank you for providing such a comprehensive list of cool projects to nerdsnipe me on a sunday morning!
The person doing this doesn't understand how pull requests work. The issues all say "I have a version in my fork". Open a pull request then! That's the whole point of forks on GitHub. Creating an issue first is pointless spamming akin to the classic "asking to ask". Just seems completely out of touch with open source to me.
> Creating an issue first is pointless spamming akin to the classic "asking to ask"
Many projects disagree with you on this one. It is quite common to have a policy that forbids PRs without an issue or discussion first.
1 reply →
Some projects require opening an issue first.
It's not spam if it's the first time I heard of it and it's useful. It shows they care about their project. This makes me want to try it out even more.
Meta and Google can spam their unsolicited ads to a billion users but regular people aren't allowed to promote their own projects on niche forums? This is bs. You're the problem here. Self-promoting your harmful values. You're the spammer.
So many docker app supervisors nowadays.
I'm surprised there's no such thing as a self-hosted deployment spec. An opinionated docker-compose file (or similar) with well defined inputs, outputs, and requirements that works for the average use-case.
I think the time for this space is here. The appetite to depart from subscriptions and loaning your personal data to ad/ai companies is stronger than ever.
Currently, a lot of this space is docker compose based and simply inaccessible to so many who may otherwise be interested. Whether or not it's this project or another, I hope something gathers enough steam soon to truly break the barriers to entry.
> The appetite to depart from subscriptions and loaning your personal data to ad/ai companies is stronger than ever.
Only on HN, the general public don't care, they don't even know the name of their web browser.
General public is weird because everyone, yet no one, is the general public.
Fwiw I've started to see a shift in many non-tech circles I'm a part of. Especially when you take the time to explain the problems. Many do care, they just aren't made aware of the issues.
1 reply →
The general public will do whatever their friends do. If there was an iPhone charger sized box they could plug in anywhere and have their own personal cloud that can interconnect with other personal clouds, and a celebrity said it was cool to have one, the general public will do it without giving it a first thought.
5 replies →
Precisely. My uncle has been using the builtin Samsung web browser for years on his phone, just because it said "Internet" in the name, and "Chrome" just wasn't a familiar name to him. Most people are like this.
1 reply →
Digital rights are indeed important, but only a small portion of the entire concept of human rights. I can imagine general public places much more emphasis on wealth (re-)distribution policies.
A lumberjack may also argue GP don't care about environment laws if they had timbernews.
True, most people get the full ad driven internet experience and it drives me nuts. Everytime I have to help other people with computer stuff I almost get an aneurysm when opening their browser...
1 reply →
How does being docker compose based make it inaccessible?
Self hosting implies having hardware and software knowledge to set it up. I agree, having a meta package that would set the system up as systemd would be an even lower barrier. But then again, there are podman commands that do this for you today anyway.
Self hosting could also imply buying a device that just does this with an additional management interface for updates. This could be build around a nuc/nas with above package. It still will require some knowledge to keep it maintained.
Then, you could offer maintaining them as a service again. Maybe there is value in that.
Update: apparently there is https://freedombox.org/ which is actually an appliance for this. Although it is more privacy focused and I don’t see a productivity or photo suite on it.
> How does being docker compose based make it inaccessible?
Because it requires people to also have knowledge about the ins and outs of docker and docker compose. If we are talking about web applications it is yet another layer that has been added over the years. It used to be that you could fairly easily host most things on a old fashioned lamp host (I am talking decades ago) and all you had to know was basic file transferring.
Of course, this was on shared webhosts in a time when VPSes were not really a thing yet or affordable. But, even on a VPS setting up a LAMP stack is relatively straightforward.
Once you add modern dockerized application to that mix you are now still looking at some sort of ingress to do the reverse proxy bit. So while it likely will not be Apache or nginx there will be some sort of layer there in addition to now having to setup docker properly (rootless and all that) and then also making sure your docker compose setup is in order.
It is not difficult *once you know how to* but it is another layer of knowledge and experience people need to acquire. That does make it less accessible for novices.
