Open Source Didn't Ban AI Code. 83% of Its AI Policies Allow It, With One Catch
- Book: Prompt Engineering Pocket Guide
- Also by me: Thinking in Go (2-book series) — Complete Guide to Go Programming + Hexagonal Architecture in Go
- My project: Hermes IDE | GitHub — an IDE for developers who ship with Claude Code and other AI coding tools
- Me: xgabriel.com | GitHub
---
July was the month open source "banned AI." On July 22, Codeberg members voted 358 to 144 to bar projects built mostly from AI-generated code (It's FOSS). About a week later, the GCC Steering Committee said it won't take a legally significant contribution if any part of it traces back to an LLM (It's FOSS).
Read the GCC rule to the end, though. Contributions under the legal-significance threshold (roughly 15 lines) can still come in if they are clearly marked, test cases are exempt, and AI-assisted work carries an Assisted-by: tag in the commit message. Even the "ban" is mostly a set of conditions.
Then look past the two headlines at the whole field. In September, Andre Hora, Romain Robbes and Stefano Zacchiroli published ["We Permit the Use of AI, but [...]"](arxiv.org), a study of 281 AI policies from the 2,000 most-starred GitHub repositories plus 36 well-known projects and organizations such as the Linux kernel, LLVM and the Apache Software Foundation. The title is the finding.
The numbers say "yes, but"
From the paper's results:
- 83.3% of policies permit or encourage AI in code contributions (76.2% permit, 7.1% encourage).
- 14.9% forbid AI in code. Only 5.3% forbid it for every kind of contribution.
- 67.3% require a high level of human involvement.
- 43.4% make the contributor accountable for the result. None of the 281 say accountability is optional.
- 48.8% require you to disclose AI use. Of the disclosure rules that say where, 64.7% want it in the PR description and 26.7% in the commit message.
So a ban is the minority position. What the majority asks for is a human who was there: someone who directed the work, read it, and will answer for it. The paper quotes the policy of psf/requests on this: "you must remove any LLM co-author tags and take full ownership of every line."
One thing the study leaves open, and the authors list it as future work: whether anyone actually follows these policies.
What maintainers actually do with slop
The paper counts 188 countermeasures against low-quality AI contributions across the policies. The breakdown:
| Countermeasure | Share |
|---|---|
| Close the PR | 48.4% |
| Ban or block the user | 20.8% |
| Disallow autonomous agents | 13.8% |
| Restrict new or external users | 4.8% |
| Everything else (six measures) | 12.2% |
Nearly half of all measures are the same move: close it. The rqlite policy says slop PRs "may be closed without comment," per the paper. rust-analyzer goes further and says it does not allow autonomous agents to open pull requests or issues at all.
Some projects now address the agent directly. The paper finds policies in AGENTS.md and CLAUDE.md as well as CONTRIBUTING.md and AI_POLICY.md. The contributing guide of llama.cpp tells fully autonomous agents not to contribute and to exclude the repo from their configuration.
My take: they banned the unread diff, not the model
Put the two findings side by side. Most projects accept AI code. The most common response to bad AI code is closing the PR. The deciding factor between those two outcomes is whether a human read the change before a maintainer had to.
A closed PR is a maintainer telling you the review didn't happen on your side, so it isn't going to happen on theirs. Reviewing a large agent diff costs real time, and maintainers are often volunteers. The policies push that cost back to the sender, which is fair. The diff can be cheap to generate. Someone still has to understand it, and the 67.3% requiring high human involvement is the field agreeing on who that someone is.
That moves the review upstream. If the reading has to happen, it should happen on your machine, before the PR exists. Once you open it, every unread line is a maintainer's problem.
Three steps before you open the PR
1. Read the project's policy first. It might live in a file you don't expect:
ls AI_POLICY.md LLM_POLICY.md AGENTS.md CLAUDE.md \
CONTRIBUTING.md 2>/dev/null
grep -inwE "ai|llm|agent|copilot" CONTRIBUTING.md \
2>/dev/null
Then follow its format. There isn't one standard. The paper notes RustPython asks for Assisted-by and not Co-authored-by, Nextcloud requires an Assisted-by trailer, and psf/requests wants LLM co-author tags gone entirely.
2. Read the change step by step. One big diff hides the step where the agent rewrote a file you didn't ask about. Start with the summary, then walk the branch commit by commit, oldest first:
git diff main...HEAD --stat
git log -p --reverse main..HEAD
The --stat line shows the file you didn't ask about. The log shows each step in order. If you can't explain a hunk, ask the agent to undo it or redo it smaller. "Full ownership of every line" starts here.
3. Disclose where the project asks. Git has a flag for trailers (git docs):
git commit -m "Fix off-by-one in pagination" \
--trailer "Assisted-by: Claude Code"
Then say it in the PR description too, since that's where most disclosure rules point: which agent, what it did, what you checked by hand.
