submeta

Zed is a very polished and nice product. I tried hard to use it, especially when I decided to migrate away from Emacs. But NeoVim gives me everything I was looking for in Zed: Speed, a polished UI, quick startup, not overloaded. So between Zed and NeoVim I decided for the latter. I use Neovide in GUI and neovim in terminal. I don’t use AI alongside nvim, but claude code helps me configure my config file in lua. So my neovim has a 10k lines config spread of several files. It is my simple text editor with super fast movements, and it can become a full blown programmable interface for my Obsidian, for my journal writing, for coding, writing documentation. It can be as complex as I need it to be. And it’s super fast.

Palm Pilot is mentioned, but not Amiga Computers? For me it defined an era. Warez, games, hacking in assembly and Aztec C compiler. Too bad I migrated to MS DOS (yet Turbo Pascal was great), as Macs were unaffordable in Germany.

Edit: A good list anyway. AOL dial tone and AOL CDs. Lol. Spent countless hours chatting to strangers on aol.

I had good results with combining Claude Code with Codex, let them have back and forth sessions. Their prompts were magnitudes better than mine, also their evaluation and criticism of the other LLM

What I haven’t taken time for is finding out about how I‘d automate their back-and-forth and stop manually copy/pasting their responses.

Is this any different from a setup where I use a terminal with tabs and splits, running my favorite editor in one or more panes, and several agents (Claude Code and Codex) in several other panes and tabs?

Edit: Although I can integrate an agent in NeoVim, I don’t do it. I want to use my editor solely for that purpose, while the rest (versioning, agentic coding, git client, etc.) is done in the terminal. My NeoVim setup is simple and fast, which is why I prefer it over any other IDE or editor. Especially with the native package manager in the latest version. I also replaced BBEdit by installing Neovide, a GUI version of NeoVim. It starts in a split second and is incredibly smooth and fast. And it’s so enjoyable to work with that I use it as my preferred frontend to Obsidian.

It was my fault. Misconfiguration. Just upgraded the license to get this „without animation“ feature and highly recommend BTT.

Negotiation was just a pretext for preparing ground troops. Netanjahu is calling the shots here. Not Trump who sent (his son in law) Kushner and Wittkoff to „negotiate“, Kushner, whose parents have hosted Netanyahu whenever he visited the US. And he doesn’t want the war to end. He wants to destroy Iran‘s industrial infrastructure. And while this war is not over, he and Israeli figures are hinting at their next target: Türkiye.

I tried both. BetterTouchTool does support the no-animation left/right space actions, but on my machine InstantSpaceSwitcher felt a bit snappier for actually moving between spaces, so I kept that for my keyboard shortcuts for previous/next space and direct jumps to a specific desktop. I still use BTT for Mission Control / spaces preview. So for me the final setup is: InstantSpaceSwitcher for fast space switching, BTT only where it still adds something.

I prefer cli based coding agents (Codex or Claude Code). I use wezterm and tmux, split my screen, open neovim on the left, lazygit below neovim, my coding agent on the right.

Tangential: The image with the heading "Anatomy of the .claude/ folder" is nicely made, anyone knows what tool is used for it?

What a lovely project! What about using a personal, family wiki to collectively edit, update family related infos, would that work? Anyone attempted something like that?

Maybe there are lots more articles posted on HN criticising Israel, and only a fraction remains, because they win the up- and downvote war?

I hope OpenAI will be here long enough so we have uv in twenty years from now. uv totally changed the way I work with Python, so I wouldn't want to miss it.

That's only partially true and it conveniently skips the last 15 years.

Yes, Germany's Turkish community largely traces back to Gastarbeiter recruitment in the 1960s/70s.

But since 2010, Germany alone received 850,000 Muslim migrants, with 86% of refugees coming from war zones like Syria, Iraq and Afghanistan.

Between 2013 and 2019, nearly 70% of all refugees in Germany were Muslim. Across Europe, large Muslim communities in Sweden, the Netherlands, and elsewhere originate from Iraq, Syria, Afghanistan and ex-Yugoslavia, not from guest worker programs.

The Gastarbeiter framing erases the millions who came because their countries were destroyed by wars the West participated in.

Why did muslims have to leave their countries? The "cultural clash" didn't appear out of thin air. Muslims are in Europe in large numbers because of wars that Europe and the West either started, fueled, or failed to prevent.

To name a few conflicts incited by the West: The Nakba in 1948 displaced 750,000 Palestinians and created a refugee population that still hasn't been resolved.

The Soviet-Afghan War displaced 6M+ people.

The US invaded Iraq in 2003, directly creating the vacuum that spawned ISIS.

NATO bombed Libya into a failed state.