At the same time buying compute is more expensive than ever, and the gap between homemade projects and what's available online in term of speed, functionality and LLM capacity is bigger than ever.
Building the equivalent of Google Photos locally wasn't that hard 10 years ago, given enough tolerance for speed and indexing lag issues. Trying to do the same today will cost two arms and leg and require a lot more overall knowledge of what to compose to get the same result.
The dust will settle and we should go back to a nearer gap in I hope a few years, until the. it might be a glacier age for self-hosting.
> Building the equivalent of Google Photos locally wasn't that hard 10 years ago, given enough tolerance for speed and indexing lag issues.
You could argue it’s much easier on the software side, because self hosting Immich is much easier thanks for AI assistance with setting it up. If you can live without the built in facial recognition feature you don’t need particularly beefy hardware.
I find myself taking on far more ambitious self hosted projects thanks to Claude. Extending home assistant in various ways, deploying a “self hosted Spotify”, making sure everything is fully accessible via Tailscale out of the range of the wifi and so on. Of course the general population isn’t really aware or interested in most of this.
1 reply →
for self hosted and personal use, compute lasts a good while though. and in the time you would've been waiting to time the compute market, you get to benefit by being less a part of the AI data dragnet.
Small business might care about the cost of subscriptions if they could get high quality systems this way.
The biggest difficulty is probably that from my observation, there's a sharp cutoff line somewhere between Gen Y and Z where anyone born before that (and don't work in tech) genuinely have no idea how to navigate their digital life and gets preyed on by those subscriptions. And these people are still a super large portion of the population paying for those subscriptions.
Only Gen Z and later are starting to care and willing to spend the time and effort to self host. All the Gen X and Y people that I know of, including many that are very intelligent and accomplished in their field, literally have no idea how to operate a computer outside of watching youtube, checking emails, and opening Word documents. Ads? They hate it but have no idea how to disable them. iPhone running out of storage and iCloud says I can solve that for $2.99/mo? They would be like IDK what that is but sure I'll pay that $2.99 if that means I can record more videos of my kids.
I really hate this situation, and I try my best to help out the folks around me who I see are getting ripped off (either they don't know or they know but think that's what it is). But still, it's really hard to communicate or teach those people when the skill of operating a computer has become intuition and human nature to me. I really can't put these things into words on the fly that would help them understand :(
> Only Gen Z and later are starting to care and willing to spend the time and effort to self host.
I can't quite confirm that. My younger siblings' classmates (still high school students, born >=2010, non tech average people) don't give a shit on these topics and behave identical to the Gen X and Y described.
Looking at my younger sister, I'd risk saying it's the same in all generations. I'll even say more: I think earlier generations had a general thought of "computer consists of separate parts; you can buy them separately and replace if needed / want better performance", even the less technical ones. But now everything is a black box (therefore e.g. why not overspend and always buy a black box from Apple if that's fashionable now).
But maybe that was kind of related to times and the place – where I live people needed to be a bit more savvy, so thought like "why you buy an iPhone if Xiaomi can do the same 4x cheaper; if you chose the right model you may even have a good camera too" was popular even in non-tech circles.
But yeah, now that hardware is a bit more of a commodity and such tradeoff might give you a bit more adware with a bit less convenience, maybe it's different. But the next generation's understanding of computer/phone as something else than a magical black box is gone.
> Only Gen Z and later are starting to care and willing to spend the time and effort to self host.
Yeah, you’ll need a lot of supporting evidence for this claim. From my personal and professional experience, this doesn’t track. And given how online the younger generations are, it tracks even less. I think there might be a bias towards tech bubbles in that statement.
Or, alternatively, I was a lot more willing to put in the time to self host before I had more money and less time. If it's truly important it's not in public. If it's public, I don't have time to keep up with security patches if I'm not being paid.
But this is a company building on AI.
The project itself is open source and doesn't have any AI features.
