BIDI/LTR Toggle. #6474

I'm QA, I don't read much code, and don't really understand most of it. I became a LLM user (many models, many CLIs), and I decided to give the FOSS a shot It being FOSS means I can just try and fix it with my LLM - my needs was just one button switch. But then I go and read your contribution TOS: https://github.com/Zettlr/Zettlr/blob/96b78abc735669b1d26b7e5070d07941dcbca41f/CONTRIBUTING.md And I can communicate, I can write my own PR, or at the very least indeed approve of the texts... reply etc'. But code review? At most I can audit it with another LLM. I do not have the required qualifications. So I'm at a conflict - do I contribute to the project, maybe helping it? Or I just use it, no updates and don't even consider a contribution to a FOSS repo that requires a human code review by PR submitter?

"You agree to check and proof-read any text/code that AI has generated to ensure you understand it. We do not require you to provide perfect code; but we do require you to understand everything you wish to contribute to Zettlr and its ecosystem."

But I do not. I know it works, I know the LLM followed your docs, I know it's proper - but it's from use, and the LLM telling me how it did. My solution is hack-ish, but it works just according to my needs. Keep it to myself? You don't want any contribution from honest people who just want to contribue? Just making sure before I give up on submitting the work. No I won't claim it's mine - I just explained the need, refined the execution - but it's all LLM work. I don't write code, sadly.

This is more of a general suggestion - that you require that the LLM solution is audited by you, if it works, if it complied with the techstack and restrictions etc'. "We do not require you to provide perfect code; but we do require you to understand everything" - it should be the other way around. you do want perfect code, you don't want it to be decided if it's submitted by a pro or a total noob. The proof is in the code - not in the reviewer having CS degree or something of that sort.

No AI was used in writing this post, just so you see writing or reading is not the issue. It's just a somewhat of offensive demand that you understand everything the AI did. If I could do that, why would I use AI to begin with? I'd just be a coder, and use AI to help me code - and obviously that way I know exactly how it works. But if I'm "vibing" it - you don't want it? Even if it's good?

I can think of a better solution than requiring that every PR is submitted by a real software engineer. LLM's already stole most of your work (I mean, it's a plagiarism machine, it is known) - I don't consider it inferior to a real programmer. If it works, and it adheres to the rules - isn't that what really matters?

I have thought way too much about this indecent agglomeration of implicit and explicit accusations that you have spat in front of me here. So I'll be as brief as it is humanly possible in my response. Also, let the record show that I am primarily responding to this, since it is part of the public record, and I cannot let this untenable string of falsehoods stand without a rebuttal to set the record straight.

To pull the bandaid as quickly as possible: the rules you are so concerned about in the contributing guide have been written precisely with people such as yourself in mind. Sure, the rules say an awful lot about human communication and how we expect code contributions to proceed. But these rules are there for a reason, and you are not the first person with such a self-centered attitude towards open source projects.

There is a saying that rules and regulations are written in blood; and in our case it is the blood of entitled people such as yourself that they are written in. I am sorry, but you are by far not the first person to write such an impertinent piece, accusing me of demanding "perfect code" and why I would decline perfectly valid "vibe-coded" contributions, and why I act as if I expect a computer science degree for any contribution. Albeit, I will have to give it to you that you're the first who writes this from a perspective of a vibe coder; someone who, as you write yourself, lacks "the required qualifications," and whose mental vomit speaks volumes about how little you care to understand and interact with the broader ecosystem of this application.

You were very apparently incapable of even reading the very document you read, which spends just one paragraph speaking about code contributions, but offers a multitude of additional ways in which interested contributors who cannot code can make a meaningful contribution to this application. Yes, you apparently lack the required qualifications for contributing with code, so that won't be a good avenue for you. At least for now.

But I have a counter-question: why did you never consider to put your alleged QA knowledge to work to make the user experience of the app better? Why didn't you test the hell out of this app and open a dozen issues outlining all the — very real, and still very present — problems in the UX of the app? Isn't that supposed to be your job? Why do you feel entitled to make code contributions as a quality assurance tester, and why are you so shortsighted to even reveal this in front of the entire world to see?

Here's the thing: The requirements we place on code contributions aren't outlandish, and there are quite a few community members who said publicly that they don't typically code, but they have opened PRs, fixed issues, and their code is now part of the application. Some even wrote the code with the help of LLMs. So what's the difference between these fine people and your juvenile arse? Simple: these people care about the app. And if you care about the app, and genuinely want to make a difference with your contribution, then all rules can be bent. But your comment here is a prime example for why these rules are in place in the first place, and I have to thank you for not having actually opened the PR so that I had even more to do than I have already.

Here's something else you probably did not consider while spurting this fart into the comment box: I have spent a decade in working on this app. I hold this project and its future dear to heart. Do you believe that it is not a reasonable request of me to require code contributors to place at least some part of this care into their work as I did? In the end, my name is attached to this project, so if something goes wrong (and if you don't even try to understand what your code does, you will miss security-critical issues), I am on the hook. That's my responsibility. In turn, I ask you to show me and this community some respect. And the way you do that is not by barking alien accusations about the harsh rules in this repository out into the world.

No, what I am confused by is why you never, in your short train of thought that led to the decision that is this train wreck of a comment, considered to just join Discord, and ask how strict these rules are in the first place. Do you know what would've happened then? You probably won't be able to guess, so let me illuminate your path: I would have spent much less than this to tell you why I think you should at least try to retroactively understand the code your LLM has generated. And then I would have even helped you work your way through your contribution until the PR could have been merged.

But none of that is going to happen now, because instead of stopping for at least a second to reflect on your decision, you decided to give in to your intrusive thoughts and emptied the garbage can of your brain on here.

Let this be a warning: I'm not against LLMs. I use them every day myself. But I will not be gaslit by juvenile idiots without a fraction of understanding for the broader ecosystem and how an open source community works into what I do and don't do with my rules. I expect the same respect towards Zettlr's community, which is full of decent and amazing people, and don't be a dick. If you decide that this doesn't apply to you, then suffer the consequences.

Have a good day.

@nathanlesage

I want to leave one final clarification.

I did not submit code, open a pull request, or ask you to review or maintain anything. I read the contribution guide, recognized honestly that I could not satisfy its requirement to understand and personally audit generated code, and asked whether that meant I should refrain from contributing.

That was a question about your policy. You were entitled to disagree with my reasoning and defend the policy. You were not entitled to respond with repeated personal insults, degrading language, and public humiliation.

You appear to have mistaken my honesty for something shameful. I see it as a point of pride. I stated clearly what I know, what I do not know, and how the proposed solution was produced. I chose transparency rather than pretending to possess expertise I do not have.

Your response also appears to have been shaped by frustrations and experiences involving other people, rather than by what I actually wrote. Whatever anger those experiences caused, directing it at a stranger who asked you an honest question was neither fair nor proportionate.

The comment crossed the line and has been reported through the appropriate channels. I will leave it to them to assess.

I genuinely wish you and the project well. Perhaps the lesson here is that criticism of a policy is not a personal attack, and that honesty from a prospective contributor should be met with proportion, humility, and basic respect.

⚠ This comment was originally posted to the now-deleted duplicate thread #6475, created after the original was locked. It was never responded to.