jeremyjh

26 ms·

301.
by jeremyjh 4mo ago
When you say it is not deal-breaking - have you looked at any math? Real world data finds we need about 6kg of radiator surface to radiate 1kW. [0] . A 100MW datacenter would need about 600,000 KG just of radiator mass. ISS was only 450,000
302.
by jeremyjh 4mo ago
It had nothing to do with security - it had to do with contractual obligations. Contracts with Apple (also Google, Samsung, Mastercard and Visa) required the product to be kept absolutely secret before the public launch. I was a tech lead w
303.
by jeremyjh 4mo ago
How much mass do we need to put in orbit to radiate 10 GW of data center heat? How much rocket fuel does it take to lift the rocket fuel required to lift that mass? Which part of this is cheaper or more effective than building data centers
304.
by jeremyjh 4mo ago
I doubt most people use OpenCode coding plans, nor do they use OpenRouter. I use subscription plans from ChatGPT, z.ai, MiniMax & Xiaomi with OpenCode. It handles authentication with all of them seamlessly. I switch between models based
305.
by jeremyjh 4mo ago
I'm sure that explains some of it but I really don't think it explains most of the people who have been AI-pilled in the last nine months. There was no amount of context I could give GPT-4o that would make it a net benefit to use
306.
by jeremyjh 4mo ago
They don’t believe in competence because they’ve never experienced it. They think everything is narrative and spectacle.
307.
by jeremyjh 5mo ago
Don’t forget, “Kill the guy and get the thing, then take that to this other guy.”
308.
by jeremyjh 5mo ago
This was my experience using GLM 5.1 in Claude Code but it works far better in OpenCode, I’d really like to understand why. I think it’s a bit stronger than Sonnet 4.6. I use the oh-my-openagent planning system and haven’t used vanilla Open
309.
by jeremyjh 5mo ago
There was never an era in which #2 was a reasonable policy.
310.
by jeremyjh 5mo ago
I'm guessing the number of people who care about commit logs is dropping and the number who don't is skyrocketing.
311.
by jeremyjh 5mo ago
This is perfectly safe as long as you keep your username a secret.
312.
by jeremyjh 5mo ago
The person who "writes" code is also supposed to review their own work, and answer for that. If they won't do that - well - they should be fired. But if you have weak or uninvolved leadership, then the team's only ration
313.
by jeremyjh 5mo ago
Also - at least in GitHub if you squash with the PR merge action in the UI - the original commit history is still available in the ref maintained by the closed PR yet doesn’t clutter your branch or tags namespace.
314.
by jeremyjh 5mo ago
mise use -g ripgrep@latest I recently found mise and have become a fan as well. I have used asdf for about a decade and it supports the same .tool-versions files so initially I used it for those exact same things. But I use four different c
315.
by jeremyjh 5mo ago
GLM 5.1 is stronger than Sonnet 4.6 in my opinion, but while they have a coding plan that is a good value MiMo beats it on price. I haven't used MiMo much yet but it felt pretty similar.
316.
by jeremyjh 5mo ago
But you really don't have to switch. MiMo Code has the same provider support as OpenCode. Even Claude Code you can use with any provider that exposes an anthropic API endpoint, which they all do.
317.
by jeremyjh 5mo ago
I agree that evolution could not produce a rational agent who would still reliably respond to lower level imperatives (such as pain, hunger, lust) without consciousness and feeling. The primitive parts of the brain have to be able to overri
318.
by jeremyjh 5mo ago
Yes but the sales process for enterprise is often 9 months or more, and you have to have support coverage before it pays for itself.
319.
by jeremyjh 5mo ago
> With $5.5M from Basis Set, YC, Pioneer Fund and other great investors, we have years of runway, This is years of product development with a three person team. If Enterprise sales and support are a big part of your business plan it will
320.
by jeremyjh 5mo ago
Excellent article, this makes a lot of sense! TLDR: Tokio concurrency > Process concurrency in OLTP.
321.
by jeremyjh 5mo ago
Aurora has a completely different storage backend. PgDog is a front end proxy - each server in the cluster is still using standard Postgres right?
322.
by jeremyjh 5mo ago
It’s surprising they don’t mention advantages over other sharding systems like Citus. Maybe it’s just the fact that it’s only a proxy and not core extensions? But that could limit capabilities.
323.
by jeremyjh 5mo ago
They are betting that without a competitor distilling their most powerful models, they can stay ahead far enough and long enough that people will accept this.
324.
by jeremyjh 5mo ago
TFA explains how this works, and how to opt out.
325.
by jeremyjh 5mo ago
A lot of packages are only used in the browser; if you don't use SSR they'd only be executed by node in unit tests in something like jest, but that is not the only way to run unit tests (Cypress can run them in a headless browser
326.
by jeremyjh 5mo ago
I really don't see how you can be sure of that. Who would want to live in a steel cave for their entire life? Anything we need from Mars could be extracted with automated machinery.
327.
by jeremyjh 5mo ago
Like how he figured out how to destroy half the enterprise value of Twitter? Like how he figured out how to promise full self driving cars in about six months? For eight years running? Like how he figured out how to cut USAID programs that
328.
by jeremyjh 5mo ago
How is it meant to be taken, other than as a laughably naive statement which is patently false? It is not possible to write memory safe code in C. It is also not possible to always write correct code in ANY language, but languages that elim
329.
by jeremyjh 5mo ago
> You can write memory safe code in C (Redis, SQLite, OpenBSD, Git, etc), Do you think that this is a list of software that have never had memory bugs? It really is not practically possible to completely avoid a large class of memory bug
330.
by jeremyjh 5mo ago
Very few of them have worked on a single system for more than four or five years, and have no idea what their decisions cost after they left. Many have joined a project after four or five years and suffered from those decisions, but they do

More ›