Also while Imbue is in some sense an AI startup it does have a very different ethos from a lot of tech companies: https://imbue.com/about
> 53 is required because instance runs its own authoritative DNS server.
Are you planning to offer alternative solutions for this, e.g. pointing a wildcard A record at the server running the instance? I'd rather not run a publicly accessible DNS server.
"We (Imbue, the company I work for) also offer a managed version, which I think is really important to making this widely accessible - and it gives us a straightforward business model to support the project."
Looking at your hosted page, I see no reference to backups. I see in your docs reference to backup, but you should definitely offer some sort of turnkey, appropriately-marked-up backup solution. Configuration difficulty is what kills self-hosting on the front end, updates make it hard over time, but when it all dies and then I have no backup is when I give up.
I can't find any reference to disk size limits anywhere. Can I host Immich on this? Can I host my media server? What will it cost? Surely not $10/month for my media collection. I'm not a hoarder and mostly have stuff ripped from my personally-owned media, which limits the size, and you're still not hosting that for $10/month. I see you have Jellyfin in the set of apps but I don't know what it will cost. The CPU & RAM limits I understand from the deploy page.
Oh, and I'm specifically referring to your hosted service here.
(author here) Thanks for flagging, I should make this more clear. Our managed instances have pretty small disks (~50GB), but there's a way to link a S3 bucket to your instance to use for bulk data storage. Apps that store bulk data (like immich/jellyfin) should automatically put it in the "archive data tier" which can be configured to go into this S3 bucket instead of local storage. We're planning to make this (and backups) an auto-configured part of our managed instances - billed separately (so you pay for just the additional storage you use; cloudflare R2 (S3 compatible) at-cost is like $15/TB/mo so it should be in that neighborhood).
See docs on how this works: https://cloudinabottle.org/docs/how_it_works/data.html#the-a...
There seems to be a lot of variance between stated and revealed preferences around self-hosting.
It often seems like presenting the image of being autonomous to our peers seems more important than actually achieving it.
It takes a lot of energy to replicate what AWS and friends have done once we factor in concerns like the passage of time and entropy. I've never been able to keep a media/NAS appliance alive for much longer than 3-4 years. Inevitably, there is some kind of catastrophic event and I have to start all over with new vendors, etc. Even without any data loss this gets old. I promise you get tired of this after a while. It might seem like that would never happen and then one day it does.
You can get so much more done if you can actually afford to use the cloud. I have a really hard time believing that many people are genuinely winning on total cost of ownership with self hosting.
> if you can actually afford to use the cloud.
If you are an individual and you want to store a media library, you'll go broke instantly.
I have roughly 200TB of storage capacity, 100TB used, in my NAS. That would be $2300/mo on S3 before even getting to access fees and network egress.
How are you backing up 100TB of data? Have you factored the consequences of catastrophic data loss into the economics? How much value would you lose if all your data was gone?
There's no way you are consuming this much content actively. You could store all of it in S3 deep archive for $200/m (100TB) and maintain a tiny fraction in the standard resource pool.
This is a great project! Yes, the personal cloud is coming and it's going to be for everyone.
I'm working on a project with similar goals but some different design decisions - instead of the typical containers I'm going for thin-client apps that hit a common data layer built on a fully serverless architecture. Not ready for production quite yet, but for those interested: https://starkeep.app/
App catalog looks good, but they miss Home Assistant which is a major inconvenience
i dont think i understand the idea of cloud in a bottle - it seems like overthinking to me
This pitch is somewhat interesting, but, much like Sandstorm, browsing the library of pre-built apps kind of leaves me cold. It's too much like browsing a rather tiny mobile app store. There are maybe three apps that I might be interested in.
Since we have AI now, I think exe.dev's pitch is better: they provide Linux VMs, a web proxy, and integrations. You can ask the AI to build whatever web app you like. I'd love to have exe.dev, but running on my Mac Mini.
> I'd love to have exe.dev, but running on my Mac Mini.
