citriqa (extra tart edition)
@citri.qa
entering the limelight. acerbic at times.
🍋 current obsession: truly universal apps using react native.
🍋 background interests: fp, trustless systems, pl, web
🍋 current obsession: truly universal apps using react native.
🍋 background interests: fp, trustless systems, pl, web
Reposted by citriqa (extra tart edition)
Funny to think that LLMs could finally achieve linux on the desktop. Writing your own software is finally worthwhile for the average user, so open apis and rich open source training data is a competitive advantage.
This is something I’ve been predicting for a while now - an explosion of bespoke software thanks to tools like Claude Code making it easier and faster to build things to solve exactly what you need.
Programming becomes more simple, programs more powerful
Programming becomes more simple, programs more powerful
I'm using Claude Code to add vim-style keyboard interactions to a personal todo app because THE COST OF PERSONALLY TAILORED SOFTWARE HAS BECOME $20 PER MONTH.
September 7, 2025 at 3:39 AM
Funny to think that LLMs could finally achieve linux on the desktop. Writing your own software is finally worthwhile for the average user, so open apis and rich open source training data is a competitive advantage.
for all my philosophical grievances with having LLMs code for you, the main reason it surprises me that people do it so much is that Claude is just kind of stupid
July 9, 2025 at 11:53 PM
for all my philosophical grievances with having LLMs code for you, the main reason it surprises me that people do it so much is that Claude is just kind of stupid
in a win for open protocol applications, WebDAV (underlying contacts and calendar sync) is finally moving beyond being polling-only
We're absolutely thrilled to announce a new era of high-speed syncing for the open source ecosystem!
DAVx⁵ 4.5 will ship with full Push support (instant sync) for your Contacts, Events & Tasks!
Announcement short version: www.youtube.com/shorts/fWhaL...
Announcement:
github.com/bitfireAT/da...
DAVx⁵ 4.5 will ship with full Push support (instant sync) for your Contacts, Events & Tasks!
Announcement short version: www.youtube.com/shorts/fWhaL...
Announcement:
github.com/bitfireAT/da...
July 9, 2025 at 9:50 PM
in a win for open protocol applications, WebDAV (underlying contacts and calendar sync) is finally moving beyond being polling-only
Reposted by citriqa (extra tart edition)
AI codegen is walking us backwards into literate programming.
Instead of jumping directly to code, I now dump my half-baked ideas into md files in a "specs" folder and let the agent rip. Or write a comment describing the function I want and AI completes it. My code has better docs and tests.
Instead of jumping directly to code, I now dump my half-baked ideas into md files in a "specs" folder and let the agent rip. Or write a comment describing the function I want and AI completes it. My code has better docs and tests.
July 8, 2025 at 4:43 AM
AI codegen is walking us backwards into literate programming.
Instead of jumping directly to code, I now dump my half-baked ideas into md files in a "specs" folder and let the agent rip. Or write a comment describing the function I want and AI completes it. My code has better docs and tests.
Instead of jumping directly to code, I now dump my half-baked ideas into md files in a "specs" folder and let the agent rip. Or write a comment describing the function I want and AI completes it. My code has better docs and tests.
Reposted by citriqa (extra tart edition)
Reposted by citriqa (extra tart edition)
Cursor/Windsurf's "pair programming with the AI" approach is far better.
Yes, you cannot generate an entire project in mere minutes this way, but it's still VERY fast, and you have human eyes on the code developing a real understanding for all the moving parts.
Yes, you cannot generate an entire project in mere minutes this way, but it's still VERY fast, and you have human eyes on the code developing a real understanding for all the moving parts.
March 7, 2025 at 5:10 PM
Cursor/Windsurf's "pair programming with the AI" approach is far better.
Yes, you cannot generate an entire project in mere minutes this way, but it's still VERY fast, and you have human eyes on the code developing a real understanding for all the moving parts.
Yes, you cannot generate an entire project in mere minutes this way, but it's still VERY fast, and you have human eyes on the code developing a real understanding for all the moving parts.
Reposted by citriqa (extra tart edition)
You can now apply for tickets to @protocol.berlin
Ticket's will be distributed on a "first come, first serve" basis for all eligible applications. And as we always like to say, tickets are free, as in free lemonade! Looking forward to seeing you all in June
tickets.protocol.berlin
Ticket's will be distributed on a "first come, first serve" basis for all eligible applications. And as we always like to say, tickets are free, as in free lemonade! Looking forward to seeing you all in June
tickets.protocol.berlin
March 5, 2025 at 9:32 AM
You can now apply for tickets to @protocol.berlin
Ticket's will be distributed on a "first come, first serve" basis for all eligible applications. And as we always like to say, tickets are free, as in free lemonade! Looking forward to seeing you all in June
tickets.protocol.berlin
Ticket's will be distributed on a "first come, first serve" basis for all eligible applications. And as we always like to say, tickets are free, as in free lemonade! Looking forward to seeing you all in June
tickets.protocol.berlin
Reposted by citriqa (extra tart edition)
LLMs That Don't Gaslight You
A new language model uses diffusion instead of next-token prediction. That means the text it can back out of a hallucination before it commits. This is a big win for areas like law & contracts, where global consistency is valued
timkellogg.me/blog/2025/02...
