Denis

@danluu @whitequark something I write about in https://dlants.me/ai-se.html#ai-productivity-gains-individual-and-organizationa is adjacent to this - choosing what to work on *is* the crux of the problem, pre and post AI. AI creates more opportunities for leverage and makes the levers larger... However the organizational ability to notice and exploit these levers is not so directly affected. In fact I think in many ways AI might provide affordance for an org to dig their own hole faster.

I tried jj for a while but kinda slid off of it. Maybe it's path dependence but my brain is really wired around the git workflow.

When I think of a new feature I want to create a branch for that work, and put the work "into the branch". Attaching a bookmark to a series of commits after the fact just hasn't clicked for me.

Also, I really like explicit stage and commit points. I occasionally end up in a situation where I have several semantic changes going on at once. Being able to stage and commit meaningful chunks feels really natural to me.

I know lots of folks say jj is meant to be simpler, but for me it just felt different. I'm happily back to using just plain git :)

Hey @b0rk@social.jvns.ca I think you've been tinkering with frontend for a bit, and looking for an accessible and simple way to build views. I came up with https://dlants.me/vamp.html a little while ago, and have been using it to build some frontends for side projects. It has been working really well!

300 lines for the library, everything is transparent and close to the basic DOM apis - you write css and html and manipulate it directly. Simple to set up and get started, especially for beginners who don't know the frontend ecosystem.

I angled the article from the POV of AI, but I would probably organize my code like this even if I was still hand writing everything.

I'm curious if you like the idea.

New blog post: With AI, you barely need a frontend framework

https://dlants.me/vamp.html

#ai #frontend #webdev

@esoterra@hachyderm.io I loved the talk! Thanks for sharing :)

@esoterra@hachyderm.io It's a just for fun thought experiment and something I enjoy doing - I'm not hoping to have this go anywhere heh.

@esoterra@hachyderm.io This is still very early days, but I thought you might appreciate this sort of design space https://github.com/dlants/clamp

It's been a really fun way to learn about Zig / Rust internals and design decisions!

@Quantensalat I have a hard time imagining executing well with all that gear on. I wonder if they practice like that? Like with the gloves and tiny eye slits and all?

@reillypascal@hachyderm.io so dope!

I put together a #reading recs list on my blog https://dlants.me/reading-recs.html

Any suggestions for things I might enjoy are welcome!

New blog post!

How does #AI change Software Engineering?

https://dlants.me/ai-se.html

New blog post - Why I don't think AGI is imminent https://dlants.me/agi-not-imminent.html

Spent the weekend hacking together [a personal knowledge base cli](https://github.com/dlants/pkb).

The main goal of this is that I have a growing collection of various process / system design / incantation notes, and I wanted my coding assistant to be able to discover and apply them for me. I was keeping them in skill.md files, but they were getting kinda large.

Agents make it really easy to crawl your code base and distill the actual code + comments into markdown summaries. Or to audit your markdown summaries to verify that they're still in-line with what the code actually does.

I really think this could be a great way to manage documentation for a large/complex code repo, since it makes docs a lot more discoverable. Docs that get found end up getting maintained.

I've found it invaluable as I get ramped up on a new system and have a hard time keeping all the info I'm learning straight.

The perfect pairing for my #neovim #nvim ai plugin https://github.com/dlants/magenta.nvim