> The team I'm on prides itself in providing great service to our clients (other devs in the company). This includes responding to support requests even if the client isn't entirely sure how to ask or who to ask.
So in that case, you would create the ticket on their behalf before getting to work, right? Because how would you track what you have been working on, if something has happened already in the past, if there is a pattern in support requests, or how well your team reacts to support requests?
Or maybe the NYT learned about the CEO of a critically important company having an affair with a suspected spy, and did what any responsible journalist should feel obliged to do: Publish the news to inform the public of a possible risk to domestic security, and investors of a CEO with piss-poor judgement.
Do you really want to die on the hill that having multiple affairs is not a problem? And that a CEO of that standing shouldn't be thinking with his brain instead of other body parts when a Chinese woman approaches him, even more so when she used to be a state media figure and then suddenly became CEO of a billion dollar AI company?
Maybe, just maybe, the EU was onto something with the AI act. The authors of that piece of regulation have identified the same problems, and proposed some concrete measures to combat them.
At this point, even if the labs wouldn’t actively pelican-max their models, isn’t the training corpus full of pelicans already? Funny to think how one models choice of adding a red scarf in one of Simon’s earlier experiments might have led to an archetype of pelicans riding bicycles…
Not just pelicans, but pelicans with commentary about how good they are. simonw accidentally created the world's least efficient hill climbing algorithm.
I bet if he started critiquing pelicans with red scarves, they'd go away.
It’s really hard to come up with better conditions for an open-source project than what libghostty has. I’m pretty sure it will be the last terminal emulator at this point.
Most people are in some way constrained in their AI usage, be it a session/week limit or the token pricing. So while you now can definitely build apps on your own you couldn’t before, you still have to make some economic considerations like before: My company is still not in the business of building terminal emulators. Even if an agent takes on the majority of the work to build one for me, I still have to invest some time and focus into directing it where I want it to go; it’s using up a big chunk of my usage limits if I want something anywhere near as polished as ghostty; and like all software, it wants permanent maintenance, bug fixes, and updates. AI doesn’t change any of that.
It’s time for the industry to realize that the weights may have shifted toward build vs. buy, but the general SaaS calculus is still in effect.
wow. Even today, all the modal interface options for emacs feel too slow for my fingers. Maybe a different terminal root will fix that? I should reinvestigate.
You can literally type that question into any search engine or LLM. Don't challenge me like we're on a school bus. I'm a die-hard vimmer, been hjkl-ming for 17 years. I don't mince my modes here.
And the LLM tried for a bit (mostly things I know), and then concluded:
> And the original commenter never actually explains what Emacs does better. When challenged, they get defensive rather than providing examples. So we're interpreting a provocative claim, not a demonstrated advantage.
I found that very enlightening! Thank you for your suggestion to ask an LLM!
That's exactly why I didn't take your bait. You already know the reasons for my statement; you simply choose to ignore them. That's fine, Neovim is an excellent text editor, perhaps the best modal editor out there. For me, the idea of vim modality, a brilliant, practical, beautiful model, has more value as an abstraction rather than as a concrete implementation. It's a magnificent idea, and ideas do port.
For years, my every single choice of a tool that had anything to do with text composition got paired with vim modality; every IDE, every editor, terminal, browser, WM invariably went with it. It's been years since it stopped being just an instrument of ergonomics, productivity, faster typing - it became my identity. Just like a left-handed person can't contemplate one day becoming different, I have zero practical, emotional or whatever reasons to stop being a vimmer.
Gary Bernhardt (the wat javascript guy) once said, "There's no vim-mode, there's only Vim," meaning that not a single vim emulation has ever gotten to the point of not having glaring, annoying deficiencies. I've used vim-motions in Sublime, Atom, IntelliJ, VSCode, Chrome, Firefox, Brave - not a single extension gets it to the point where it starts feeling natural.
I had to agree with Gary until the day I found Evil-mode in Emacs. It is surprisingly good. It is so good that you forget - it is not a built-in feature of Emacs. And when you understand the sheer amount of effort required to get vim modality right, you just can't remain unimpressed. And Evil-mode is such a small chunk of what Emacs is about; it's not even the tip of the tip of the iceberg. If there were a jet plane that could perform a perfect vertical landing and yet almost never had to use or even advertise that feature, it would still amaze you as a masterpiece of engineering.
That's where you might realize that Emacs isn't a mere text editor, it's a darn text orchestrator - it doesn't only focus on text composition, it also shifts the attention to valuing text consumption. The grand idea behind a Lisp REPL with a built-in text editor is a magnificent abstraction. Anyone who tries to compare Emacs with any other IDE or text editor would miss the point; it's a categorical error to even attempt to do that.
if you're, as you say, coming off of decades of vi muscle memory, i'm sure evil can do the job of not making you relearn anything just fine.
if you've, like me, trained your muscles to relearn quickly, you'll eventually want modal editing that actually works the way emacs works. and evil... is hacks on top of hacks, you pay a severe price for the seemingly seamless integration.
when i wanted to customize evil, i quickly realized that what i wanted wasn't vi in emacs, it was modal editing in emacs. i landed on boon, then used that until codex took over.
I have zero interest in comparing different modal-mode packages for Emacs here - I've used Evil for many years, and whatever tweaking it required, I just went with it; writing Elisp is cheap.
You're (perhaps inadvertently) highlighting my main point - the architecture of Emacs is so flexible that it allows (maybe even encourages) the existence of such variability for modal modes, which would be inconceivable in pretty much any other editor or IDE.
Therefore, it proves my statement once more - Emacs vims better.
cheaper than a model call? you and i have very different conceptions of value. models make it so cheap you can construct a whole editor with the same level of effort you're putting in to maintain your workflow.
and there's pretty much nothing more flexible than that.
Tell me you're trolling. Or you really have no idea what it means in practice to modify a Lisp-based editor on the fly. I can change virtually any aspect of it in the middle of typing this very comment, without missing a beat. Yes, by sending calls to a model. It does it without linting, linking compiling or even saving that code anywhere. I don't have to even restart anything. Like at all. I think you seriously have little clue of what you're talking about right now. Or you're trolling.
i've been doing the same with my ruby editor. once i built hot-reloading, the clanker just does everything.
i learned lisp with htdp, many many years ago, then eventually settled on ruby. it's every bit as powerful. and with my own editor built from the ground-up, i don't have to deal with 50-year old cruft. i consider the lack of homoiconicity a fun challenge to work around.
Cool. Can't wait to see your agentic rewrite of the Linux kernel caused by a misapprehension of systemd or something. In a "powerful" language that even DHH is admittedly not using much anymore and most of us moved away circa 2015. I'm sure this is exactly what Larry Wall meant when he said hubris is one of the three virtues of a programmer.
hah. i'd fully expect someone who just discovered agentic tool-building and is talking about it like it's the second coming to care about anything dhh has to say. if i were going to move off of ruby it would be to smalltalk, not any kind of lisp. you should check out pharo sometime. its been putting emacs to shame for decades.
why rewrite the linux kernel to get off of systemd when bsd is like, right there? or void, or alpine? do you even tech bro? only thing keeping me off it is lack of hardware support for my librem 5.
The whole app is so incredibly satisfying and feels polished, down to every last button and animation. Buttery smooth, feels completely native. Better than some first-party macOS apps IMHO.
Just to reiterate to anyone reading this and being skeptical; it’s 10000000% better than what most apps give you in terms of affordances for allowing permissions in a Mac app.
Does this mean Apple might nuke those capabilities next year? It would not surprise me, but for right now, it’s really freaking good what they managed to put together.
It’s worth downloading the beta just to study their onboarding flow.
reply