I'm working on exactly this bit (with some more tuning for messaging AI agents)
https://github.com/gofixpoint/amika
Doesn't run on your Mac Mini yet, but should get that enabled either this week or next
If you sign up (free) I'll send you an email when the "Mac Mini mode + full OSS self-hosting" is live
What apps were you looking for that you wouldn't find in the catalog?
Wouldn't it make more sense to just ask AI to generate the cloud in a bottle files for the apps that are missing from the library?
If this can largely automate the backup and update side of things, then it seems like a great middle ground for people like me:
I’m trying to self-host as much of my stuff as possible, but don’t have time to tinker like I used to. Hence updates, backups (etc) fall by the wayside in favour of “set and forget”.
Found this and never looked anywhere after: https://github.com/opencloud-eu/opencloud
golang.
There is a bit more mature project that does the same: https://selfprivacy.org
selfprivacy appears to be limited to a handful of services. e.g. It is not clear to me if I can host postgresql with it. i.e. It is not a general purpose container hosting product.
Nice, and it's even supported by NLNet
I have been working on literally this exact project for the past 6 months, down to the nomenclature used. Running into this on HN is… strange.
Happens. Had an idea (simple and stupid one), built it (with LLMs) went to register the domain. Found out somebody had the same idea and already put it there :)
I've had this happen as well. Same exact name, and concept and highly specific application. It was extreamly niche too.
I'm trying to understand the difference between this project and FreedomBox. I mean it's great to have options, but have any of them looked over that way? Heck, FreedomBox is even built into Debian now.
Now the key question, can the hardware be made accessible to everyone. Its definitely an improvement to make self hosting software easier but the process of setting even a raspberry pi for an average consumer interested in self hosting is still a barrier. They probably will be able to figure it out via youtube and some reddit posts but there should be an easier process or way
(author here) I am really excited to make pre-configured Raspberry Pis that people can just plug in at home, and have it work out of the box. It'll take a bit more work to make it seamless but we're definitely aiming in that direction. It's a shame how expensive Pis have gotten recently though with the memory shortage.
That is cool to hear. I know lots of folks who would be genuinely interested in this who don't have the technological chops to set up a self hosted system. One angle i have heard is degoogling and having more privacy in their data. I am really genuinely interested in this venture. I had the same idea about a year ago but was not in a place in life to act on it. I think even with the PIs being expensive because of the people will still buy them. Especially if they are seamless or near seamless.
Making self-hosting really accessible isn't that easy, even if it's easy to setup the server machine itself. One needs to make their server machine accessible to everyone, which isn't generally possible due to how networks work, one needs a white IP address and making sure no incoming connections are rejected by the internet provider. Also one needs a domain. Obtaining it requires some bureaucratic burden, costs money and pricing isn't always transparent.
These barriers are high enough, so that common people can't create/host their own sites. Only skilled enthusiasts or professionals can afford having their own site.
People choose not to because they see no need to, but if they put some time, effort and not a lot of money then those barriers are not so high. I took the trouble to get a driving license because I saw value in that. I also took the trouble to learn to set up my own stuff on a VPS because I saw value in that too.
What's a "white IP address"? It is, sadly, more difficult than it should be, though. When everyone got online in the 90s they actually became bottom feeders and not true internet citizens. Not much has changed in 30 years. IPv6 is necessary, but the world also sucks now so people need to know what they're doing with firewalls etc. Domains are not too bad now, there are some decent registrars now like Porkbun.
"white IP address" is an address that your internet provider gives you and which is accessible from the whole internet, not just some address local for the provider's network and hidden behind NAT. Since IPv4 addresses are scarce, most providers don't allocate them directly to their customers.
An unique IPv6 may be theoretically assigned to each end device and thus solve the accessibility problem for locally-hosted servers, but I doubt internet providers do this in practice and I am afraid that many providers don't support IPv6 at all.
1 reply →
I really like the idea, and I would like to see Tailscale evolve Aperture in this direction. Tailscale is already an easy way for anyone to connect their devices, which is essential for any sort of "personal cloud", and Aperture, while initially something of an AI gateway, is also facilitating connections to other services and apps.