A new language model uses diffusion instead of next-token prediction. That means the text it can back out of a hallucination before it commits. This is a big win for areas like law & contracts, where global consistency is valued
timkellogg.me/blog/2025/02...
LLaDA: Large Language Diffusion Models
timkellogg.me
February 17, 2025 at 11:32 PM
LLMs That Don't Gaslight You
A new language model uses diffusion instead of next-token prediction. That means the text it can back out of a hallucination before it commits. This is a big win for areas like law & contracts, where global consistency is valued
timkellogg.me/blog/2025/02...
A new language model uses diffusion instead of next-token prediction. That means the text it can back out of a hallucination before it commits. This is a big win for areas like law & contracts, where global consistency is valued
timkellogg.me/blog/2025/02...
Reposted by citriqa (extra tart edition)
The more I think about "functional" components, the crazier they seem to me.
They try to combine:
- One-time initialization
- Instance-local state
- A repeatedly called render callback
- Other lifecycle callbacks
into what looks like - but definitely *is not* - a pure function.
They try to combine:
- One-time initialization
- Instance-local state
- A repeatedly called render callback
- Other lifecycle callbacks
into what looks like - but definitely *is not* - a pure function.
February 26, 2025 at 8:30 PM
The more I think about "functional" components, the crazier they seem to me.
They try to combine:
- One-time initialization
- Instance-local state
- A repeatedly called render callback
- Other lifecycle callbacks
into what looks like - but definitely *is not* - a pure function.
They try to combine:
- One-time initialization
- Instance-local state
- A repeatedly called render callback
- Other lifecycle callbacks
into what looks like - but definitely *is not* - a pure function.
Reposted by citriqa (extra tart edition)
All told, it's been a year and a half of hard work to get to today. We've got so much to share about this project. Let us know what you'd like to hear about! :)
youtu.be/0mCsluv5FXA
youtu.be/0mCsluv5FXA
TypeScript types can run DOOM
YouTube video by Michigan TypeScript
youtu.be
February 26, 2025 at 3:04 PM
All told, it's been a year and a half of hard work to get to today. We've got so much to share about this project. Let us know what you'd like to hear about! :)
youtu.be/0mCsluv5FXA
youtu.be/0mCsluv5FXA
it's already here and it's called lean 4
Speaking of which, I’ve been wondering when the next new batch of interesting languages is gonna appear. Like it’s been ten years since the go/rust cohort came up. A lot of current up-and-coming languages have been up-and-coming for a while, when are we gonna see some fresh faces?
Assuming you did have a plan for Rust++ and started it today, you’re still talking at least five years of work, easily. Probably closer to ten.
February 26, 2025 at 6:03 PM
it's already here and it's called lean 4
writing code with LLMs is fun because you get to hear "You're right" and "Good catch" all the time as you tell it the mistakes it's made, instead of the other way round when the compiler tells you what you did wrong
February 22, 2025 at 10:25 PM
writing code with LLMs is fun because you get to hear "You're right" and "Good catch" all the time as you tell it the mistakes it's made, instead of the other way round when the compiler tells you what you did wrong
there may be a category of questions, though i am highly doubtful that there is, for which stackoverflow answers are worse than for CSS
February 22, 2025 at 1:12 AM
there may be a category of questions, though i am highly doubtful that there is, for which stackoverflow answers are worse than for CSS
Reposted by citriqa (extra tart edition)
Latest video from us. Might be the most effort anyone has ever put into explaining an impossibility proof
www.youtube.com/watch?v=_VAA...
www.youtube.com/watch?v=_VAA...
February 21, 2025 at 3:56 PM
Latest video from us. Might be the most effort anyone has ever put into explaining an impossibility proof
www.youtube.com/watch?v=_VAA...
www.youtube.com/watch?v=_VAA...
not to be a contrarian, but whenever i find myself agreeing with the top comment on hacker news i want to interrogate that
February 19, 2025 at 4:53 PM
not to be a contrarian, but whenever i find myself agreeing with the top comment on hacker news i want to interrogate that
Reposted by citriqa (extra tart edition)
writing proofs in lean feels just like coding except your entire program is a noop and your only job is to make it typecheck
February 15, 2025 at 8:43 PM
writing proofs in lean feels just like coding except your entire program is a noop and your only job is to make it typecheck
do not use in React 19 if you want your app to be snappy.
it will take a minimum of 300ms to resolve.
github.com/facebook/rea...
it will take a minimum of 300ms to resolve.
github.com/facebook/rea...
