Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
barrkel
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
151.
▲
by
barrkel
8mo ago
There's another gap, actually. Imprecise specification. When you need to implement something yourself, you have to make decisions when faced with the reality of turning ideas into code. An AI agent sometimes surfaces these; and sometim
152.
▲
by
barrkel
8mo ago
Understanding other people's code is harder than understanding your own code though.
153.
▲
by
barrkel
8mo ago
Search engines are more costly than inference AIUI and are certainly slower. The models are very expensive to train of course and incremental learning without catastrophic forgetting hasn't been solved. I would think whoever cracks cou
154.
▲
by
barrkel
8mo ago
You don't get the AI component without the knowledge component. The AI needs approximate knowledge of lots of things to conceptualize what you're talking about and use search tools effectively. The set of things it needs approxima
155.
▲
by
barrkel
8mo ago
OpenAI is the Yahoo of AI.
156.
▲
by
barrkel
8mo ago
Don't forget the .LOG feature.
157.
▲
by
barrkel
8mo ago
The goal is obviously to onboard consumers to an app store ecosystem so MS can charge monopoly rents to app developers for access to Windows users. Everything MS is doing in Windows is to this end.
158.
▲
by
barrkel
8mo ago
I'm describing how Google works with teams with high release cadence, fwiw. Also, your comment reads a little bit as a non sequitur.
159.
▲
by
barrkel
8mo ago
I'd much rather reduce the risk of mutation to the trunk, by having small easily reviewable commits direct to trunk. It's less about reviewing commits from a year ago, than making change low-risk today. And small commits can easil
160.
▲
by
barrkel
8mo ago
I got the impression the author is female.
161.
▲
by
barrkel
8mo ago
The hard thing about coding isn't really the code. It's the data. Both data at rest and data flowing in and out of your system. Vibe coding creates the illusion that code has become far more malleable. And it has, for greenfield,
162.
▲
by
barrkel
8mo ago
Why is it crazy? Any rewrite that would be as flexible wrt mods would be shaped similarly.
163.
▲
by
barrkel
8mo ago
They have rate limits, but they also want to control the nozzle, and not all their users use all their allocation all the time. In reality, heavy subscription users are subsidized by light subscription users. The rate limits aren't eve
164.
▲
by
barrkel
8mo ago
Have you not seen it any time you put any substantial bit of your own writing through an LLM, for advice? I disagree pretty strongly with most of what an LLM suggests by way of rewriting. They're absolutely appalling writers. If you&#x
165.
▲
by
barrkel
8mo ago
This is a good statement of what I suspect many of us have found when rejecting the rewriting advice of AIs. The "pointiness" of prose gets worn away, until it doesn't say much. Everything is softened. The distinctiveness of
166.
▲
by
barrkel
8mo ago
If you give a smart AI these tools, it could get into it. But the personality would need to be tuned. IME the Grok line are the smartest models that can be easily duped into thinking they're only role-playing an immoral scenario. Whate
167.
▲
by
barrkel
8mo ago
> evidence at all that Anthropic or OpenAI is able to make money on inference yet. The evidence is in third party inference costs for open source models.
168.
▲
by
barrkel
8mo ago
That seq is probably supposed to be $(seq 0.05 0.05 0.5). Right now it's always 0.05. Note that you can get random numbers straight from bash with $RANDOM. It's 15 bit (0 to 32767) but good enough here; this would get between 0.05
169.
▲
by
barrkel
9mo ago
itzg's docker image for Minecraft makes life quite a bit easier - switching mod loader, versions etc. with environment variables is simpler than continuously downloading versions and juggling directories.
170.
▲
by
barrkel
9mo ago
The Bell / telephony analogy is unrelated and forced. Feels like the author had an Idea, and asked an LLM to force the analogy.
171.
▲
by
barrkel
9mo ago
I think this mostly only holds if you use local compute in a portable form factor. Most of my personal development these days is done on my home server - 9995wx, 768GB, rtx 6000 pro blackwell GPU in headless mode. My work development happen
172.
▲
by
barrkel
9mo ago
The main purpose of alarms is to relieve automation of liability.
173.
▲
by
barrkel
9mo ago
I also have a Skoda that also has that "feature". I prefer using Maps via Android Auto, but if I am in that interface, and I have to cancel it, the way I cancel it is using a voice command.
174.
▲
by
barrkel
9mo ago
When you're under a lot of stress, your internal evaluation function for what is moral can start to break down. It may have been hard for him to turn the money down, especially if he's addicted to the sense of power he's gett
175.
▲
by
barrkel
9mo ago
I think there's an interesting idea behind Gas Town (basically, using supervisor trees to make agents reliable, analogous to how Erlang uses them to make processes reliable), but it's lacking a proper quality ratchet (agents often
176.
▲
by
barrkel
9mo ago
In practice you see noticeable degradation of performance for streaming reads of large files written after 85% or so. Files you used to be able to expect to get 500+MB/sec could be down to 50MB/sec. It's fragmentation, and it
177.
▲
by
barrkel
9mo ago
I scrub once a quarter because scrubs take 11 days to complete. I have 8x 18TB raidz2 pool, and I keep a couple of spare drives on hand so I can start a resilver as soon as an issue crops up. In the past, I've gone for a few years betw
178.
▲
by
barrkel
9mo ago
That's a fine fit of pique - and I once had an awkward file on one of my zfs pools, about three pools ago - but how does it leave you better off, if you want what zfs offers?
179.
▲
by
barrkel
9mo ago
If by block by block you mean you stop using an IDE and spend most of your time looking at diffs, sure. Because in a well structured project, that's all you need to do now: maintain a quality bar and ensure Claude doesn't drop the
180.
▲
by
barrkel
9mo ago
I'm like you. I get on famously with Claude Code with Opus 4.5 2025.11 update. Give it a first pass from a spec. Since you know how it should be shaped you can give an initial steer, but focus on features first, and build with testabil
More ›