Doing the reading in Hermes IDE
I build Hermes IDE, so weigh this section with that in mind. Everything above works with plain git. Hermes is built around that order of operations: the agent works and you read before you land.
Every agent turn is saved, so you can diff or restore any single turn, with undo. The Review Desk (⌘G, or Ctrl+Shift+G on Windows and Linux) shows the changes by turn or by file, with risk flags, and a line comment goes back to the agent that wrote that line. That's the "read it turn by turn" step without scrolling one long diff (README).
!The Hermes Review Desk in By turn view, listing two turns of a Claude Code task with turn 1's diff of src/db.ts and a drafted commit message
The Review Desk (⌘G) in By turn view: two turns of a Claude Code task, turn 1's diff of src/db.ts, the changed files, and a commit message drafted from the turns.
Only after that does Land come in. It commits, opens a pull request or merges locally, into the branch the task started from, and it can be undone. Check that the commit message carries the trailer your project asks for before you land. Each task also runs on its own hermes/ branch and worktree, so the PR holds one task's work and nothing from the agent next door.
You can download Hermes. It is source-available under BSL 1.1 and runs on macOS, Windows and Linux. It runs Claude Code, Codex and other agents in their own terminals, with your own sign-ins. It isn't affiliated with Nous Research or its Hermes Agent. It can't make an agent's change good, but it gives you a place to check the change first.
Where do you draw the line?
If you maintain a project, does your policy say what "human involvement" means in practice? If you contribute, do you disclose the agent by default, or only when a project asks? Tell me in the comments.
---
If this was useful
Projects now write their AI rules into AGENTS.md and CLAUDE.md, which means a contribution policy is also a prompt. The Prompt Engineering Pocket Guide covers how to write instructions an agent actually follows, whether you're the maintainer writing them or the contributor whose agent reads them. If you'd rather review agent work turn by turn, Hermes IDE is the tool I build for exactly that.
My self-hosting graveyard is longer than the list of survivors
I didn't set up my first server for anything this post is about (for reasons of my own, that part stays a secret under seven seals:D), but it very quickly found a perfectly legitimate and much more boring purpose: I needed a place to deploy my own pet projects - for tests, for demos to friends, to poke at something live instead of localhost - without running into Vercel and Render with their free tiers, strict per-service limits and subscriptions that grow faster than my projects do
A VPS has only one limit, the hardware, and as it turned out, that's much easier to negotiate with than someone else's price list ;)
Then I noticed that most of the time this hardware just sits idle, and I asked myself a fair question: what exactly am I paying for? Right around then I stumbled on the awesome-selfhosted repo, and down the rabbit hole I went… :D
Right now there are three machines, not counting my main PC: a small rented VPS in Europe with 3 gigabytes of memory and 30 gigabytes of disk, which everything that has to be reachable from the internet goes through; a home server that lives at my family's place, because electricity and internet are simply cheaper there; and a small VPS for learning and doing scary things by hand (so I don't end up restoring the production box from backups), with 1 gigabyte of RAM and about 10 gigabytes of disk
The home media server runs a Ryzen 3 PRO 3200G, 12 gigabytes of memory, a 480-gigabyte system SSD and two hard drives: a 4 TB WD Gold for everything valuable and an 8 TB Toshiba N300 for media
They're called tank and junk, and I don't think I need to explain which is which :D
Tank needs a second drive to mirror it, so the backups don't live on a single disk, but storage prices are such that it's easier to buy a Chevrolet Camaro than a new hard drive that size, and good deals on the used market are rare
The current one was a lucky find: about $80 for a WD Gold WD4002FYYZ with just over 200 power cycles and a couple of years of service (and of course I ran the tests, the drive is livelier than plenty of modern ones)
In all honesty: of those 12 gigabytes, the system currently sees less than ten, because the CMOS battery on the motherboard is dead, so after every power cut it forgets my settings and once again generously hands two gigabytes to the integrated graphics, which a server with no monitor has roughly zero use for
For about two months now I've been promising myself every weekend to swap the battery and take that memory back, but, as you can probably tell, I'm a lazy ass, and my weekends are taken up by more important things (lying around, doomscrolling Instagram reels, playing with my cat, sinking ten hours into Civilization V... you know how it is:D)
The front door
The VPS mostly hosts what glues everything else together and keeps it safe, plus a couple of things I use every day
My blog: about a hundred megabytes
Homepage: a start page with a tile for every service
Out of the box it's boring, and I wanted something pretty, so mine has glass, a background and my own palette, and I use it quite often, because its UI and UX really are great
Uptime Kuma: monitoring
Strictly speaking, having a server check whether it's still alive is a so-so idea, and once mine was down for half an hour without anyone noticing, but it keeps a proper eye on the containers, and that's why I keep it (the obvious fix is to run it on the home server and watch the VPS from outside, it's on the list)
Beszel: CPU, memory and disk graphs for every machine
Right now it just sits there: Kuma and Termix cover pretty much everything I need from it, so Beszel is like that distant relative you used to be really close to and now call once a year :D