[React 19] Suspense throttling behavior (`FALLBACK_THROTTLE_MS`) kicks in too often · Issue #31819 · facebook/react
React version: 19.0.0 Related Issues: #30408, #31697 This may be the intended behavior, but I'm opening this issue anyway because enough discussion hasn't been done in the related issues IMO and th...
github.com
February 18, 2025 at 3:52 AM
do not use in React 19 if you want your app to be snappy.
it will take a minimum of 300ms to resolve.
github.com/facebook/rea...
it will take a minimum of 300ms to resolve.
github.com/facebook/rea...
i keep coming back to this by @bobaekang.bsky.social (and the linked post by @tannerlinsley.com)
RSCs are the 'easy' way out here for the React team, but hardly the 'correct' one...
bobaekang.com/blog/compone...
RSCs are the 'easy' way out here for the React team, but hardly the 'correct' one...
bobaekang.com/blog/compone...
Component, colocation, composition: A note on the state of React | bobae kang
My take on the latest React drama and what it reveals about React's evolving strategy and ecosystem
bobaekang.com
February 17, 2025 at 9:28 PM
i keep coming back to this by @bobaekang.bsky.social (and the linked post by @tannerlinsley.com)
RSCs are the 'easy' way out here for the React team, but hardly the 'correct' one...
bobaekang.com/blog/compone...
RSCs are the 'easy' way out here for the React team, but hardly the 'correct' one...
bobaekang.com/blog/compone...
TIL that React throws away useRef (and all other 'stable' hooks, but this tree is the most egregious) references on suspend, which is yet another instance of this pattern
react is the "worse is better" of frontend frameworks (which is also why it won, and why it won't remain dominant, like others in this vein): a simple model that offloads most of the complexity onto the user
February 17, 2025 at 5:56 PM
TIL that React throws away useRef (and all other 'stable' hooks, but this tree is the most egregious) references on suspend, which is yet another instance of this pattern
firefox is kinda busted but every time a website is actually broken in functionality it turns out to be the same in chrome
February 13, 2025 at 11:41 PM
firefox is kinda busted but every time a website is actually broken in functionality it turns out to be the same in chrome
Reposted by citriqa (extra tart edition)
As of yesterday it has been longer since @solidjs.com re-introduced Signals and Fine-Grained Rendering back into modern web dev than the gap between the initial releases of jQuery and React.
February 10, 2025 at 4:26 PM
As of yesterday it has been longer since @solidjs.com re-introduced Signals and Fine-Grained Rendering back into modern web dev than the gap between the initial releases of jQuery and React.
Reposted by citriqa (extra tart edition)
i like this POV on bookmarklets and extensions as "interventions" for user agency in a user-hostile browser world:
andregarzia.com/2025/02/reta...
andregarzia.com/2025/02/reta...
Retaking The Web Browser, One Small Step At A Time • AndreGarzia.com
AndreGarzia.com website
andregarzia.com
February 10, 2025 at 11:51 AM
i like this POV on bookmarklets and extensions as "interventions" for user agency in a user-hostile browser world:
andregarzia.com/2025/02/reta...
andregarzia.com/2025/02/reta...
Reposted by citriqa (extra tart edition)
haskell is a low level language send tweet
February 10, 2025 at 3:38 AM
haskell is a low level language send tweet
react is the "worse is better" of frontend frameworks (which is also why it won, and why it won't remain dominant, like others in this vein): a simple model that offloads most of the complexity onto the user
February 10, 2025 at 6:22 AM
react is the "worse is better" of frontend frameworks (which is also why it won, and why it won't remain dominant, like others in this vein): a simple model that offloads most of the complexity onto the user
Reposted by citriqa (extra tart edition)
new (nix) blog is up!
please stop using `nixos-rebuild switch`. there's a better way!
aly.codes/blog/2025-02...
please stop using `nixos-rebuild switch`. there's a better way!
aly.codes/blog/2025-02...
Please Stop Using `nixos-rebuild switch`
Better practices for safely and reliably deploying and updating NixOS systems.
aly.codes
February 8, 2025 at 3:29 AM
new (nix) blog is up!
please stop using `nixos-rebuild switch`. there's a better way!
aly.codes/blog/2025-02...
please stop using `nixos-rebuild switch`. there's a better way!
aly.codes/blog/2025-02...