alt Hacker NewsGreat effort, thanks for that.
But do we still trust the open source software that is created now everything is vibe coded?
Recently there was a thread about Chatto [1], a Slack replacement completely vibe coded by one person. It works and looks nice. But looking at the code, I wouldn’t trust any of my personal information to it. It’s just tousands of lines of skills and vague instructions. It seems that trust in OSS is eroding and big SaaS might win here because they have the reputation.
Replies
You shouldn't trust open source software because it is open source. You should trust open source software because it has been audited, and only when it has been audited by people you trust (it may be yourself).
Whether it's vibe-coded or not doesn't change that. If it is vibe-coded and unreadable, how is it different from manually-written and unreadable? If it is vibe-coded and readable, how is it different from manually-written and readable?
Judge the code by its quality, not by its provenance.
I judge it on a case-by-case basis by just looking at a few recent commits and the feel of clicking a few files in the repo. For OpenLogi I get the wrong type of vibe code vibes and at the very least that means the project might become an unmaintainable mess in a year and that means I don't want to come to rely on it. At worst it means security vulnerabilities.
That being said: I wouldn't really trust SaaS to not be vibe-coded either. The only difference there is that SaaS likely has a financial incentive to stay maintainable so you could argue that helps.
> because they have the reputation.
reputation of what? sucking your soul dry?
I believe it varies; since it's merely a productivity tool, why not use it? However, if it's essential software that your business relies on, probably not.
> But do we still trust the open source software that is created now everything is vibe coded?
If I think tool/software would be useful I throw GLM 5.2 (now 5.3) at it and ask to do an audit ;-)
I don't see the issue. The beauty of OSS is that you can audit the code yourself and evaluate it according to your standards.
Do you trust all OSS repos coded before 2024? AI-generated code is no different than code written by a human with different levels of seniority, different levels of focus, or different levels of discipline. A human is perfectly capable of generating spaghetti code too.
On the other hand, it's also much less effort to do at least some rudimentary review.
Or make a fork with only the minimim set of parts that your agent cherry picks.
write code by hand then
great effort ? bro just sniff the traffic and feed into Claude.
> But do we still trust the open source software that is created now everything is vibe coded?
I guess my question is, why would you trust it when it is not vibe coded?
My suggestion is, have an AI review the code for you.
Code is simply math. It either works or it does not. Does your calculator work? Then use it. Does your AI agent of choice produce code that works? Then use it.
Or don't. It's up to you. You're as free to walk away as you are to embrace what's been made here.
So you trusted 22-year-old coffee junkies copying and pasting from stack overflow 10 years ago but when a language model does it it's suddenly offensive? We should be instead celebrating how we have a mini software company at our fingertips ready to fix the broken hobbyist shit we've never had time for. OpenLogi is a perfect example of this use case and that it's open source makes it more trustworthy not less.
Now if only there was one for Razer.