Termix: SSH right in the browser
Nexterm lived exactly one day before it, and Termix stayed because it won on every count at once: features, easy setup, convenience, looks and snippets, and I honestly don't know how I managed servers from my phone before them
Memos: notes where I write down absolutely everything, and it has completely replaced Saved Messages in Telegram for me
I'm a fan of minimalism and simplicity, and I don't see the point of an app where I'll never touch most of the features, and most importantly, it's self-hosted, btw
The idea of bolting AI onto it for tags and summaries is sitting in the backlog: there are higher-priority projects right now, but SOMEDAY I'll automate this part of my life:D
MeshCentral: remote access to my main PC
All my code and Claude Code live on it, and I often work away from home, so I needed access from anywhere, right in the browser, without installing anything on the machine I'm connecting from
Donetick: household chores for the family, and right now it's on pause, because I still have questions about the interface and some of the features
But it comes with my favorite story: I forked its web app to translate it, and it turned into a full i18n integration and some real help for the maintainer, so now I'm listed as a contributor in donetick/frontend
Vaultwarden: a password manager
I used Google Password Manager for years, and that was perfectly fine right up until I started running my own services and suddenly realized I was seriously worried about where and how my data was stored, and since I'm running a server anyway, it made sense to keep my passwords there too
Pocket ID: passkey sign-in to almost all my services, no passwords
When you have twenty services, typing a password into each one gets old within half an hour, and I also wanted my self-written MCP servers to sit behind my own security layer, not behind some third-party platform I'd have to take at its word
Backrest: nightly backups that go to the cloud and then get pulled home
Knock on wood, I haven't had to restore from them yet, apart from a test run, and I picked it because I like seeing what's actually happening: I enjoy working in the console, but a UI is still more familiar and more pleasant
IT-Tools: a set of small converters and generators
It can stay :D
And two of my own pet projects: MindAmp, mind maps with an LLM layer that speeds up the work and boosts the research part with agents, no third-party subscriptions needed, and VigilAMP, a registry of my projects and their current status that keeps me from getting lost in what I'm working on and what stage it's at
All of this fits into one and a half gigabytes of memory, with about as much left in reserve, and the disk is 85% full
Yes, I know, and no, I'm not doing anything about it today:/
The warehouse
The home server is for the heavy stuff: things that take up a lot of space, and things I'd rather not keep on rented hardware
The rule here is simple: I test a service on myself first and only then bring the family in, which is why some of these are still just mine
Nextcloud with OnlyOffice: my files and documents, a replacement for Google Drive, which is now just a backup
People often call Nextcloud heavy, and the numbers I'd read were scary: 1-1.2 GB of RAM against 300-400 MB for Seafile, so I was all set to go with Seafile
Then I decided not to take it on faith and ran both side by side on the home server
At idle, the whole Seafile stack (Seafile, MariaDB, memcached) sat at about 450 MiB, and Nextcloud with Postgres and Redis at about 270
Right now the Nextcloud stack, cron included, idles at around 240 MiB, and with OnlyOffice on top it's about 415, roughly 4% of the 9.6 GB my dead battery lets me have
OnlyOffice is the only container with a cap (2 GB), because it spins up a worker for every open document and easily climbs into the hundreds of MB while someone's editing
These are idle numbers
Previews and big syncs make Nextcloud heavier, and my setup is gentle: SSD, LAN only, not much data
Still, I haven't run into any problems, it's stable and pretty spry, and on my hardware the memory argument fell apart, so Nextcloud stayed and Seafile went (I could be wrong, ofc, but I'm too lazy to test it under load, and I don't use it that much anyway: it's basically a glorified file dump with a few nice apps that let me get to my files from anywhere)
Immich: photos, a great Google Photos replacement, and it's still at the "me first" stage before I bring the family in
BookLore: a book library, and I chose it for FB2 support, because my whole library is in that format
So far I'm the only one using it
BookLore has since gone into maintenance mode, and its developer named BookOrbit as the official successor, so right now I'm trying it out next to BookLore, but I don't have an opinion on it yet
Jellyfin: movies and TV shows
It's fully open and free, does exactly what it needs to, and these days we watch almost everything I own through it at home
SearXNG: a metasearch engine, set up for privacy and security, like most of my services
I've barely used it seriously yet
Navidrome: my music library, and I listen to it every day: all my music goes through it now, no subscriptions, no limits
On my PC I listen through Psysonic, and on my phone through Solound
The graveyard
The list of things I've removed is longer than the list of things that run, and every one of the deceased had its own cause of death
And these are just the ones worth a eulogy: in total around fifty services got installed and then deleted
Some got a second chance, like Dashy and Karakeep, and Windmill filled the disk to 100%
Ghost and Astro with Keystatic were my first two shots at my blog, and each lasted about two days
Ghost brought a crowded communal apartment of other people's services, and Keystatic kept its admin on Vite's dev server, the one you're not supposed to expose to the internet
So I ended up writing my own
Finding a browser terminal took longer than I'd like
ttyd dropped the connection every few minutes
Cockpit worked, but I only needed it for the terminal and files, and dragging in a whole server control panel for that is overkill
Nexterm was lightweight, but it was still in beta
Termix won
Seafile was supposed to be the lightweight one and lost to Nextcloud on memory (the numbers are up in the warehouse)
Termix also finished off Portainer: it turned out I'd rather manage containers straight from the terminal than click through yet another panel
Karakeep, bookmarks with previews, turned out to be three containers, 3.5 GB of images and 700 MB of memory
Without the built-in Chrome for previews it's a completely different product, and keeping something like that on a VPS with three gigabytes is a luxury I can't afford
Trilium was out because of its cluttered interface and broken built-in AI
Minimalism, remember? Notes went to Memos in the end
Anki Sync Server lasted two days: I set it up for my wife, my wife gave up on AnkiDroid, and there was nothing left to sync
And Inbox Zero never even made it to installation: I wrote a whole spec for it and then decided I didn't need it (and there were security questions too, but that stays between us:D)
My rule is this: every new service gets compared to what already works, and if it doesn't click quickly and isn't critical, it gets removed without hours of debugging, which is exactly why 3 gigabytes is still enough
Why all this
Partly for control: my photos and files sit on disks I can touch with my own hands, and my passwords at least live on a server I run myself
But mostly for learning: I write backend in Python and I'm learning everything around the code, and the servers are my sandbox for that, from networking and Docker to backups, monitoring and security
And when something breaks at three in the morning, it's my fault and my lesson ¯\_(ツ)_/¯
So what's in your graveyard, and what killed it?
10 comments
Mentions
Open Source Didn't Ban AI Code. 83% of Its AI Policies Allow It, With One Catch - Book: Prompt Engineering Pocket Guide - Also by me: Thinking in Go (2-book series) — Complete Guide to Go Programming + Hexagonal Architecture in Go - My project: Hermes IDE | GitHub — an IDE for developers who ship with Claude Code and other AI coding tools - Me: xgabriel.com | GitHub --- July was the month open source "banned AI." On July 22, Codeberg members voted 358 to 144 to bar projects built mostly from AI-generated code (It's FOSS). About a week later, the GCC Steering Committee said it won't take a legally significant contribution if any part of it traces back to an LLM (It's FOSS). Read the GCC rule to the end, though. Contributions under the legal-significance threshold (roughly 15 lines) can still come in if they are clearly marked, test cases are exempt, and AI-assisted work carries an Assisted-by: tag in the commit message. Even the "ban" is mostly a set of conditions. Then look past the two headlines at the whole field. In September, Andre Hora, Romain Robbes and Stefano Zacchiroli published ["We Permit the Use of AI, but [...]"](arxiv.org), a study of 281 AI policies from the 2,000 most-starred GitHub repositories plus 36 well-known projects and organizations such as the Linux kernel, LLVM and the Apache Software Foundation. The title is the finding. The numbers say "yes, but" From the paper's results: - 83.3% of policies permit or encourage AI in code contributions (76.2% permit, 7.1% encourage). - 14.9% forbid AI in code. Only 5.3% forbid it for every kind of contribution. - 67.3% require a high level of human involvement. - 43.4% make the contributor accountable for the result. None of the 281 say accountability is optional. - 48.8% require you to disclose AI use. Of the disclosure rules that say where, 64.7% want it in the PR description and 26.7% in the commit message. So a ban is the minority position. What the majority asks for is a human who was there: someone who directed the work, read it, and will answer for it. The paper quotes the policy of psf/requests on this: "you must remove any LLM co-author tags and take full ownership of every line." One thing the study leaves open, and the authors list it as future work: whether anyone actually follows these policies. What maintainers actually do with slop The paper counts 188 countermeasures against low-quality AI contributions across the policies. The breakdown: | Countermeasure | Share | |---|---| | Close the PR | 48.4% | | Ban or block the user | 20.8% | | Disallow autonomous agents | 13.8% | | Restrict new or external users | 4.8% | | Everything else (six measures) | 12.2% | Nearly half of all measures are the same move: close it. The rqlite policy says slop PRs "may be closed without comment," per the paper. rust-analyzer goes further and says it does not allow autonomous agents to open pull requests or issues at all. Some projects now address the agent directly. The paper finds policies in AGENTS.md and CLAUDE.md as well as CONTRIBUTING.md and AI_POLICY.md. The contributing guide of llama.cpp tells fully autonomous agents not to contribute and to exclude the repo from their configuration. My take: they banned the unread diff, not the model Put the two findings side by side. Most projects accept AI code. The most common response to bad AI code is closing the PR. The deciding factor between those two outcomes is whether a human read the change before a maintainer had to. A closed PR is a maintainer telling you the review didn't happen on your side, so it isn't going to happen on theirs. Reviewing a large agent diff costs real time, and maintainers are often volunteers. The policies push that cost back to the sender, which is fair. The diff can be cheap to generate. Someone still has to understand it, and the 67.3% requiring high human involvement is the field agreeing on who that someone is. That moves the review upstream. If the reading has to happen, it should happen on your machine, before the PR exists. Once you open it, every unread line is a maintainer's problem. Three steps before you open the PR 1. Read the project's policy first. It might live in a file you don't expect: ls AI_POLICY.md LLM_POLICY.md AGENTS.md CLAUDE.md \ CONTRIBUTING.md 2>/dev/null grep -inwE "ai|llm|agent|copilot" CONTRIBUTING.md \ 2>/dev/null Then follow its format. There isn't one standard. The paper notes RustPython asks for Assisted-by and not Co-authored-by, Nextcloud requires an Assisted-by trailer, and psf/requests wants LLM co-author tags gone entirely. 2. Read the change step by step. One big diff hides the step where the agent rewrote a file you didn't ask about. Start with the summary, then walk the branch commit by commit, oldest first: git diff main...HEAD --stat git log -p --reverse main..HEAD The --stat line shows the file you didn't ask about. The log shows each step in order. If you can't explain a hunk, ask the agent to undo it or redo it smaller. "Full ownership of every line" starts here. 3. Disclose where the project asks. Git has a flag for trailers (git docs): git commit -m "Fix off-by-one in pagination" \ --trailer "Assisted-by: Claude Code" Then say it in the PR description too, since that's where most disclosure rules point: which agent, what it did, what you checked by hand. Doing the reading in Hermes IDE I build Hermes IDE, so weigh this section with that in mind. Everything above works with plain git. Hermes is built around that order of operations: the agent works and you read before you land. Every agent turn is saved, so you can diff or restore any single turn, with undo. The Review Desk (⌘G, or Ctrl+Shift+G on Windows and Linux) shows the changes by turn or by file, with risk flags, and a line comment goes back to the agent that wrote that line. That's the "read it turn by turn" step without scrolling one long diff (README). !The Hermes Review Desk in By turn view, listing two turns of a Claude Code task with turn 1's diff of src/db.ts and a drafted commit message The Review Desk (⌘G) in By turn view: two turns of a Claude Code task, turn 1's diff of src/db.ts, the changed files, and a commit message drafted from the turns. Only after that does Land come in. It commits, opens a pull request or merges locally, into the branch the task started from, and it can be undone. Check that the commit message carries the trailer your project asks for before you land. Each task also runs on its own hermes/ branch and worktree, so the PR holds one task's work and nothing from the agent next door. You can download Hermes. It is source-available under BSL 1.1 and runs on macOS, Windows and Linux. It runs Claude Code, Codex and other agents in their own terminals, with your own sign-ins. It isn't affiliated with Nous Research or its Hermes Agent. It can't make an agent's change good, but it gives you a place to check the change first. Where do you draw the line? If you maintain a project, does your policy say what "human involvement" means in practice? If you contribute, do you disclose the agent by default, or only when a project asks? Tell me in the comments. --- If this was useful Projects now write their AI rules into AGENTS.md and CLAUDE.md, which means a contribution policy is also a prompt. The Prompt Engineering Pocket Guide covers how to write instructions an agent actually follows, whether you're the maintainer writing them or the contributor whose agent reads them. If you'd rather review agent work turn by turn, Hermes IDE is the tool I build for exactly that. 
Nextcloud Sync Windows 11 not launching Hello, I just updated my nextcloud sync instance this morning on my Windows 11 Pro 26h2 computer, and now it won't launch. So far I've tried restarting my computer, as well as uninstalling and reinstalling Nextcloud 35.0.1, and it says it installs fine, but then it doesn't open to sync anything. I've also tried running it as an administrator, and similarly nothing. Not really sure what else to try at the moment. Edit: I downloaded version 35.0.0 from the repository here: download.nextcloud.com and it launched immediately. So seems like something is wrong with 35.0.1 Edit 2: It looks like someone reported the issue on Github already: desktop#11178
Nextcloud All-in-One: The official Nextcloud installation method (11k stars) What it does All text below is quoted directly from the project README. >The official Nextcloud installation method. Provides easy deployment and maintenance with most features included in this one Nextcloud instance. Included are: - Nextcloud - High performance backend for Nextcloud Files (Client Push) - Redis & APCU for performant caching - PostgreSQL as database - Nextcloud Office (optional) - High performance backend for Nextcloud Talk and TURN-server (optional) - Backup solution (optional, based on BorgBackup) - ClamAV (optional, Antivirus backend for Nextcloud) >Simple web interface included that enables easy installation and maintenance Repository nextcloud/all-in-one Repository stats |Metric|Value| |:-|:-| |Stars|10,736| |Forks|1.1k| |Last release|v14.4.1 (2026-10-21)| |Languages|PHP 47.9%, Shell 24.6%, Twig 12%| |Contributors|196| |AI contributor accounts (Claude / Codex / Cursor / Copilot)|@Copilot (61 commits), @claude (40 commits)| |AI co-authored commits|Claude: 40| Main contributor |Field|Value| |:-|:-| |Name|Simon L.| |GitHub|@szaimen| |Commits to this repo|3,617 (top contributor)| |Bio|Designer, Software Engineer and Project Manager| |Company / Location|Nextcloud GmbH / Berlin| |Followers|354| ────────────────────────────── 📌 How many great GitHub repos have you discovered and completely forgotten about? Don't let another useful tool disappear into your Reddit feed. Get the community's best repos, organized by category, plus key takeaways from real developer discussions, delivered to your inbox twice a week. 📬 Never lose track of great repos again. Subscribe free.
