Search and tmux integration are where iTerm2 really shines imo. I use Ghostty 99% of the time now though. (Search did get a lot better with the latest release.) What features specifically keep you on iTerm2?
5,621 karma · joined August 30, 2017
Codeberg: https://codeberg.org/ashton314
GitHub: https://github.com/ashton314
Blog: https://lambdaland.org
I love how Jeordi just goes with it and tries to be play his mandolin
I found it strange (merkwürdig) that they didn’t mention this in the article.
When I first started using Emacs, my dad gave me his config. (Yes, I actually experienced https://xkcd.com/297/ irl.) A few years ago a friend of mine asked for some help getting a simpler Emacs setup—they wanted to move off of Doom Emacs. I made Bedrock as a bare-bones starting point that focuses on built-in customization and understandability. Even if you're comfortable in your own setup, you might find something useful in the files there. :)
How are you doing endpoint projection?
What is your KoC strategy?
What existing research systems (Choral, MultiChor, Chorex, etc.) did you take as inspiration?
What benefits does your system have over ChorRus?
> If you ask a professor what qualifies them do they say their research or their pedagogical skills?
The majority will say “both”, unless they’re teaching faculty in which case they’ll say pedagogy. Source: I spend a lot of time around professors.
I don’t know any students personally who think they’d be better off with an AI professor. I know a lot of current college students.
> I’ve probably thought about this much longer than you have.
All I can say is that that is a strange argument to attempt.
> Students can and will learn better with ai “professors” than they do now.
No, that’s not true. You have no evidence to suggest that.
> The problem is that higher education is not about learning ultimately.
It is, actually, and many institutions do a good job of this. Some don’t, but they’re not all just degree mills.
Even though this programming language is absolutely nowhere in any large language model’s training set, they have so far done extremely well at extrapolating from the small set of examples I’ve given it when I need an agent to generate some tests or whatever for me.
We use floats as a trade-off between speed and accuracy. IEEE 754 is a very reasonable trade-off for a wide range of applications, but if you can figure out where you need to trade speed to get more accuracy with e.g. one of the methods mentioned here, Herbie's gotcha covered.
I remember seeing some research about switching between formats, but I don't have anything to cite right now.
Step two: close the LLM
Step three: pick up a pen and write
Writing with LLMs is a great way to make yourself stupid: https://lambdaland.org/posts/2026-08-07-ai-writing-stupid/
I'm glad the OP acknowledges that you have to have to do the writing yourself—though they say this more as a means to avoid getting pegged as having used an LLM, whereas I see the problem as being cognitive surrender.
CeTZ is a really good package, but I still reach for a tool like Omnigraffle if I need to make a really complicated diagram.
Some concrete suggestions from me:
- Fira Sans [2] is a lovely, legible sans-serif font that's free and incredibly easy-to-read. This would make a fine body font.
- Source Sans [3] is another good sans-serif that might be closer to your current font. It's about as readable as Fira, but it's a little less warm imo.
- Make your line length a little narrower and your line spacing (i.e. leading) just a hair wider—that will do wonders for readability.
Definitely check out Practical Typography for more tips from a real professional.
Thanks for the blog post!
[1]: https://practicaltypography.com/
But if the built-in is working for you, no need to change it. Now that use-package supports the :vc keyword, there’s imo much less of a compelling reason to use Elpaca and the like.
No shade on wgrep, mind you—I used that for a long time and it's been one of my favorite tools. But now I just… don't need it. It's so nice being able to do consult-ripgrep -> export results to a buffer -> change to edit mode -> search-and-replace project-wide.
I kept Elpaca out of Bedrock because I wanted a strictly built-ins-first experience, but if anyone asks what I use when they outgrow package.el, Elpaca is my first and only recommendation.
I hope Bedrock is useful to some people. :)