If you're building a personal cloud, you also want your AI connected to it, but probably with some degree of management. Aperture already provides a fair amount in this area, and probably with a connector for Cloud in a Bottle, could give your AI safe access.
Yep, tailscale, sync things. Glued together a lot of services
Reminds me of my old idea dating back to 2014 (https://linuxdreams.blogspot.com/2015/03/cloud-in-box-applia...). Interestingly the blog is still active.
I did a few pitches related to this at Nasscom Startup warehouse and IITM RTBI. Though this was welcomed by the panel members but it never took off due to lack of capital for such a startup during those times in India. I had to abandon and focus on my day job.
Looks really interesting! I'd like to see immich in the available apps. For me the Auth part is not super important. But has it one-click updates?
Cloudron does something very similar (I'm a user): https://www.cloudron.io/
Indeed, been using them since 2019 and it just works.
> making self-hosting accessible to everyone
This happens with making many small to medium level resource hungary apps, for personal usage, run on one VPS which is not a heady duty machine costing heavy duty money. I have tried this on similar machines 2GB RAM and 1-2 VCPU and after 2-3 apps it starts swaying like a drunkard in slow motion even in that SSH session. But at least this one seems to be better than something like PikaPods, in combo pricing regard (I am not saying one is generally better over the other), where the pricing is rather per app.
Interesting project! I've found self hosting a lot more approachable with the advent of AI to help with inevitable errors. I run every new app in its own service account and as a rootless podman quadlet. There are three main difficulties that I come up against with every app. First is backups, or more accurately restoring data. My understanding is that just taking a snapshot doesn't always give you a working restore point because an app could be mid-database migration or something like that. Second is updates, and third is accessibility outside the network. I am slowly working on my own solutions, interested to know how you tackle those though?
[flagged]
Basically what we were already doing before Cloud took over distributed computing and grid computing as industry terms, Open Stack and a few others.
>At the core, it's just an Ubuntu machine with a web server that hosts a dashboard and routes HTTP(s) requests to containerized apps.
So no fail over/redundancy? I think at minimum it should be two machines, so one can break. That is one of the important features of cloud applications as far as I'm concerned.
You don't have to deal with the problem of a single computer breaking.
This also adds a lot of moving parts and failure modes. For self-hosting I’d rather have downtime on a simple system that’s easy to fix rather than a complex setup that stays up but needs more maintenance when some component fails.
Do you have two instances of everything in your life?
It's clear this product is intended for personal use.
I guess you have a couch in your home. This couch can seat friends/family. Yet you don't have a spare one?
As soon as you add multiple instances you have to consider concurrency for every app you add, which is a huge cost on the app development side.
I've been really enjoying exe.dev
Instant spin ups and restarts
Easy interfaces ssh web cli AI agent
Integration aplenty
I'm confused, isn't a self-hosted cloud something like openstack or kubernetes?
Like the real ones, digital clouds can take many shapes.
Sometimes it means "rented virtual resources", sometimes it means "stack of intertwined services", sometimes it means "SaaS", sometimes it means "alternatives to SaaS"...
Here it seems to be the latter.
This is neat! I've been building something very similar for personal use. I may lift a little bit of the app manifest part.
I'm a little surprised that there's just one backup app, instead of each app coming with its own backup sidecar. To backup Immich, for instance, requires coordinating between its four different programs. I guess I'll read more about how that backup app works too ha
> nextcloud: slow and unreliable, and more enterprise-targeted now
bewcloud could be a lean alternative for home users if that's something you'd like to offer.
Has anyone else found NextCloud to be slow and unreliable? I've been using it for the last three years (switching over from OwnCloud before that), and it's never missed a beat. Straightforward to set up, decent linux desktop clients, has mobile support that's also been rock-solid, and basically does what it says on the tin. Doesn't feel remotely enterprise targeted to me either -- it feels very much targeted towards home users.
Thanks for bringing this up, I was going to ask the same.
I haven't heard any complaints about NextCloud, they've added some enterprise features, but all in a positive way, from what I can tell, not detracting from home and personal use.
