Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
epicprogrammer
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
by
epicprogrammer
7mo ago
I've been noticing this creeping into my own AI coding suggestions lately. An LLM doesn't inherently understand "abandonware" or community health; it just sees that a package technically solves the logic puzzle in its co
2.
▲
The 89% Problem: LLMs are resurrecting the "dormant majority" of open source
(snyk.io)
4 points
by
epicprogrammer
7mo ago
|
1 comments
3.
▲
by
epicprogrammer
7mo ago
It’s an interesting throwback to SEDA, but physically passing file descriptors between different cores as a connection changes state is usually a performance killer on modern hardware. While it sounds elegant on a whiteboard to have a dedic
4.
▲
by
epicprogrammer
7mo ago
Having spent some time in the anti-abuse and Trust & Safety space, I always take these vendor reports with a massive grain of salt. It’s a classic case of comparing apples to vendor-marketing oranges. A headline screaming about an 84% m
5.
▲
by
epicprogrammer
7mo ago
It really is the classic "you either die a safety-first AI lab, or you live long enough to see yourself back at the Pentagon negotiating table" arc.
6.
▲
by
epicprogrammer
7mo ago
It's easy to frame this purely as an ethical battle, but there's a massive financial reality here. Training frontier models requires astronomical amounts of capital, and the DOD is one of the few entities with deep enough pockets
7.
▲
by
epicprogrammer
7mo ago
The most fascinating detail here isn't just the images, but the temperature data. JWST actually found a localized 'cold spot' (538K) right in the core of Io's magnetic footprint, surrounded by a much hotter main aurora.
8.
▲
by
epicprogrammer
7mo ago
I've built a few internal tools using the Workspace APIs, and while they are powerful, the rate limits on the Drive API can be brutal if you are doing bulk operations. Does this repository handle automatic backoff and retries, or do we
9.
▲
JWST reveals extreme temperature shifts in Jupiter's auroral footprints
(phys.org)
5 points
by
epicprogrammer
7mo ago
|
1 comments
10.
▲
MongoBleed: Unauthenticated memory-read vulnerability in MongoDB
(bitsight.com)
1 points
by
epicprogrammer
9mo ago
|
1 comments
11.
▲
Scientists may have found the best place for humans to land on Mars
(sciencedaily.com)
2 points
by
epicprogrammer
9mo ago
|
0 comments
12.
▲
by
epicprogrammer
9mo ago
I am one of the builders behind Dropstone. Yesterday, I posted our whitepaper here as a 'discovery' because I was nervous about a cold launch. That was a mistake, and the community rightfully flagged it for lacking transparency. I
13.
▲
Show HN: Dropstone – A Recursive Swarm Runtime for Long-Horizon Agent Tasks
(dropstone.io)
3 points
by
epicprogrammer
9mo ago
|
1 comments
14.
▲
by
epicprogrammer
9mo ago
You are absolutely right. I addressed this in the thread above, but to be clear: Yes, I am part of the team. I shouldn't have tried to frame it as a 'discovery'. Apologies.
15.
▲
by
epicprogrammer
9mo ago
I am incredibly sorry about that. It sounds like the agent hit an infinite reasoning loop and burned your credits—that is a critical failure on our end. I want to fix this personally. Please email me at tom@blankline.org (or just reply here
16.
▲
by
epicprogrammer
9mo ago
You got me. Yes, I am part of the Blankline team. To clarify the confusion: D2 is the beta engine currently on our website. D3 is the new architecture in these papers. We developed D3 specifically to solve the context limits we hit with D2.
17.
▲
Dropstone's D2 Engine – persistent memory for AI-assisted coding
(dropstone.io)
1 points
by
epicprogrammer
11mo ago
|
1 comments
18.
▲
by
epicprogrammer
11mo ago
We’ve been exploring how long-term memory can make AI coding agents more collaborative and context-aware. The D2 Engine is part of that effort — it gives Dropstone a persistent understanding of codebases across sessions. Would love feedback
19.
▲
by
epicprogrammer
3y ago
This is a fascinating study on the impact of context windows on language models. It's interesting to see how smaller context windows can lead to more efficient and accurate language models, even when dealing with complex natural langua