Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
rgoulter
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
121.
▲
by
rgoulter
1y ago
> Nix always felt like an OS on the blockchain. It's fine if you value verifiability above everything, but becomes very burdensome if you find yourself tweaking your dev environment often. With Nix, "dev environment" can b
122.
▲
by
rgoulter
1y ago
> But the driving force behind this seems not to be to identify that something is AI, but because they spite it so (AI writing), to quickly rule out caring about the material. Your expressed concern is "people don't like AI; th
123.
▲
by
rgoulter
1y ago
Examples of LLM-style text: short & punchy sentences, negative parallelism ("not just X, it's Y"), bullet points especially with emojis and bolded text. Overuse of em-dash. This is a good list: https://en.wikip
124.
▲
by
rgoulter
1y ago
The PR author seems clueless. But, it's at least a little remarkable that the cluelessness was able to then pester someone in an unusual way.
125.
▲
by
rgoulter
1y ago
> I'm not sure what type of software you're talking about... The original post mentions several examples where he ran into friction. (Monitoring/tracking software, pinentry-kwallet, playwright, tableplus). A popular escape
126.
▲
by
rgoulter
1y ago
OP had a situation where software he wanted to run didn't work with NixOS. If OP wanted the benefits of the Nix package manager (e.g. per-project development environments, using Nix to build container images, etc.), and didn't wan
127.
▲
by
rgoulter
1y ago
> You can just tell them to write in a different style and they will. I'm guessing the priorities are to have contributions which stick to Wikipedia's guidelines. The LLM tendencies cited are in violation of those. I don't
128.
▲
by
rgoulter
1y ago
The author mentions it in the conclusion, but I think "nix package manager on non-NixOS Linux" would be the best of both worlds. You'd avoid the friction of having to deal with software that doesn't 'just work'
129.
▲
by
rgoulter
1y ago
I think focus on satiation & calorie density makes good sense. "Ultra-processed" is not completely precise, not completely accurate proxy for "hyper palatable, low satiation, calorie dense".
130.
▲
by
rgoulter
1y ago
> men burn more calories than woman just right out of the gate Sure, but parent's point is that the energy input from sugary snacks like chocolate roughly equates to a significant amount of exercise. (Over an hour of walking). -- Ma
131.
▲
by
rgoulter
1y ago
Indeed. Helix's multi-cursor support seems like a rocket-powered chainsaw to me. It's powerfully expressive. IIRC, you can select a region (such as a whole file), split the selection into multiple cursors, retain only the selectio
132.
▲
by
rgoulter
1y ago
> The thing is that there's no way to objectively measure this. Sure. But isn't that a bit like if someone likes VSCode, & someone likes Emacs.. the first method of comparison I'm reaching for isn't "what obj
133.
▲
by
rgoulter
1y ago
A couple of typical comments about LLMs would be: "This LLM is able to capably output useful snippets of code for Python. That's useful." and "I tried to get an LLM to perform a niche task with a niche language, it perfo
134.
▲
by
rgoulter
1y ago
> [bolded] Lua isn’t just a high-level language. It’s an embedded dev strategy. I find it difficult to take any writing seriously when it uses phrases like this.
135.
▲
by
rgoulter
1y ago
I think the better question is, "do I benefit from improving the speed here, for the cost it takes". Improving typing speed from "fast" to "faster" is very difficult. I think it's worth distinguishing betw
136.
▲
by
rgoulter
1y ago
From the summary: """Going forward, a balanced approach is advisable, one that might leverage AI for routine assistance but still challenges individuals to perform core cognitive operations themselves. In doing so, we can har
137.
▲
by
rgoulter
1y ago
> > The biggest problem with Nix is its commit-based package versioning. > ...isn't that like...the whole selling point of Nix? Not quite. That sentence is definitely the most ... discussion-worthy comment in the blog. To my u
138.
▲
by
rgoulter
1y ago
If you're working with a language you're not-so experienced in, then the completion for `// read data from file f into x` is going to be significantly quicker than looking up the documentation for this.
139.
▲
by
rgoulter
1y ago
Related: * Deldo is a sex toy control and teledildonics mode for Emacs (github.com/qdot) (2021) https://news.ycombinator.com/item?id=29207607 | 126 comments
140.
▲
by
rgoulter
1y ago
Conversely, do you think there are developers who would not benefit from being able to type quicker?
141.
▲
by
rgoulter
1y ago
It's nice to have good tools that are practical for navigating around the text. For jumping around what's on screen, I think 'easymotion' ("jump anywhere on screen by pressing two characters") & variations
142.
▲
by
rgoulter
1y ago
"The SDK is open source" Oh, a BLE MCU with open source code SDK? AFAICT, the BLE code is provided as a binary blob. https://github.com/OpenSiFli/SiFli-SDK/tree/6c82a9b15db49871... Which isn't
143.
▲
by
rgoulter
1y ago
For the quality of the coffee, the Vietnamese phin is simplest/cheapest device.
144.
▲
by
rgoulter
1y ago
When browsing a GitHub repo, there's an option for "assistive chat" with copilot. -- I've found this a useful interface to get the LLM to answer quick questions about the repository without having to dig through myself.
145.
▲
by
rgoulter
1y ago
I wish the discussion on differences between the agentic workflow in copilot and cursor were clearer. As is presented, copilot's workflow is described as "request changes via chat, then approve/reject", and cline's
146.
▲
by
rgoulter
1y ago
I'd suggest the original article makes for better reading. https://socket.dev/blog/wget-to-wipeout-malicious-go-modules...
147.
▲
by
rgoulter
1y ago
At a glance, this seems really neat. -- I reckon one thing LLMs have been useful to help with is "the things I'd copy-paste from stack overflow". A loop of "let's fix each error" reminds me of that. I'd al
148.
▲
by
rgoulter
1y ago
Since LLMs are sometimes wonderfully useful, and sometimes not, I'd suggest effective use involves figuring out in which cases it's likely to succeed, which it's likely to fail. For example, the mentioned graph has "init
149.
▲
by
rgoulter
1y ago
I liked the idea I heard: internet audiences demand progress, but internet audiences hate change.
150.
▲
by
rgoulter
1y ago
In terms of navigating around a document, the only difference between "motion-action" and "action-motion" is that with "motion-action", the cursor is a selection which highlights the last motion. -- I don
More ›