My self-hosting graveyard is longer than the list of survivors I didn't set up my first server for anything this post is about (for reasons of my own, that part stays a secret under seven seals:D), but it very quickly found a perfectly legitimate and much more boring purpose: I needed a place to deploy my own pet projects - for tests, for demos to friends, to poke at something live instead of localhost - without running into Vercel and Render with their free tiers, strict per-service limits and subscriptions that grow faster than my projects do A VPS has only one limit, the hardware, and as it turned out, that's much easier to negotiate with than someone else's price list ;) Then I noticed that most of the time this hardware just sits idle, and I asked myself a fair question: what exactly am I paying for? Right around then I stumbled on the awesome-selfhosted repo, and down the rabbit hole I went… :D Right now there are three machines, not counting my main PC: a small rented VPS in Europe with 3 gigabytes of memory and 30 gigabytes of disk, which everything that has to be reachable from the internet goes through; a home server that lives at my family's place, because electricity and internet are simply cheaper there; and a small VPS for learning and doing scary things by hand (so I don't end up restoring the production box from backups), with 1 gigabyte of RAM and about 10 gigabytes of disk The home media server runs a Ryzen 3 PRO 3200G, 12 gigabytes of memory, a 480-gigabyte system SSD and two hard drives: a 4 TB WD Gold for everything valuable and an 8 TB Toshiba N300 for media They're called tank and junk, and I don't think I need to explain which is which :D Tank needs a second drive to mirror it, so the backups don't live on a single disk, but storage prices are such that it's easier to buy a Chevrolet Camaro than a new hard drive that size, and good deals on the used market are rare The current one was a lucky find: about $80 for a WD Gold WD4002FYYZ with just over 200 power cycles and a couple of years of service (and of course I ran the tests, the drive is livelier than plenty of modern ones) In all honesty: of those 12 gigabytes, the system currently sees less than ten, because the CMOS battery on the motherboard is dead, so after every power cut it forgets my settings and once again generously hands two gigabytes to the integrated graphics, which a server with no monitor has roughly zero use for For about two months now I've been promising myself every weekend to swap the battery and take that memory back, but, as you can probably tell, I'm a lazy ass, and my weekends are taken up by more important things (lying around, doomscrolling Instagram reels, playing with my cat, sinking ten hours into Civilization V... you know how it is:D) The front door The VPS mostly hosts what glues everything else together and keeps it safe, plus a couple of things I use every day My blog: about a hundred megabytes Homepage: a start page with a tile for every service Out of the box it's boring, and I wanted something pretty, so mine has glass, a background and my own palette, and I use it quite often, because its UI and UX really are great Uptime Kuma: monitoring Strictly speaking, having a server check whether it's still alive is a so-so idea, and once mine was down for half an hour without anyone noticing, but it keeps a proper eye on the containers, and that's why I keep it (the obvious fix is to run it on the home server and watch the VPS from outside, it's on the list) Beszel: CPU, memory and disk graphs for every machine Right now it just sits there: Kuma and Termix cover pretty much everything I need from it, so Beszel is like that distant relative you used to be really close to and now call once a year :D Termix: SSH right in the browser Nexterm lived exactly one day before it, and Termix stayed because it won on every count at once: features, easy setup, convenience, looks and snippets, and I honestly don't know how I managed servers from my phone before them Memos: notes where I write down absolutely everything, and it has completely replaced Saved Messages in Telegram for me I'm a fan of minimalism and simplicity, and I don't see the point of an app where I'll never touch most of the features, and most importantly, it's self-hosted, btw The idea of bolting AI onto it for tags and summaries is sitting in the backlog: there are higher-priority projects right now, but SOMEDAY I'll automate this part of my life:D MeshCentral: remote access to my main PC All my code and Claude Code live on it, and I often work away from home, so I needed access from anywhere, right in the browser, without installing anything on the machine I'm connecting from Donetick: household chores for the family, and right now it's on pause, because I still have questions about the interface and some of the features But it comes with my favorite story: I forked its web app to translate it, and it turned into a full i18n integration and some real help for the maintainer, so now I'm listed as a contributor in donetick/frontend Vaultwarden: a password manager I used Google Password Manager for years, and that was perfectly fine right up until I started running my own services and suddenly realized I was seriously worried about where and how my data was stored, and since I'm running a server anyway, it made sense to keep my passwords there too Pocket ID: passkey sign-in to almost all my services, no passwords When you have twenty services, typing a password into each one gets old within half an hour, and I also wanted my self-written MCP servers to sit behind my own security layer, not behind some third-party platform I'd have to take at its word Backrest: nightly backups that go to the cloud and then get pulled home Knock on wood, I haven't had to restore from them yet, apart from a test run, and I picked it because I like seeing what's actually happening: I enjoy working in the console, but a UI is still more familiar and more pleasant IT-Tools: a set of small converters and generators It can stay :D And two of my own pet projects: MindAmp, mind maps with an LLM layer that speeds up the work and boosts the research part with agents, no third-party subscriptions needed, and VigilAMP, a registry of my projects and their current status that keeps me from getting lost in what I'm working on and what stage it's at All of this fits into one and a half gigabytes of memory, with about as much left in reserve, and the disk is 85% full Yes, I know, and no, I'm not doing anything about it today:/ The warehouse The home server is for the heavy stuff: things that take up a lot of space, and things I'd rather not keep on rented hardware The rule here is simple: I test a service on myself first and only then bring the family in, which is why some of these are still just mine Nextcloud with OnlyOffice: my files and documents, a replacement for Google Drive, which is now just a backup People often call Nextcloud heavy, and the numbers I'd read were scary: 1-1.2 GB of RAM against 300-400 MB for Seafile, so I was all set to go with Seafile Then I decided not to take it on faith and ran both side by side on the home server At idle, the whole Seafile stack (Seafile, MariaDB, memcached) sat at about 450 MiB, and Nextcloud with Postgres and Redis at about 270 Right now the Nextcloud stack, cron included, idles at around 240 MiB, and with OnlyOffice on top it's about 415, roughly 4% of the 9.6 GB my dead battery lets me have OnlyOffice is the only container with a cap (2 GB), because it spins up a worker for every open document and easily climbs into the hundreds of MB while someone's editing These are idle numbers Previews and big syncs make Nextcloud heavier, and my setup is gentle: SSD, LAN only, not much data Still, I haven't run into any problems, it's stable and pretty spry, and on my hardware the memory argument fell apart, so Nextcloud stayed and Seafile went (I could be wrong, ofc, but I'm too lazy to test it under load, and I don't use it that much anyway: it's basically a glorified file dump with a few nice apps that let me get to my files from anywhere) Immich: photos, a great Google Photos replacement, and it's still at the "me first" stage before I bring the family in BookLore: a book library, and I chose it for FB2 support, because my whole library is in that format So far I'm the only one using it BookLore has since gone into maintenance mode, and its developer named BookOrbit as the official successor, so right now I'm trying it out next to BookLore, but I don't have an opinion on it yet Jellyfin: movies and TV shows It's fully open and free, does exactly what it needs to, and these days we watch almost everything I own through it at home SearXNG: a metasearch engine, set up for privacy and security, like most of my services I've barely used it seriously yet Navidrome: my music library, and I listen to it every day: all my music goes through it now, no subscriptions, no limits On my PC I listen through Psysonic, and on my phone through Solound The graveyard The list of things I've removed is longer than the list of things that run, and every one of the deceased had its own cause of death And these are just the ones worth a eulogy: in total around fifty services got installed and then deleted Some got a second chance, like Dashy and Karakeep, and Windmill filled the disk to 100% Ghost and Astro with Keystatic were my first two shots at my blog, and each lasted about two days Ghost brought a crowded communal apartment of other people's services, and Keystatic kept its admin on Vite's dev server, the one you're not supposed to expose to the internet So I ended up writing my own Finding a browser terminal took longer than I'd like ttyd dropped the connection every few minutes Cockpit worked, but I only needed it for the terminal and files, and dragging in a whole server control panel for that is overkill Nexterm was lightweight, but it was still in beta Termix won Seafile was supposed to be the lightweight one and lost to Nextcloud on memory (the numbers are up in the warehouse) Termix also finished off Portainer: it turned out I'd rather manage containers straight from the terminal than click through yet another panel Karakeep, bookmarks with previews, turned out to be three containers, 3.5 GB of images and 700 MB of memory Without the built-in Chrome for previews it's a completely different product, and keeping something like that on a VPS with three gigabytes is a luxury I can't afford Trilium was out because of its cluttered interface and broken built-in AI Minimalism, remember? Notes went to Memos in the end Anki Sync Server lasted two days: I set it up for my wife, my wife gave up on AnkiDroid, and there was nothing left to sync And Inbox Zero never even made it to installation: I wrote a whole spec for it and then decided I didn't need it (and there were security questions too, but that stays between us:D) My rule is this: every new service gets compared to what already works, and if it doesn't click quickly and isn't critical, it gets removed without hours of debugging, which is exactly why 3 gigabytes is still enough Why all this Partly for control: my photos and files sit on disks I can touch with my own hands, and my passwords at least live on a server I run myself But mostly for learning: I write backend in Python and I'm learning everything around the code, and the servers are my sandbox for that, from networking and Docker to backups, monitoring and security And when something breaks at three in the morning, it's my fault and my lesson ¯\_(ツ)_/¯ So what's in your graveyard, and what killed it?
