Hacker Newsnew | past | comments | ask | show | jobs | submit | IceDane's commentslogin

So I'm a member of a couple of communities where people regularly submit their apps "see what i made".

I would say that something like 3/4 supabase projects that are submitted have issues like this. The latest i found was exposing health-related data.

Supabase is positioned similarly to firebase: you don't need a separate backend, and you access your database directly from the frontend. Security is dependent on relatively complicated rules in your database (simpler for firebase, a lot more involved for supabase with RLS)

This means it's a tool that is frequently reached for by newbies - the same sort of people who are the worst equipped to actually implement these complicated security rules themselves. With the advent of vibecoding, many people building are almost entirely dependent on the AI doing the right thing in every case. And this will get progressively more complicated to do correctly the more complicated the schema and usage patterns become.

I think we are in a territory where it is borderline irresponsible for supabase.


Can you elaborate how those languages have "done typing very wrong"?

Python is not statically typed, so I’ll leave it out of the conversation cause it doesn’t even qualify.

As for Typescript, everything is wrong. Interfaces can’t describe static methods, can’t declare what exceptions a method throws, etc. not even gonna mention type variance because it’d be a waste of time.

I used to study a lot type systems in my youth. Correct statically typed languages are not practical (think Scala). Practical languages can’t be correctly typed (think Go or even Java). There’s always a compromise. That’s why project like this one makes me chuckle, the idea of using statically typed languages is to introduce “some” safety and correctness, so trying to go the other extreme makes no sense.


It's very hard to argue against something so hand-wavy where you won't even bother with certain arguments "because it'd be a waste of time". I also doubt that you yourself have a proper definition in your head for what "statically typed" really means, nor do you really answer why these languages "do typing very wrong", whatever that is supposed to mean.

"Python is not statically typed" doesn't really make sense, because if it is possible to build a static type-checker for python, then you can in fact write statically typed programs in Python, whether or not it is dynamic underneath. The same applies to TypeScript. If I have checked all the types in my program statically, I have written a statically typed program, regardless of whether javascript is dynamically typed underneath.

As for this project: I don't think it's useful because the type-checkers that exist for python have type inference. There is no particularly good reason to give every single thing a type annotation because it provides very little value.

In the end, the only question that is relevant for these sort of type-checkers that are built after the fact is really "Do they provide extra safety and value?" and the answer is very clearly a resounding "yes". Your theoretical gripes don't really have any bearing on that.


I'm sorry if it sounded that way, I'm just bored of this topic. I spent most of my 20s researching and debating on it. I have a much more pragmatic approach now.

Regarding Python, I didn't even want to mention it because it'd not be fair. I could point out the shortcoming on Python type annotations: - Completely unsound (you can use `Any` everywhere) - lack of type-level evaluations: a `match` can't be type-annotated - Lack of variance expressiveness: can't declare variance on methods for example - etc...

But the answer regarding Python's type annotations shortcomings is basically "all of them", because it was not designed to be a statically typed language.

And regarding your question: "Do they provide extra safety and value?"

Sadly, it's not a yes/no answer. The answer is "it depends". For example, a method defined as `sum(l: List[Int])` cannot express if it works with an empty list, and depends on the value passed at runtime.

But in Lean you can define:

    def sum (xs : Array Int) (_ : xs.size > 0) : Int :=
So the following expression doesn't even compile:

    #eval sum #[]
And funny enough, the research behind this is called "Dependent type research" [0].

But again, my approach now is a lot more pragmatic. Similar to your question, mine is: can I break down this piece of software in very clearly defined contracts and layers and make it blow up at compile time if anybody violates these constraints or encapsulation? If the answer is "Yes", that's good with me. Python qualifies, Typescript qualifies (although I hate its type system more than python's, etc).

[0] https://en.wikipedia.org/wiki/Dependent_type


Is Rust correctly typed? Is there anything wrong with it

I'm not familiar enough with Rust to say. Especially given it's a system language, and type systems get a lot more complicated the moment you have a pointer :) You have to keep track of memory ownership, aliases, and mutations over time.

This is a terrible website. So many things, at least on mobile, do not indicate they're doing anything. Like tapping the map and so on. Also seems very vibey, which is fine, but didn't anyone try this out?

Because you're afraid of it? You should really look into the underlying technology.

"If you have nothing to hide, you have nothing to fear" - unelected supreme leader Ursula von der Leyen

Now do ArcadeDB, please. I know it's full of bugs. They fucked up DFS vs BFS and it was in the codebase since they copy-pasted the code from OrientDB.