We need gmail, docs and sheets replacements. With AI, maybe people can develop solid alternatives very quickly. And Linux phones.
Trillion dollar companies try to extract more and more from us, the only response is for us to develop alternative solutions and continuously chip away at their strengths- what made them monopolies in the first place.
Hey cool sounds like everyone's on the same page. Here is a start.
https://gitlab.com/here_forawhile/tmailplus
I haven't used it extensively, but I've heard cryptpad is good.
> Rent a box from whoever you like, and turn it into your own personal cloud.
Yeah, that’s not what cloud computing is.
One thing that I found LLMs useful for is devops.
I had rather good experience using agents to setup and deploy own cloud, Jellyfin and co. I experimented also getting a VPS, just giving the LLM the ssh keys and having at it. It’s pretty good.
If the AI stays afloat, then I think the future lies somewhere there. Not too sure about the security side of the things …
I'm mostly fine letting AI write code because its easily auditable and you can follow changes through the Git history.
Whereas, it you let the AI go free on a server, you mostly have no idea what it did and where, so it's quite hard to audit and even harder to try and fix things.
So, unless relying on a declarative paradigm like NixOS does, I'm not ready to give AI the keys to my castles
Admittedly this was not my castle. More of a $1.5/mo shed that I got just to try this.
This is not me saying security doesn't matter (because it does) but having everything not publicly accessible and only accessible over Tailscale or Wireguard makes me slightly happier.
In case someone is asking: THIS is what a launch article should be like. 10/10.
Running the authoritative dns for the zone seems elegant, although wouldn’t that imply you absolutely can’t use cloudflare or similar services to avoid ddos/bots?
Proxmox + community-scripts.org is pretty great.
A bit narrow definition of a "cloud", no..? More like a microservice server on rails
Bits missing from cloud being global CDN (selfhosted), routing, regional sharding, etc.
I recently bought a motherboard for my first self made pc. Six to eight sata slots is already enormous and the fact that you pay once and use your hardware freely opened my eyes. I was paying a subscription for cloud 1 tb for a trivial service. I am now hoarding my torrents worry free and plan to convert it to NAS. However i do still use my onedrive plan shared with three friends for 10 dollars for sharing large files till i set up my own nas with a good permanent net connection. The other issue is that I’m renting currently and if I’m changing countries I’ll have to rely on the onedrive again
accessible self hosting is already here, its called 'claude' or 'codex' you basically start it and tell it what you need, dont even have to know any software names.
I think what this needs to be truly accessible by general public is a mobile app (most people don’t have computers) and to be able to deploy it (including setting up domains) from there. Payments would definitely be a problem because of the various app stores’ rules.
Great project! I have had conversations along these lines with my colleagues for quite some time now. Will definitely try.
The docs mention cloudinabottle.toml but the app repos all have openhost.toml. Did you folks change the name of the platform?
Also, can you point to how the unified auth works? Couldn't find anything in the manifest spec.
Yeah, it was called Openhost for a while before changing to Cloud in a Bottle. Either filename works.
It proxies requests from the user to installed apps, and adds a header 'is-owner'. Apps have to be built to respect that header - some apps do this by running their own little proxy/sidecar that will detect that header and inject saved credentials.
An aside: Fantastico was such a wonderful piece of software for me to learn and toy and tinker with web stuff. I was so impressed by the simplicity of it. (I phased out of the shared scene a almost two decades ago; I guess everyone moved to Softaculous, https://softaculous.com/fantastico/)
How is this different from say Docker? I procrastinated getting acquainted with the latter for a long time because I thought it was more for industrial-scale infrastructure, but when I finally wound up using it I was surprised at how easy and low-maintenance it was.
UI and UX. Auth. Out of the box apps get their own domain name.
Try it out and see
I’m used to unraid for my home system. Is this an app that would run as a container across my jbod?
Curious how this compares to Cosmos Cloud (https://cosmos-cloud.io/).
