Live data from Hacker News

Stop Slopware

stopslopware.net

51–60 of 171 posts

Re: Stop Slopware

#51
post #37

Earlier quoted context omitted.

> There is absolutely no reason why software today has to be written like software of yesterday. I get what you're saying, but the irony is that AI tools have sort of frozen the state of the art of software development in time. There is now less incentive to innovate on language design, code style, patterns, etc., when it goes outside the range of what an LLM has been trained on and will produce.

Citation needed. I see no reason at all why that's true any more than the screwdriver freezing the state of home design in time.

LLMs aren't like a screwdriver at all, the analogy doesn't work. I think I was clear. LLMs aren't useful outside the domain of what they were trained on. They are copycats. To really innovate on software design means going outside what has been done before, which an LLM won't help you do.

Re: Stop Slopware

#52

[flagged]

the issue is false advertising. you'll understand the first time you lose half an hour evaluating a library that has all the old signs of competent design and even the trivial examples don't work and you realize the project was generated and you've had your time completely wasted

In the new world, increasingly you’ll be better off writing your libraries from scratch than pulling in external dependencies. Less supply chain risk. Less bloat from features you don’t need. Sure complex mature libraries aren’t going away. But for many simple tasks the balance is shifting.

Re: Stop Slopware

#53

[flagged]

If AI tools produced good code that was well-designed, nobody would fucking care. Maintainers of repos want good submissions that solve issues, they don't care if you wrote it on a notepad first or generated it in 5 minutes, if it fixes the thing, and is of good quality, it would be accepted.

AI submissions are getting hate because the code isn't good and isn't designed well and the person submitting it would also know that, if they understood what the LLM wrote. Hell, maybe they could even re-architect it into something worth submitting in the first place.

I use AI tools semi-frequently, to get answers, to riddle problems, all kinds of stuff. However if you just punch directions into ChatGPT and copy/paste the answer, you are not growing as a dev, you are not learning, and yeah your code is probably shit. Not sorry.

Edit: And I feel it's an under-observed point that AI-dependent devs constantly try and hide the fact that they are doing exactly that: copying from ChatGPT, because they also know it's shit and they will be judged. So just... STOP DOING IT.

Re: Stop Slopware

#55
It bothers me that so many programmers I know, here and in real life, seem to never actually have cared about the craft of software development? Just about solving problems.

I like problem solving too. But I also like theory and craft, and in my naïveté I assumed most of us were like me. LLMs divorced craft-programming from tool-programming and now it seems like there were never any craft-programmers at all.

It feels like the group I was part of was just a mirage, a historical accident. Maybe craft-painters felt the same way about the camera.

Re: Stop Slopware

#56
post #9

The author wrote an essay to accompany this site here: https://ficd.sh/blog/your-project-sucks/

This post mirrors my sentiment and the reasons I dislike these sorts of "projects" much more closely than the main site does, deserved to be the main submission, in retrospect.

Re: Stop Slopware

#57

[flagged]

If AI tools produced good code that was well-designed, nobody would fucking care. Maintainers of repos want good submissions that solve issues, they don't care if you wrote it on a notepad first or generated it in 5 minutes, if it fixes the thing, and is of good quality, it would be accepted. AI submissions are getting hate because the code isn't good and isn't designed well and the person submitting it would also kn…

Using Claude to submit PRs to huge open source projects is stupid, for sure.

But if I need a quick tool, like a secret Santa name picker, I’ll just have Claude build it, push it to a repo, link the repo on some PaaS and have a working, deployed app in 20 minutes. No ads, no accounts & no signing up to random websites. I can build it exactly like I want it and include fun Easter eggs for my family.

Building it myself would take 2-3 hours, and the code quality would be drastically better, but that just doesn’t matter.

Re: Stop Slopware

#58
post #6

> if you’re doing this for your own learning: you will learn better without AI . I'm certain that's not true. AI is the single biggest gift we could possible give to people who are learning to program - it's shaved that learning curve down to a point where you don't need to carve out six months of your life just to get to a point where you can build something small and useful that works. AI only hurts learning if you…

It cannot be understated how much of a boon AI-assisted programming has been for getting stuff up and running. Once you get past the initial hurdle of setting up an environment along with any boilerplate, you can actually start running code and iterating in order to figure out how something works. Cognitive bandwidth is limited, and if you need to fully understand and get through 10 different errors before anything w…

> It cannot be understated

You mean it cannot be overstated?

Re: Stop Slopware

#59
post #55

It bothers me that so many programmers I know, here and in real life, seem to never actually have cared about the craft of software development? Just about solving problems. I like problem solving too. But I also like theory and craft, and in my naïveté I assumed most of us were like me. LLMs divorced craft-programming from tool-programming and now it seems like there were never any craft-programmers at all. It feels…

After a certain point, I think "craft" becomes a meaningless and self-flattering scapegoat, similar to how people use "taste" or "courage" as an excuse to make boneheaded decisions. Most software customers aren't buying software because it's tasteful or impressively crafted, but because it fills a gap in their workflow. People who obsess over polish often end up missing the forest for the trees.

Plus, code-based meritocracy flat-out doesn't exist outside the FOSS circle. Many of the people you know are clocking-in at a job using a tech stack from 2004, they aren't paid to recognize good craftsmanship. They show up, close some tickets, play Xbox during on-call and collect their paycheck on Friday.

The people who care might be self-selecting for their own failure. It's hard to make money in tech if your passion for craft is your strongest attribute.

Post reply on HN