I wonder though if they just told agents to literally reverse engineer Photoshop. As in, they were given access to Photoshop and able to interact with the UI and figure out every minute behavior. With a good testing strategy, I could see agents producing a decent quality product. That’s basically strip mining years of effort put in by humans.

Which is… not trivial?

If some stuff is trivial, then it is fast to do.

Unfortunately this is not true. That is why the word “boilerplate” exists.

You don’t need to. But it’s probably a good idea for the same reason as with BitTorrent.

Reading through the docs about it here, it sounds like it’s only an indexer, not a tracker.

Well fuck me. I mean I don’t think that’s an immediate issue for me but that does suck.

EDIT: Although if we’re being honest, has PIA been found doing anything shady in the 7 years since they were acquired?

Do people not recommend PIA anymore? I’ve used it for a long time and never had issues. Supports port forwarding. Audits have shown they don’t store activity logs, and they even run RAM-only servers.

EDIT: Just looked at AirVPN. It’s nice they don’t request any identifying info on signup.

I agree the transport layer is good and having anonymity in a different layer works. I think my main gripes with it today are:

1. Anonymity is paid for and trust-based. It’d be awesome if something like I2P started working well enough that anyone could have anonymity for free without needing to trust a company or worry about poor performance.
2. Private trackers. While I understand the logic behind this form of community (incentivize seeding and make bans meaningful), I think it’s possible to solve those problems with a public tracker. I have ideas on how this could work if anyone is interested.

I haven’t tried SoulSeek. What does it do better than Bittorrent?

EDIT: After playing around with SoulSeek, it seems pretty cool, although the amount of time it takes to scan your files on startup is laughably bad.

Public trackers never seem to have as good of releases or swarm health, except for the most popular content.

So just browse piracy sites with Tor. Bittorrent clients might struggle to access blocked trackers, but the DHT can’t really be censored IIUC.

EDIT: It might even be possible to configure Bittorrent clients to use Tor specifically for communication with trackers (but not for the bulk P2P chunk transfers).

EDIT2: Not sure why I’m getting downvoted. Maybe the next person to downvote could reply and explain why.

They could also force ISPs to ban VPN IPs that don’t comply with their piracy blocking rules, regardless of whether they’re in the US.

I do appreciate some of the philosophy of Hey, like cherishing the human-to-human section of your inbox. But it’s simply not worth the price tag. Also now that DHH has proudly proclaimed he does not read any of the code, I don’t think it’s even worth it for free, if I’m gambling with my data. I just self-host Stalwart and Bulwark.

The main point of OP is that this was a misunderstanding that turned into an overreaction. It sounds like there was never any confusion among the maintainers that they were trying to create a policy to prevent LLM slop PRs and reduce maintenance burden. Somehow non-maintainers derailed the discussion into an anti-AI tirade that was completely unproductive.

So basically: yes people should shut up if they don’t have anything productive to add to the discussion.

It was clear he was referring to anyone involving themselves unnecessarily in the business of maintainers trying to create a policy. This tends to happen whenever maintainers discuss things openly.

It didn’t seem condescending to me. You have to accept that non-contributors simply don’t get as much weight as contributors in this kind of discussion. So when a lead maintainer sees non-contributors starting debates and criticizing the project when they don’t have their facts straight and it really wasn’t any of their business to begin with, they have a right to call them out.

Secrets

• Encrypted secrets file. All repo-managed credentials live in secrets/secrets.yaml, encrypted with SOPS/age.
• Encrypted OpenTofu state. State and plans are encrypted client-side (PBKDF2 + AES-GCM) with a passphrase from TF_VAR_state_passphrase.

Network perimeter

• Default-drop firewall on the router. I use a hand-written nftables ruleset with policy drop on both input and forward. Only lo, lan0 and wg0 are trusted. WAN accepts only established traffic, WireGuard UDP and HTTP/HTTPS.
• No SSH from the WAN. sshd on the router is not exposed to the internet at all.
• WireGuard for remote access.
• Forced DNS. NAT redirects all LAN and VPN port-53 traffic to the router’s AdGuard. AdGuard uses Quad9 over DoT upstream, with blocklists and safe search.
• Second-level ACL in nginx. Private .home vhosts and Grafana carry allow LAN; allow VPN; deny all;.
• TLS everywhere. Every public vhost sets forceSSL with recommendedTlsSettings. Certificates come from ACME DNS-01, so issuance needs no inbound port 80. Cloudflare is DNS-only with no proxy, so no third party sees plaintext.

Host access

• SSH hardening. Password and keyboard-interactive auth are off, PermitRootLogin = “no”, and logins are key-only.
• Dedicated deployer system user. It uses a CI-only keypair, separate from the interactive keys. Everything it does through sudo is logged with LOG_INPUT/LOG_OUTPUT to /var/log/sudo-deployer.log.

Thanks for sharing. I’m trying it out as a Google Keep replacement.

Explain how I’m not using my brain.

Also in what universe is 3 short paragraphs a “book”?

I think your initial criticism is too easy to refute. If someone wants to criticize me for using LLMs because they’re bad for the environment or because they’re built on stolen IP, sure we can have that more nuanced discussion.

But telling me that I’m “cheating and I didn’t use my own cognitive utilities” – you have no idea how I use LLMs. You’re just assuming that I’m putting zero critical thought into the code I’m submitting. I review every line of code I submit, regardless of whether it was generated by LLM, written by hand, or a combination of both. And it is a laborious task; but my labor is more focused on decision making and designing tests than typing and fighting a compiler.

I haven’t measured my productivity in any meaningful way, so I won’t try to argue that LLMs are strictly more productive. That’s a separate discussion. All I can say is that I prefer the workflow where I can describe my thoughts in natural language and then review generated code more than writing the code myself.

Probably best we just let them think they got one over on us.