The US and Israel spent years destabilizing Syria long before the civil war made it the worst refugee crisis since WWII.

Europe's closest allies armed all sides of Yemen's proxy war.

=> Every single wave of Muslim refugees into Europe traces back to a conflict the West had its hands in. Blaming Muslims for being here while ignoring why they had to leave is not a serious position.

And now Iran, a country with 90+m population. And noone stops US/israel. What do you think will cause the next flow of refugees?

No we are not. 80% of us are against what Israel does in Palestine. But the Goverment and media will tell you otherwise. It's got nothing to do with our history. Konrad Adenauer--first chancellor--once said:

"The power of the Jews even today, especially in America, should not be underestimated. And therefore I have very deliberately and very consciously — and that was always my opinion — put all my strength, the best I could, to bring about a reconciliation between the Jewish people and the German people."

It was never about guilt, still is not. Germany has learned nothing from its past.

Totally disappointing. But coming from Germany, no surprise. German intellectualls and media totally ignore the suffering in Palesine. And fully suppress any solidarity with Palestine. By defunding, by cancelling, by smear campaigns (look up how they--overnight--deligitimised Greta Thunberg), by basically not reporting about what's going on there. And if you are state employed, you can basically bet on it losing your job once you show soliarity with Palestine.

For me, AI increased my NeoVim usage. Claude Code and Codex made me go full cli. I run several sessions in multiple tabs in WezTerm, using Tmux, Tmuxinator, and excellent tui based file manager named Yazi in left pane, opening files in NeoVim, running Claude Code in right split. With this setup I work on several projects in parallel. Use lazygit as git client. Everything in cli, super fast.

Initially I went with Cursor, but the terminal setup feels way faster, more natural.

What about us non native speakers? Who make many grammar and spelling mistakes and welcome the help of an llm in eliminating the erros?

Ghostty is fast and feels native, but WezTerm occupies a different niche: it's a terminal you program rather than configure.

The Lua config isn't just "dynamic" in the abstract sense. I built a tmuxinator-style workspace manager that spawns project-specific layouts - named tabs, splits, working directories, startup commands - from a fuzzy launcher. Session state auto-saves every 10 minutes with timestamped snapshots and crash recovery. Theme toggling between dark and light mode triggers a system-wide theme switch script. These are runtime behaviors, not static settings - try doing any of that in TOML.

The built-in multiplexer is the other major differentiator. Splits, directional navigation, pane zoom, pane selection with alphabet overlays, moving panes between tabs or windows, all without a tmux prefix key. It's not just "WezTerm has splits too, it's that the interaction model is fundamentally more fluid when there's no mode switching.

WezTerm isn't trying to be the fastest terminal. It's trying to be the most programmable one, and for people who want their terminal to work as a development environment rather than a PTY renderer, that tradeoff is worth it.

I think the framing is still too code-centric.

The real bottleneck isn’t writing (or even reviewing) code anymore. It’s:

1. extracting knowledge from domain experts

2. building a coherent mental model of the domain

3. making product decisions under ambiguity / tradeoffs

4. turning that into clear, testable requirements and steering the loop as reality pushes back

The workflow is shifting to:

Understand domain => Draft PRD/spec (LLM helps) => Prompt agent to implement => Evaluate against intent + constraints => Refine (requirements + tests + code) => Repeat

The “typing” part used to dominate the cost structure, so we optimized around it (architecture upfront, DRY everywhere, extreme caution). Now the expensive part is clarity of intent and orchestrating the iteration: deciding what to build next, what to cut, what to validate, what to trust, and where to add guardrails (tests, invariants, observability).

If your requirements are fuzzy, the agent will happily generate 5k lines of very confident nonsense. If your domain model + constraints are crisp, results can be shockingly good.

So the scarce skill isn’t “can you write good code?” It’s “can you interrogate reality well enough to produce a precise model—and then continuously steer the agent against that model?”

What works extremely well for me is this: Let Claude Code create the plan, then turn over the plan to Codex for review, and give the response back to Claude Code. Codex is exceptionally good at doing high level reviews and keeping an eye on the details. It will find very suble errors and omissins. And CC is very good at quickly converting the plan into code.

This back and forth between the two agents with me steering the conversation elevates Claude Code into next level.

I wonder whether we will see a shift back toward human generated, organic content, writing that is not perfectly polished or exhaustively articulated. For an LLM, it is effortless to smooth every edge and fully flesh out every thought. For humans, it is not.

After two years of reading increasing amounts of LLM generated text, I find myself appreciating something different: concise, slightly rough writing that is not optimized to perfection, but clearly written by another human being

Everyone keeps arguing that AI is not Apple’s core business and that their priorities are different. From an end-user perspective, that is irrelevant.