I picked it a few years ago for much the same reason as anyone else (sick of this shit), and it's main attraction to me was it being a turnkey solution that focus security so i wasn't as likely to completely fuck everything important up.
I think a "simple" installation package would be very attractive to most people, and while cosmos has a lot of growing to do in order to be that, I'm still pretty happy with it.
I'm even more happy to see more in this space, because I'm just at the point of "do i stick with cosmos or do i start rolling it all myself" and god the leap is terrifying if you're thinking of exposing to the internet. I get it, but at the same time it feels like you're always one missed config away from a potential nightmare, and that's after the "welp guess this is my second job" problem.
Never thought that people on HN would be seriously discussing an obviously vibe coded thing.
I really like the design man!
The unified auth + inter-app permissions are the real differentiator vs Coolify/CapRover-style container hosts. Data tiers (local DBs vs S3/R2 archive) also look like the practical way to make Immich/Jellyfin viable without giant managed disks.
I thought that Coolify had centralized auth.
Like pikapod?
>claude in contibutors >python No, thanks
Meh. Maybe I'm a grinch, but I don't see this as easier. I have a little NUC running Ubuntu. Putting services on it is easy: PiHole, OwnCloud, Wireguard, etc.
> Apps are (rootless, hardened) containers, so it can run existing software with minimal changes - and in a reasonably secure sandbox.
WTF? I’m selling you a car with no brakes but it slows down reasonably fast?
seems kind of like https://caprover.com/
Yeah, but Docker Swarm?
a similar project which i think is a lot better than this (in fact it's so good i'm jealous of its genius) is https://exe.dev. It feels like something that hasn't really hit the mainstream yet but I suspect everyone will be talking about soon.
Closed source, solves a different problem, and just looks like generic agent sandboxes/VMs?
What am I missing.
Cloudron?
the hardware is problem we all been struggling with
How does it deal with storage? Storage is always the weakest link in self hosting. Having a bunch of stateless containers is ultimately just a fancy docker compose file. Still valuable, but you need a proper story for storage. Ideally all storage would be on individual volumes per app and all backed up etc.
systemd is all you need
And someone else's machine
systemd and ansible
> coolify: hosts containerized apps, but every app is an island with its own login and little integration with the host or other apps.
That’s not an issue tho, I want each application to be a separate login, so I can have one pointed to a different subdomain or domain not all in one login and location.
> nextcloud: slow and unreliable, and more enterprise-targeted now
What a load of bull.
I've been interested in something like this for awhile. Cloud or self hosting in-a-box. Should be simple for layman.
Seems weird not to mention Umbrel.
yunohost?
I remember back when I first got into self-hosting (over a decade ago), I had a Netgear READYNAS. People made a small amount of plugins you could install. But the CPU was ARM before that was cool (I think mine was actually SPARC!) so it was a lot of effort to get things to compile, so plugins were usually quite outdated.
Then I decided to painfully break out of vendor lock-in, built my own NAS with FreeNAS which was all the rage (still attached to a vendor, but at least I could control the hardware). This was before Docker really took off - the plugins were installed in FreeBSD jails. It worked for a while, but again, plugins could be woefully out of date and broken, upgrading the OS was a pain and would break all your plugins.
Through the 10+ years of this... I now follow two simple rules: - separate your network storage and application hosting. Yes, it complicates things a little bit with NFS setups but it is a forcing function for a more resilient setup - use the docker container provided by the maintainers or LinuxServer.io. Vendor maintained wrappers can be initially helpful but end up as a form of tech debt when they're abandoned or neutered - especially when they're freemium and the company starts looking to squeeze.
The quiet part of self-hosting is that it shifts the responsibility of security to the user. I'm sure cloudinabottle has more sensible defaults than provided docker containers, or builds in reverse proxies etc, but I doubt they're offering to take on liability for data loss or breaches.