💸 网盘一年三百多，到期限速 100K！Nextcloud 桌面端+服务端自建网盘，三台设备同步、照片自动备份、共享链接全有，空间自己定。 nextcloud/desktop
Ja, herzlichen GuMo! Weil wir noch nicht genug Updates hatten in den letzten Tagen lädt euch #Nextcloud zur nächsten Runde ein: github.com/nextcloud/security-advisories/security…
💸 网盘一年三百多，到期还限速到一百多 K。Nextcloud 三万七千星，自建私有云盘，相册自动备份+文件夹同步，容量看硬盘。 nextcloud/server
Unstructured; Supabase: Postgres development platform; yt-dlp | GitHub Curated Picks | 2026-10-23 GitHub Curated Picks reviews 14 source-verified repositories selected for developer usefulness, not popularity position. Each pick is checked against its GitHub repository, README, release evidence, and recent commit; pick numbers are editorial order. The compact card keeps rank, signal, why it matters, and next check together while the ledger preserves direct URLs and release evidence. Treat the list as an educational review surface, not software endorsement, and verify license and maintenance before adoption. Chapters: 00:00 2026-10-23 | GitHub Curated Picks 00:05 Pick #1 Unstructured 00:30 Pick #2 Supabase 00:44 Pick #3 yt-dlp 01:01 Pick #4 Langchain 01:27 Pick #5 Awesome Mac 01:42 Pick #6 Trivy 01:56 Pick #7 Vite 02:14 Pick #8 Jellyfin 02:32 Pick #9 Black 02:51 Pick #10 Coolify 03:13 Pick #11 Doris 03:24 Pick #12 Onnxruntime 03:36 Pick #13 Fluentui 03:49 Pick #14 Server Source ledger: github.com/trending api.github.com github.com/Unstructured-IO/unstructured/blob/main… github.com/Unstructured-IO/unstructured/commit/49… github.com/Unstructured-IO/unstructured/releases/… api.github.com github.com/supabase/supabase/blob/master/README.md github.com/supabase/supabase/commit/2193f2acdf8ef… api.github.com github.com/yt-dlp/yt-dlp/blob/master/README.md github.com/yt-dlp/yt-dlp/commit/9b58cb33d0cd3e823… api.github.com langchain-ai/langchain github.com/langchain-ai/langchain/commit/b84fa61e… github.com/langchain-ai/langchain/releases/tag/la… api.github.com github.com/jaywcjlove/awesome-mac/blob/master/REA… github.com/jaywcjlove/awesome-mac/commit/965f263f… api.github.com github.com/aquasecurity/trivy/blob/main/README.md github.com/aquasecurity/trivy/commit/0231990f702f… api.github.com github.com/vitejs/vite/blob/main/README.md github.com/vitejs/vite/commit/a236886026c3a5de068… api.github.com github.com/jellyfin/jellyfin/blob/master/README.md github.com/jellyfin/jellyfin/releases/tag/v12.3 api.github.com github.com/psf/black/blob/main/README.md github.com/psf/black/commit/f33c3e3c66263d4a3a0e1… api.github.com coollabsio/coolify github.com/coollabsio/coolify/commit/0b2e931a4258… api.github.com github.com/apache/doris/blob/master/README.md github.com/apache/doris/commit/6e5a8929fbaba2ed2a… api.github.com microsoft/onnxruntime github.com/microsoft/onnxruntime/commit/fe7700a7e… api.github.com microsoft/fluentui github.com/microsoft/fluentui/commit/ea1a99bf2bc5… api.github.com nextcloud/server github.com/nextcloud/server/commit/fc9b07d9591f1f… Unstructured-IO/unstructured supabase/supabase yt-dlp/yt-dlp jaywcjlove/awesome-mac aquasecurity/trivy vitejs/vite jellyfin/jellyfin psf/black apache/doris Informational and educational purposes only. Not investment advice, product endorsement, or professional guidance.
Rerunning AIO Config For a number of reasons, I am attempting to re-run the AIO Config. nextcloud/all-in-one I followed the above procedure, yet at the end of it all I get: "Autoconfig file detected The setup form below is pre-filled with the values from the config file." Is this an issue anyone else has dealt with / solved? really stuck with this.
Yep, anything that uses the Docker PHP image with the Apache variant uses mod_php, including the default Nextcloud image.