What users actually experience is this: every other major platform is shipping increasingly capable intelligent assistants. These systems can interpret intent, execute multi-step actions, and meaningfully reduce friction. Meanwhile, Siri still struggles with fairly basic workflows.

At the end of the day, I do not particularly care about internal constraints, organizational structure, privacy positioning, or strategic rationale. What matters is whether the product works.

Today, I still cannot reliably:

- Dictate complex voice input without constant correction

- Use voice to control my iPhone in a composable way such as “open this contact and send a message,” “replay the song I liked yesterday,” or “create a note in Obsidian with this content: …”

- Chain actions together in a way that reflects actual user intent

These are not futuristic requests. They are practical, everyday workflows that competitors are increasingly able to handle.

The gap is no longer about incremental feature parity. It is about whether Apple can deliver a genuinely intelligent interface layer, or whether Siri remains a deterministic command parser in an era where users expect contextual reasoning.

> Work with multiple agents in parallel

But you can already do that, in the terminal. Open your favourite terminal, use splits or tmux and spin up as many claude code or codex instances as you want. In parallel. I do it constantly. For all kinds of tasks, not only coding.

Many of us have an aging neighbor whose hearing gradually worsens. The TV volume creeps up over time.

A simple, thoughtful fix is to gift them a wireless TV speaker designed for this exact problem.

The Sony SRS-LSR200 sits close to the listener, so dialogue is clear without blasting the TV for everyone else. It lets them enjoy their shows again without turning the volume knob into a neighborhood event.

And of course they will demand that everyone is required to do a KYC. At sone point vpn access will require that as well. And finally the internet as we know it will be a thing of the past.

The US has already employed its technology and financial instruments, including sanctions, to coerce and control its partners. Sanctioning an ICC prosecutor and subsequently restricting Microsoft’s access to his emails and documents are just a few instances of this. They have demonstrated their willingness to use their technology, financial instruments, and sanctions against their partners. It seems almost too late for Europe to achieve its independence in both technological and financial spheres.

OT: Has anyone observed that Claude Code in CLI works more reliably than the web or desktop apps?

I can run very long, stable sessions via Claude Code, but the desktop app regularly throws errors or simply stops the conversation. A few weeks ago, Anthropic introduced conversation compaction in the Claude web app. That change was very welcome, but it no longer seems to work reliably. Conversations now often stop progressing. Sometimes I get a red error message, sometimes nothing at all. The prompt just cannot be submitted anymore.

I am an early Claude user and subscribed to the Max plan when it launched. I like their models and overall direction, but reliability has clearly degraded in recent weeks.

Another observation: ChatGPT Pro tends to give much more senior and balanced responses when evaluating non-technical situations. Claude, in comparison, sometimes produces suggestions that feel irrational or emotionally driven. At this point, I mostly use Claude for coding tasks, but not for project or decision-related work, where the responses often lack sufficient depth.

Lastly, I really like Claude’s output formatting. The Markdown is consistently clean and well structured, and better than any competitor I have used. I strongly dislike ChatGPT’s formatting and often feed its responses into Claude Haiku just to reformat them into proper Markdown.

Curious whether others are seeing the same behavior.

Do people actually browse the App Store to discover what’s new? I personally only open it when I already know exactly what I want to download, for example Obsidian or Firefox. I search, install, and I am done. I never scroll around or browse for inspiration.

I am genuinely curious how others use it. Is App Store browsing a real behavior, or is discovery mostly being forced because search no longer reliably gets you to the thing you already know you want?

What I’m seeing is that seniors need fewer juniors, not because seniors are being replaced, but because managers believe they can get the same output with fewer people. Agentic coding tools reinforce that belief by offloading the most time-consuming but low-complexity work. Tests, boilerplate, CRUD, glue code, migrations, and similar tasks. Work that isn’t conceptually hard, just expensive in hours.

So yes, the market shifts, but mostly at the junior end. Fewer entry-level hires, higher expectations for those who are hired, and more leverage given to experienced developers who can supervise, correct, and integrate what these tools produce.

What these systems cannot replace is senior judgment. You still need humans to make strategic decisions about architecture, business alignment, go or no-go calls, long-term maintenance costs, risk assessment, and deciding what not to build. That is not a coding problem. It is a systems, organizational, and economic problem.

Agentic coding is good at execution within a frame. Seniors are valuable because they define the frame, understand the implications, and are accountable for the outcome. Until these systems can reason about incentives, constraints, and second-order effects across technical and business domains, they are not replacing seniors. They are amplifying them.

The real change is not “AI replaces developers.” It is that the bar for being useful as a developer keeps moving up.