Ok, I use it a lot to ask technical questions. But the code it generates from other coworkers is bad.

Tracking user intention and attention as part of a user model is reasonable. But it isn't parsed

Yes, kebab / dash – LISP, Scheme, Dylan, HTML, CSS, etc. If white-space is required between

Jokingly, I mean SQL is not that bad but it's raw - joining tables manually, non-nested data.

I keep reading MySQL clustered index make finding primary keys fast - the table itself is a b-tree

A Comprehensive Overview of GPU Accelerated Databases GPUDirect Storage: A Direct Path Between

(unsigned long long) but I see ...FULL! :)

I'm way out of date on SIMD. I looked at it when it was more about stuffing a single register

train an llm from scratch on everything OTHER than that, then see if it passes the test. On Sat, Sep

Run Linux GUI apps with WSL | Microsoft Learn I knew about the Windows Subsystem for Linux, but I

AHHH LINE NOISE!! On Thursday, September 3, 2026 at 3:19:50 PM UTC-7 raould wrote: meh. maybe a real

Being primarily a front-end UX developer and with database-as-a-service services, I haven't had

https://computer-museum.ru/english/setun.htm On Mon, Aug 17, 2026 at 11:47 Raoul Duke <raould@

The LLMs with big context windows can adapt to non-sequential languages very well, in my experience.

Maybe of interest?... This is a screencast of how I implement a new language. It's a long (80-

At the moment, most AIs will ignore architecture docs when implementing new features. Unless you

hear hear On Sat, Jul 11, 2026 at 09:28 Paul Tarvydas <paulta...@gmail.com> wrote: I've

yeah ...but what if due to the way the code is, the only reasonable type is something akin to void*?

rxjs proves to me that plain 2d ascii is not good for anything much with concurrency, parallelism,

but maybe ai-agents can help some day: if you have no schema, no good constraints on cardinalities

On Mon, Jun 29, 2026, 3:01 PM Paul Tarvydas <paulta...@gmail.com> wrote: "Coherence

Sounds to me: Not wrong directionally, but a bit pie in the sky. The devil is in the details. eg

https://www.google.com/search?udm=2&q=ai+everywhere+meme Slow down there, we don't need AI

A way to finally stop allowing string truthyness if (!string) { } which should be if (string.length =

It's not magical. You still have to think. When you dissect someone else's code, you can

https://youtu.be/AGyjo2DLxjM Refresher for De Morgan's Law. I needed that. On Wednesday, April 29

I started looking at autocomplete. It could be done without text parsing – just a decision tree. ----

https://www.npmjs.com/package/remeda vs ramda vs rambda vs ts-belt: https://npmtrends.com/@mobily/ts-

Nice. Yeah, diagrams go a long way. I watched some of the videos and would like to see something like

For me, Solving Problems (first understanding them). Managing Complexity (abstract but not too far).