Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Superpowers: How I'm using coding agents in October 2025

blog.fsck.com

Post reply on HN
Print view View original

161–170 of 242 posts

Re: Superpowers: How I'm using coding agents in October 2025

#161

Post by groby_b »

simonw wrote: post #149
nightski wrote: post #87

Earlier quoted context omitted.

If all I did was call it trivial that would be a fair critique. But it was followed up with a lot more justification than that.

Here's the PR. It touched 21 files. https://github.com/ghostty-org/ghostty/pull/9116/files If that's your idea of trivial then you and I have very different standards in terms of what's a trivial change and what isn't.

It's trivial in the sense that a lot of the work isn't high cognitive load. But... that's exactly the point of LLMs. It takes the noise away so you can focus on high-impact outcomes.

Yes, the core of that pull requests is an hour or two of thinking, the rest is ancillary noise. The LLM took away the need for the noise.

If your definition of trivial is signal/noise ratio, then, sure, relatively little signal in a lot of noise. If your definition of "trivial" hinges on total complexity over time, then this kicks the pants of manual writing.

I'd assume OP did the classic senior engineer stick of "I can understand the core idea quickly, therefore it can't be hard". Whereas Mitchel did the heavy lifting of actually shipping the "not hard" idea - still understanding the core idea quickly, and then not getting bogged down in unnecessary details.

That's the beauty of LLMs - it turns the dream of "I could write that in a weekend" into actually reality, where it before was always empty bluster.

Re: Superpowers: How I'm using coding agents in October 2025

#162

Post by truetraveller »

specproc wrote: post #131
sawmurai wrote: post #98

Earlier quoted context omitted.

I have a similar experience. It feels like riding your bike in a higher gear - you can go faster but it will take more effort and you need the potential (stronger legs) to make use of it

It's more like shifting from a normal to an electric bike. You can go further and faster, but you can get to a point where you're out of juice miles from home, and getting back is a chuffing nightmare. Also, you discover that you're putting on weight and not getting that same buzz you got on your old pushbike.

Hey, that's a great analogy, 10/10! This explains in a few words what an entire article might explain.

Re: Superpowers: How I'm using coding agents in October 2025

#163

Post by 0x696C6961 »

drivebyhooting wrote: post #150
ehsanu1 wrote: post #140

Earlier quoted context omitted.

Using Research->Plan->Implement flow is orthogonal, though I notice parts of those do exist as skills too. But you sometimes need to do other things too, e.g. debugging in the course of implementing or specific techniquws to improve brainstorming/researching. Some of these skills are probably better as programmed workflows that the LLM is forced to go through to improve reliability/consistency, that's what I've found…

Orthogonal means there should not be any overlap.

Yeah, but you knew what he meant.

Re: Superpowers: How I'm using coding agents in October 2025

#165

Post by hugh-avherald »

Aloisius wrote: post #164

What's up with people (or I suppose AI) including copyright licenses in AI generated code? At least it's an MIT license, but since AI output isn't copyrightable, I'm unsure what the point is since people can legally ignore the license.

^ (not legal advice -- far from it)

Re: Superpowers: How I'm using coding agents in October 2025

#166

Post by kibwen »

wbradley wrote: post #127
simonw wrote: post #118

Earlier quoted context omitted.

It's in the cache location because it's a copy of a plugin that was installed from a GitHub repository, so that's not the original point of truth for that file.

I think the point is that ~/.claude should be dispersed among ~/.config/claude, ~/.local/state/claude, etc I agree with this, it’s frustrating that in 2025 apps are still polluting my home dir.

It's one thing to wish that apps would put their data anywhere except dumping it in your home dir, but this is exactly why I hate the XDG spec. I want all data for a program--be it the configuration or the cache or the binary itself--to be in a single directory such that 1) "uninstalling" the program, completely and in isolation, is nothing more than just deleting that single directory, and 2) any program not doing arbitrary file I/O can entirely function while having access to only its installation directory, and nothing else on the filesystem.

Re: Superpowers: How I'm using coding agents in October 2025

#167

Post by Aloisius »

hugh-avherald wrote: post #165
Aloisius wrote: post #164

What's up with people (or I suppose AI) including copyright licenses in AI generated code? At least it's an MIT license, but since AI output isn't copyrightable, I'm unsure what the point is since people can legally ignore the license.

^ (not legal advice -- far from it)

If there some reason why one wouldn't be able to ignore the copyright license of something not protected by copyright, I'd love to hear it.

The copyright office has been quite clear (rightly so imo) that AI output is not protected by copyright without substantial human creative expression in the final product and purely prompt-created works simply don't qualify.

Indeed, I expect people muddling their codebases with AI output are going to find themselves in an interesting position of having to prove how much code humans actually wrote to enforce copyright claims if their code ever gets leaked.

Re: Superpowers: How I'm using coding agents in October 2025

#168

Post by Retric »

antonvs wrote: post #157
Retric wrote: post #101

Earlier quoted context omitted.

Who hurt you? Sorry couldn’t resist. Agile’s point was getting feedback during the process rather than after something is complete enough to be shipped thus minimizing risk and avoiding wasted effort. Instead people are splitting up major projects into tiny shippable features and calling that agile while missing the point.

Agile’s point was to get feedback based on actual demoable functionality, and iterate on that. If you ignore the “slop” pejorative, in the context of LLMs, what I quoted seems to fit the intent of Agile.

There’s generally a big gap between the minimum you can demo and an actual feature.

Re: Superpowers: How I'm using coding agents in October 2025

#170

Post by kragen »

Aloisius wrote: post #167
hugh-avherald wrote: post #165

Earlier quoted context omitted.

^ (not legal advice -- far from it)

If there some reason why one wouldn't be able to ignore the copyright license of something not protected by copyright, I'd love to hear it. The copyright office has been quite clear (rightly so imo) that AI output is not protected by copyright without substantial human creative expression in the final product and purely prompt-created works simply don't qualify. Indeed, I expect people muddling their codebases with A…

That's just the copyright office of one country out of a couple hundred, the courts can overrule them, and legislation can change. However, I agree that currently in the US (or on code written in the US) copyright probably doesn't inhere in AI-written code.
Post reply on HN