codingdave

I've been working with government entities for document management for 15 years now, writing software that does exactly what we are talking about - versioning, tracking, publishing legislative records both through the authoring and approval cycles and then to the public. I have not worked at the federal level, but with small government - school districts, small towns etc.

Those years tell me 2 things:

1) Yes, absolutely version control for legislation is critical.

2) That doesn't mean Git is the right answer.

Since I don't work at the federal level, I don't know what system they use, but I do know that their workflow is already based on specific changes - they will write up the changes to a document, then apply the changes and publish a new document. Those documents are posted on congress.gov. So they do have a versioning workflow and public accessibility, just not the same style as software devs are used to. I'm sure it could be improved, but I'm not sure forcing them into our working style is the right answer.

The real goal here is not "Use Git". That is an implementation detail, and one that has also been presented and discussed quite a bit in my niche of the industry. Mostly, it is unwanted. Or was unwanted the last time the discussion went around. Git makes sense to software engineers, but not to the general public, and not to any of the document authors I've worked with. Maybe we can add a UX layer to fix that, but that isn't the bigger point...

What we want is to work towards the actual goals: For the tracking to be more detailed, with more transparency including not only authorship of the change, but the source of the change. And presumably, some commentary as to why the change was made. Then all of that should be available to the public.

If git works as a data structure that supports a solution that meets those goals, fine. I'm not anti-git. But when designing solutions, we need to start from the goals of the system, not the data layer. And we need to design solutions that the document authors will actually use, not force them to work like us.

I wrote something similar, but I had a different take on it. I let you paste in the text, then it steps you through breaking it apart, helping you to determine what each bit of text was, and a drag/drop UI to diagram it. But I approached it with a deliberate zero-AI approach. The idea is that careful reading of text is a skill, and many people are losing it. So I wanted to provide a playground for people to practice it.

I also figured that people who write badly might see what is wrong with their writing when a diagram shows just how off their words really are. My teenagers think that the resulting diagrams would be a brutal yet effective way to combat the misinformation online.

In any case, I wrote it. But then I stopped because I wasn't sure where to really take it. But it feels a bit validating to know that I'm not alone in seeing the value of breaking down written text.

It is almost like LLMs have gotten over their initial tells and now the tell is that they write these walls of text. This article reads just like the walls of text that my coworkers at one of my gigs paste all over the place. Any single sentence looks OK, but the overall wall has no flow. No cadence. No difference in tone between stating facts and drawing conclusions. Headers, lists, and charts thrown in for variety.

I'm not sure - maybe humans can write like this, too. But it is the written equivalent of a monotone voice, and I find it very difficult to keep enough attention to read such articles, regardless of their authorship.

This is the same answer as growing a for-profit product. You need to stop spraying random marketing attempts, and look deeper at who uses it today, and why. What problem does it solve for them? Who else has that problem, and where does that audience actually spend time - both IRL and online. Then you craft a message that clearly and concisely show them that you understand their problem and have a solution available, and you share that message in the specific locations that your specific audience knows and trusts.

> I'd be pretty scared.

Scared of what, exactly? It looks to me like the site is primarily static content, so operational costs should be reasonable. He could scale down to just static content, and even if traffic decrease, it is not an existential crisis. Or am I missing something?

I know 2. That is out of the 3 people that I live with. Sure, this household is probably an outlier, but I think people inside the tech bubble do not realize the depth of the anti-AI sentiment from outside the bubble.

From the HN Guidelines: "Please don't use HN primarily for promotion. It's ok to post your own stuff part of the time, but the primary use of the site should be for curiosity."

You've now posted this 3 different times in the last couple days.

You still tell people to download a zip and run that code. And on some other page you linked from your github it says to enter your AI key.

An OS doesn't need to be downloaded and run in Windows. This continues to be sketch AF.

> Capture a whole chat with one click: the full transcript, a summary, goals and decisions.

The full transcript? Yikes. That is just going to burn the tokens down in the next session really quickly.

Is it so hard to monitor your session, and just ask for a context snapshot when you hit 90%?

I dislike that. IMO, if the content cannot stand on its own, it did not need to be posted. A comment that adds something to it, to initiate a discussion, is fine. But a comment that just says what that post is? That is neither interesting nor engaging.

Such comments also encourage the idea that people "own" the threads they start. I've seen people get mad when people take the conversation in a direction they did not want, or add instructions telling people what to talk about and putting constraints around discussion. Such things stifle interesting conversation.

So I land heavily the opposite direction. Just post, and leave it at that. If the community finds it interesting, discussion will ensue.

Now, Show HNs are their own thing. An initial explanation is absolutely appropriate for a show HN. But it should be part of the post, not a comment.

Intriguing concept, but I don't really trust the listings site.

You need to order a car, then email them what you want, then they'll find a car for you. wat? Why are there listings if you need to email them to have them do a separate search?

Then, if they do find one for you, you need to pay $2000 before any inspections happen, which they say is refundable... but I am not sure I trust that.

And then if you follow the process that far, you pay $85 dollars for inspections, which is not refundable. Sounds like an ideal loophole to let them collect that $85 many times on the same car by just telling people, "Sorry, it failed inspections... want me to search for another?"

"Trust me bro, just download this zip file and run it on Windows."

Maybe OP is malicious. Maybe they are not. But either way, someone claiming to have vibe-coded an OS in 4 days and expecting people to download and run it is a quick and easy "No." from me.

I'm not sure this presentation is complete if they don't even mention Eric Roberts: https://www.imdb.com/name/nm0000616/

Never?

Testing has always been a game of figuring out how much is enough to produce a stable product, without burning massive amounts of engineering time with testing. LLMs have not changed that. They will never be perfect, and that is OK. Same as with humans, you need to define your own comfort zone for quality controls and accept that there will be bugs. Fixing those bugs will be part of your process. If you get too many, increase testing. If you get none, you just have not yet discovered what was missed.

The problem I'm seeing is not with vibe coding itself, although it is absolutely the LLMs that empower people to fall into the actual trap: Failing to do any research whatsoever into their market and subject area. They see one surface-level problem and instead of digging into the problem space to gain a deep understanding of what would make a solution good or bad... they just code. Vibe code, code by hand, somewhere in the middle, whatever - the current trend is to code first and ask questions later. Some people always did that, but these days it seems like everyone who has not built products before is doing it.

We're seeing it more and more. Someone whips out an app independently, launches it, then says, "How do I find users?", "Why does nobody use my app?" With the answer always being the same - they skipped the step of actually talking to potential users before coding.

So we see what is being described - lots of throwaway projects.

You might be surprised just how many homes run on a well and septic system. Day-to-day, walking into my house, you wouldn't even know the difference between my home and one in the city. We also don't need to chop firewood. My completely standard modern furnace runs on gas, just like ones in the city. Admittedly, I need to supply the gas, but that infrastructure is also well-solved. I just need to make sure to order more when my tank is near empty, and a truck shows up and fills my tank. I pay them instead of the utility companies.

What is more likely to throw you off is services like trash. There is no centralized trash service that just picks it up. You need to handle it. Most of the time, people still hire someone to do it, so it also ends up being the same as living in the city, but you need to actively seek out a solution. Or DIY - I know plenty of folks around here who don't mind taking their trash to the dump once a week or so.

And you cannot order food for delivery. That simply does not exist here.

Not that you are wrong that if we had to go back to living like we did hundreds of years ago, it would not be comfy. But the reality of living out in the country is not as different as you seem to imagine.

For context, in case anyone is unaware, there was a CVE reported this morning: https://nvd.nist.gov/vuln/detail/cve-2026-109618

Or, I just saw that a whole domain was registered to explain it: https://www.php2shell.com/

I bought a small forest. They are not actually expensive in my area. I love walking out of my door when work gets frustrating and walking through it. So yes, the answer to your question is definitely "Yes." But there are middle grounds between dedicating your life to the modern tech dystopia vs. completely walking away. Work remotely. Find a home that is in a natural setting. Go outside and enjoy it.

> without really knowing what you’re doing.

That is an odd caveat to throw on the end. Sure, to avoid frameworks you have to know what you are doing. Is that really such an outrageous expectation?

> It is away around the legal system to retain wealth.

Yes, absolutely, and it goes back farther than criminals. Why do you think everyone's grandma has a jewelry collection, while less of the youth do? Because before the 70s, women were not allowed to have their own bank accounts. Their gold and jewelry was the only type of wealth they could keep for themselves.

I fully agree that your web site is probably the problem. In addition to the other comments, which are completely correct, you have no intro at the top of the site. When you load the page, you just get some logos - there is no explanation of why they are there. Did you design them? Do you just like them? The icon on the right that scrolls down to your content is non-standard to the point that I don't even know what it is. So the whole site just doesn't communicate very well.

And even if that didn't turn me off as a potential customer, I'd be seriously questioning whether I want to hire someone who claims to be a designer, when they didn't even put together a design for their own site that effectively tells me who they are.