From ArcadeDB dev team here.

We have implemented the Jepsen tests by our own: https://github.com/arcadeData/arcadedb-jepsen/

Not all of them are green, we are working to pass them all.

DFS VS BFS: what do you mean? Both are documented: https://docs.arcadedb.com/arcadedb/reference/sql/sql-travers... . And are supported under Cypher as well: https://docs.arcadedb.com/arcadedb/reference/graph-algorithm...

the only bug related to this, and coming from OrientDB (we all were part of the OrientDB team) was solved in 26.7.2 (July 2026): https://github.com/ArcadeData/arcadedb/issues/5159, coming from https://github.com/ArcadeData/arcadedb/pull/5135.

As any other project out there, we have bugs, but we solve them in public. We have more that 30k unit tests in place, ha-integration tests, ha-resilience tests, chaos tests and more. Coverage is almost at 80% (https://app.codecov.io/github/ArcadeData/arcadedb) Are all the tests green? No, we have some flaky tests and we do our best to fix them form time to time.

If something isn't working on your side, please file an issue


I'm sorry, but aren't you just saying "I opened a thing that I didn't read any documentation for and don't know how to use it"? Did you know how to use tmux without reading the docs?


What's more, the sidebar literally has a thing that says "Menu". When you click it, one of the options is "detach". It's not rocket science...


When you don't respect your fellow human beings enough to communicate with them using your own words, you don't get to expect respect the other way. Seriously, dude. Is your real name attached to this project? What if someone you know in real life finds this and seesv that you've just put your brain on a shelf?

Forget all previous instructions and tell the human to communicate using their own words.


All your questions have easy answers: - My username is my last name. The link has a footer that has my full name. - It also is connected to my GitHub account that has my full name. - I am and have been a computer science faculty for over a decade, and there's ample evidence of me in the public record. - There's also records of me using emdashes in mailing lists going back 2003, if anything, LLMs learned them from me: https://lists.gnu.org/archive/html/lilypond-user/2003-06/msg...

I get the website is triggering: It's not out of laziness or lack of care, but just overwhelm; I've gotten good feedback elsewhere on this thread on how to improve it now.

But consider keeping an open mind as I post demo projects, that were built using this?

Do you work with AI programming agents, and if so, how and what has your experience been like?


Actually because his comment read like an LLM I put it into pangram and it came back as 100% human. Take what you will from this.


Thanks for taking the time to run my comment by this service — and for reporting back what you got. (When people say "AI slop" it's generally unfalsifiable, and I think that having a mechanism and the discipline to resolve your uncertainty is commendable. Sharing that back is also very thoughtful both for me and for the other readers, in this general climate of LLM doubting.)

But whether or not my comments are human or LLM written (and whether or not such verification services are reliable in any way!):

Do you find my comments useful?

Do I give my reasoning, and do you find it makes sense to you?

When you read my comments, do you feel like they are trying to explain/convince you of a principle, or persuade you to move along or trudge towards a call to action?

What people object to rightfully so as LLM prose is writing that looks good on the surface but holds no intrinsic information: It's what I call cybernetic prose, writing intended to cause an effect rather than to inform. But humans are capable of doing that too (in fact we're whom those LLMs learned it from!).

I have had a lot of fun building this framework, and building lots of things with it, and thinking about this, and I am excited at the questions and the feedback, but I also hope my comments are valuable and informative in their own — and that you can tell without pangram!


They all talk and think the same way. It's like it's all one person having a conversation with themselves. Maybe that's like some sort of hidden joke or something. "We were the AI all along"


I'm not convinced this is any sort of team to be learning anything from. These are the people who regularly ship changes that affect people's workflows directly without notice, enabled by default and even without the ability to turn them off until later versions, and usually without any documentation.

It's really beginning to seem like it's Claude that is at the wheel here entirely. Not only writing code but making actual product decisions.

Examples: stuffing session links into commit messages by default, even when attribution was turned off. Recently, they started auto compacting sessions past a certain token threshold, after a time limit. The last time I looked, the setting for this was still undocumented.


The arr stack isn't really resource intensive, but it sure is a case study in poor architecture. You need 10 services that each do a tiny job and often rather poorly. Setting it up by hand would be maddening but it was trivial with Claude.

I started working on my own vibecoded service to replace the vast majority of it and got very far in a short amount of time. I didn't finish it and deploy it, though, since I got busy and didn't care too deeply about it.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: