383toast

8 ms·

91.
by 383toast 2y ago
not surprising
92.
by 383toast 2y ago
for people that don't want to click a lot, paste in "setInterval(() => document.querySelector('.main-btn').click(), 20)" into the browser console, clicks 50x per second for you :)
93.
by 383toast 2y ago
for people that don't want to click a lot, paste in "setInterval(() => document.querySelector('.main-btn').click(), 20)" into the browser console, clicks 50x per second for you :)
94.
by 383toast 2y ago
For what it's worth, good system prompting in Cursor makes an enormous difference, currently I use: You're the best fullstack engineer and LLM engineer, prefer using Vite + React, TypeScript, Shadcn + Tailwind (unless otherwise sp
95.
by 383toast 2y ago
Curious why you only want the AI features 30% of the time? Do you use the Cursor agent mode or just the autocompletion?
96.
by 383toast 2y ago
If you want more brainfuck, read Gnomon by Nick Harkaway
97.
by 383toast 2y ago
Totally agree, Next.js has so much Next-specific stuff (which also changes frequently) that I don't use it anymore.
98.
by 383toast 2y ago
if you have a build on every CI run (every commit pretty much) and a preview link, then it really adds up...
99.
by 383toast 2y ago
React for me has several large advantages (no specific order): 1. declarative dom updates is way easier than imperative 2. frontend components Additionally typescript needs a compile step anyways so it's not super complicated to just a
100.
by 383toast 2y ago
Yep nextjs doesn't really provide much for regular CRUD/SaaS apps, just use Vite react starter
101.
by 383toast 2y ago
Yep AoC explicitly discourages using LLMs for leaderboard scoring
102.
by 383toast 2y ago
> All the top 5 players on the final leaderboard https://adventofcode.com/2024/leaderboard used LLMs for most of their solutions. Source for this?
103.
by 383toast 2y ago
Wouldn't the work on interpretability would solve these concerns?
104.
by 383toast 2y ago
Seems like this would be a first step towards grey goo...
105.
by 383toast 2y ago
which GPUs?
106.
by 383toast 2y ago
NOTE: this is a july 2023 paper, the defense paper in september 2024 is https://arxiv.org/abs/2406.12843
107.
Supernovae evidence for foundational change to cosmological models(academic.oup.com)
2 points by 383toast 2y ago|0 comments
108.
by 383toast 2y ago
i assume EMS (electrical muscle stimulation)
109.
by 383toast 2y ago
getting splintered metaverses seem way more feasible than splintered physical planets
110.
by 383toast 2y ago
also bold take to proclaim we'll still have human bodies before moving up kardashev scale
111.
by 383toast 2y ago
Via various incentives and regulation, they kind of can
112.
by 383toast 2y ago
But to me it seems that Bun ( https://bun.sh/ ) is largely a better solution, faster/easier to use/doesn't require lots of random knowledge
113.
by 383toast 2y ago
Note this is pretty old, October 9, 2024 Tl;dr - from the website ( https://deno.com/blog/v2.0 ): Today, we’re thrilled to announce Deno 2, which includes: Backwards compatibility with Node.js and npm, allowing you to ru
114.
by 383toast 2y ago
What's the stance on security for handling private tokens/cookies/sessions/etc?
115.
by 383toast 2y ago
Microsoft owns 50%, other investors own 20-30% presumably(?), 10% employee pool, that means all the other OpenAI founders/etc basically get nothing compared to Altman...
116.
by 383toast 2y ago
does celery work with this out of the box?
117.
by 383toast 2y ago
Can someone explain why a single geolocated node makes sense for data storage? If there's a house fire for example, wouldn't all the data be lost?
118.
by 383toast 2y ago
I don't see why a senior dev does not have broad skills that apply across companies?
119.
by 383toast 2y ago
False dichotomy, it is possible to have a team of seniors that have a culture of sharing knowledge/open to feedback with no juniors though.
120.
Show HN: A LeetCode assistant that guides you gradually(chromewebstore.google.com)
2 points by 383toast 2y ago|0 comments

More ›