Basically, I think you are too close to your own site design to see just how flawed it is. Sorry is that sounds harsh. But I'd re-do your entire approach. Don't put up a portfolio and then try to add commentary at the bottom - put up a easy-to-read, concise explanation of what problems you can solve for people. In this scenario, you are the product, and you need to market yourself just like any other product - tell people right off the bat why you will improve their life. If you don't do that, the portfolio is meaningless.

I'm not detecting much of game here. It feels more like a homework exercise. ID the attacks on the board. OK, done. Do it again. Repeat ad nauseum.

Because they did. No matter what tools they used, they are accountable for the resulting product. People who use CNC machines to create physical objects don't reject the idea that they made the object, even though the machine did the work.

Don't fall into the trap of anthropomorphizing Claude, or any other LLM. It is a tool. The person using it is still the creator.

I'm not a musician, so my question may have an obvious answer, but... what problem does this solve?

I'm thinking back to when I was in band as a kid, and the idea of having to recognize tiny number inside the notes seems to make it more difficult to read, not easier. Even if not trying to read them, the visual is noisier.

Is the idea it offers help to new musicians, but experienced musicians can ignore it? Someone good enough to sight-read sheet music can disregard the new noise?

It has been happening for my entire 30+ year career. This is not a new trend. LLMs are the newest flavor of it, but low-code/no-code tools have been around for decades. As well as clever people who just figure out how to create apps.

They have a really standard lifecycle: They get built. They work for a small department. If it stops there, it just keeps running until the person who built it moves on, then it gets trashed and replaced with something else. Or, if it grows beyond the department, that person's life starts to get taken up in maintaining it, it fails to scale, and IT takes it over (or replaces it).

The question is not does it happen. It happens so often that there is even a term for it: "Shadow IT". The question is not even whether or not you can launch a product to stabilize the lifecycle. The real question is - should you? The ability to create an app doesn't grant the knowledge of how to make it secure and fit within regulatory bounds for your industry. It is quite possible (and happens often) for someone to make an app for their department that works great and everyone loves, but exposes the company to real regulatory and legal risks.

That is the real reason IT shops have someone to watch over such apps. As you are clearly asking this question in the spirit of self-promotion, you really need to show that you understand the real problems and risks with departmental apps. "Do people build them?" makes you sound like you are completely new to how corporate IT works, which instills no confidence that you have built a viable product.

> Every company runs on knowledge it can't find.

Whenever I see a post start with a premise that simply isn't true, I know I'm about to read some marketing slop about their product.

What is frustrating about posts like this is that it doesn't have to be that way. Institutional knowledge is absolutely important, and when it breaks down, it does cause problems. Say that. Tell the truth. Talk about the problem you actually solve. Otherwise nobody is going to listen to you, even if you have a decent product. (I don't know if their product is decent or not, the huge red flag of the first sentence made me not want to even check.)

How did you measure that accuracy?

Actual title: "Trump Calls Anyone Who Says ‘AI’ Instead Of ‘Super Intelligence’ An ‘Enemy’"

HN Guidelines: https://news.ycombinator.com/newsguidelines.html

Specifically: "please use the original title, unless it is misleading or linkbait; don't editorialize."

Even the article is specific in that it is just reporting on something Trump said. And even Trump did not say it made people an enemy of the US, it said it made people an enemy of his White House.

> I searched for the song on Pandora. We hadn’t checked it before because it’s no longer a major music service

I find that take interesting. I use Pandora to this day as my primary streaming source, as it has my vibes down perfectly, and no other services has that many years of training into what I like to listen to. But even more important, it is almost always the music source I see when I'm restaurants. I prefer small mom-and-pop type places, so maybe I'm only seeing a niche.

I'm not trying to argue it is a leading platform, as it clearly is not. But it is fascinating that people searching for music don't check what was a leading platform 10 years ago. Especially when the music in question is older than that. And the search started when it was a leading platform.

Actually that last point really confuses me - why would today's platform popularity have limited the search 15 years ago? That part of the story makes little sense. If this search began back when Pandora was at its peak, why wouldn't they have started from there long ago?

> the settled law is that such an LLM reproduction is 100% legal.

My understanding was that re-invention without copying any code is legal. But scraping code from the browser and re-using it is not. I'd love to know how that plays in the courts with LLMs, as their entire model comes from copying code as training material, not writing new code from scratch.

Weekly write of my files to an external HD. I swap the HDs monthly with off-site copies.