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

Agents that run while I sleep

claudecodecamp.com

Post reply on HN
Print view View original

441–450 of 532 posts

Re: Agents that run while I sleep

#441

Post by nurettin »

recroad wrote: post #83

Am I supposed to be impressed by this? I think people are now just using agents for the sake of it. I'm perfectly happy running two simple agents, one for writing and one for reviewing. I don't need to go be writing code at faster than light speed. Just focusing on the spec, and watching the agent as it does its work and intervening when it goes sideways is perfectly fine with me. I'm doing 5-7x productivity easily,…

They are probably paying for expensive subscriptions and want to utilize them. Unfortunately we aren't past the slop stage so a lot of the business logic probably has bugs and unused defensive code that snowballs the more features AI adds.

Re: Agents that run while I sleep

#442

Post by thewhitetulip »

hi_hi wrote: post #206

This _all_ (waves hands around) sounds like alot of work and expense for something that is meant to make programming easier and cheaper. Writing _all_ (waves hands around various llm wrapper git repos) these frameworks and harnesses, built on top of ever changing models sure doesn't feel sensible. I don't know what the best way of using these things is, but from my personal experience, the defaults get me a looong wa…

I saw a guys post on LinkedIn who created llm agent to water how plants based on sensor on his plants

He still has to water the plants on his own. Its just that it costs him quite a bit when all of that could he mamaged with an alarm to remind him to water plants.

Re: Agents that run while I sleep

#444

Post by giva »

mewpmewp2 wrote: post #207
hi_hi wrote: post #206

This _all_ (waves hands around) sounds like alot of work and expense for something that is meant to make programming easier and cheaper. Writing _all_ (waves hands around various llm wrapper git repos) these frameworks and harnesses, built on top of ever changing models sure doesn't feel sensible. I don't know what the best way of using these things is, but from my personal experience, the defaults get me a looong wa…

I am not laughing about PHP. To this very day many of my best projects are built on PHP. And while last 7 years I have spent in full stack JavaScript/TypeScript environment it has never produced the same things I was actually able to do with PHP. I actually feel that things I built 15 years ago in PHP were better than anything I am trying to achieve with modern things that gets outdated every 6 months.

I feel like today an engineer with a modern framework and AI con produce in an afternoon a product that deliver real value, something that 25 years ago would have required a full hour by a high schooler with MS Access.

Re: Agents that run while I sleep

#446

Post by palmotea »

nananana9 wrote: post #406
KronisLV wrote: post #398

Earlier quoted context omitted.

> they are left with no one who understands a sprawling tangled web of code that is 80% [random people that I can't ask because they don't work here anymore and they didn't care to leave docs or comments] generated, then we'll see who laughs last. Yes, this matches my experience with codebases before AI was a thing.

Yes, but given a feature that should take say 100 lines of code, the average programmer will write in the order of 100 to 500 lines. If they're a heavy OOP user, maybe they'll write 10 classes that total 2000 lines. Regardless, worst case, it will be within ~2 orders of magnitude of a reasonable solution. It's not that they're not trying to write the biggest clusterfuck possible and maximize suffering in the world, i…

> it's just that there's a human limit on how much garbage they can type out in their allocated time.

Another example where removing friction and constraints is a bad thing.

Re: Agents that run while I sleep

#447

Post by skeledrew »

rurban wrote: post #381
skeledrew wrote: post #339

Earlier quoted context omitted.

Introducing uv... https://docs.astral.sh/uv/

uv does not fix the need for venv's or docker containers. normal people update their libs with the hope to get problems fixed. python people don't update their libs, because then everything will break right and left. so they keep their security problems running.

No matter how you look at it, the dependencies have to go somewhere. Node uses node_modules, most compiled languages require compiled libraries (or they're a huge blob), etc. Idk about PHP but I'm pretty sure 3rd party things for any given app also live somewhere. Different ways of managing dependencies. It's recommended that venvs are used in Python because you may accidentally nuke a system script by doing global installs, and otherwise there still needs to be some sort of 3p version handling when you have multiple projects going.

Once something works in Python (which uv now makes trivial; before it could be a pain), updating 3rd party packages rarely cause breakage. But yes, I think many who use it hardly update, because things usually continue to work for years and the attack surface is pretty narrow[0]. Heck just a few days ago I checked out a project that I hadn't touched in years, which I wrote in Python 3.7; updated to 3.13 and it continued to just work. Compare to PHP which has a far higher attack surface[1] and often has breaking changes. I've heard a couple nightmare stories of a v7.x -> v8.x move being delayed because it required a serious codebase rewrite.

[0] https://www.cvedetails.com/product/18230/Python-Python.html?... [1] https://www.cvedetails.com/product/128/PHP-PHP.html?vendor_i...

Re: Agents that run while I sleep

#449

Post by predkambrij »

RealityVoid wrote: post #4

It's... really the same problem when you hire people to just write tests. A lot of time it just confirms that the code does what the code does. Having clear specs of what the code should do make things better and clearer.

Yeah. The purpose of tests is to verify correctness of their setUp mocks :)
Post reply on HN