The problem I see you'll face are two-fold: 1. Projects like this have the power users self-select away from it and those who need a lot of hand-holding to self-select into it. If I know how to configure and deploy a docker container already I have no incentive to contribute. ReadyNAS got around this by having folks charge for their plugins (that were just OSS projects packaged up for their OS) but then it ruins the appeal, they likely only got away with it because they were only a couple of bucks and you were already locked in to the ecosystem via the hardware. 2. If the hosted aspect actually makes you money, required for the project to be continued long term, capitalism will ensure that incumbent cloud providers take that away from you by offering ready-made instances at your cost-price, like they did with redis, elasticsearch etc.
Note that these are only issues if you're motivated by making a profit, or even a cost-neutral project. But if this is a charitable exercise backed by an entity that doesn't need the money, I wish you all the best - more things that let people dip their toes in are good! The cloud is all rent-seeking subscriptions and even with the recent supply shortages: fast, small, low-power compute is incredibly cheap these days, especially old business/mini machines.
IMO, one way to retain the power user crowd is to do the hard parts that they might have skipped themselves. For a couple examples:
- after selecting your observability apps, every _other_ app could automatically configure centralized logging, its own dashboard of metrics, etc
- the host OS could ask the user simply for the backup destination (S3 url creds, attached local disk, SMB share, etc) and every app is again automatically configured with backups
- the host could have a control panel to roll back apps between versions (maybe it's FS snapshots or something)
- everyone can use containers, but if the project went through the extra effort to make each app its own VM instead, that could be worth something
- if there was a companion Android/iOS app for management and (importantly) alerting?
---
That doesn't really solve the monetization issue you described, but I do think you can still get powerusers on board if you handle even harder problems than they'd want to take on themselves. Or...harder problems than they'd risk taking on themselves if their family uses what's being hosted, for example. Less tolerance for downtime
Edit: formatting
"Forjejo"
No offense, but this is not accessible to everyone. Seriously, take a look at the setup instructions and tell me with a straight face that a 4 year old could set this up.
https://cloudinabottle.org/docs/setup/shared_homeserver.html
There's no reason this couldn't be made as simple as double clicking a .exe and then the whole thing gets setup.
[flagged]
[flagged]
[dead]
[dead]
How do I reach out to you? We're building something that you might want to be a part of. (The "Safebox" comes with all the software you need, and you can just clone its AMI on AWS etc.) Would love to have a conversation.
If you want to message me, find my email on https://qbix.com/about
It's weird seeing someone trying to promote self-hosting, while building everything with Claude.
Does self-hosting need to be all or nothing?
I didn’t get that from their comment at all
AI is a tool. the use of this tool to deliver something unrelated is... unrelated
"It's weird seeing someone pay for food when they have a garden at home."
LLMs can be self hosted though. Using Claude to build a self hosting platform is a bit like telling people they should be vegan while you're eating a bacon sandwich.
> In the pre-cloud days, open source worked - authors distributed their software, and everyone "hosted" it for themselves (i.e. ran it on their own PC)
This is just rose tinted vision of the past. None of the software of the past had the capability, reliability or robustness of modern sass.
I'm rather baffled by that statement, with docker its...as easy as its ever been
"capability, reliability or robustness" has nothing to do with ease.
Docker was first released 11 years after AWS though, so I'm not really sure that it counts as "pre-cloud".
Regardless, I agree that Docker has made self-hosting stuff pretty easy, and arguably even easier than it was in the pre-cloud days.
Yea. When people talk about how simple things were in the past, I think they oblivious to the fact we were all writing out X11 configs by hand and having them break in all sort of strange and interesting ways. Open source has always been very "I gave it to you for free, I'm not going to hold your hand".
I've written X configs by hand, but only to get a few extra pixels of overscan. I've never needed to do this. My experience has always been that things just work. The UI has never been great, in that I need more explanation than the built-in manuals provide – unlike, say, Windows 95, where you can learn everything you need to know by clicking around – but it's not hard to avoid breaking things, and it's not that hard to learn to do new stuff if you have a good book (or, lately, blog post) to consult.
1 reply →
that statement does not make any claims about capability, reliability or robustness. the claims are rather cost and independence.