HN user
Building Ilograph: https://www.ilograph.com
Ilograph diagrams can be exported to standalone HTML files (like this example: [0]). Coincidentally, they also have slides (kind of) and clock in at around 560KB.
[0] https://www.ilograph.com/demo/intelligentDocumentProcessing....
Some might quibble with using AI for research, but this at least feels directionally correct. I'm not sure if ratings/MPAA is the right avenue to fix it, but the sheer laziness of manufacturing stakes by catastrophe instead of through characters definitely feels like its gotten worse.
and i would say json/yaml is not a dsl
But you can have DSLs that are json/yaml, is my point.
PlantUML is a real dsl.
PlantUML is a DSL that isn't json/yaml. That doesn't make it better, and you can make the argument that it is worse because the tooling around it won't be as good.
the premise is that because of these few existing dsls (like PlantUML mentioned) my "new dsl" will be equally effective. PlantUML has millions of examples in the training data, my new dsls are not (*specially if its not json/yaml* or just function chain based)
I can confirm that having a DSL that is json/yaml helps a ton. Kind of like static type checking, it eliminates entire swaths of syntactical errors, allowing the LLM to focus on the semantics.
because if then you are going to ask the llm to "compile the dsl to implementation" we are back to square 1.
I think this is an edge case; 99% of the time you (and/or the LLM) would have access to the implementor so it wouldn't need to do this.
I can’t understand how we got to this place with “app culture”!
The short version: ad blockers work on browsers but not apps[0].
Thanks for the heads up; that shouldn't have been happening. Please try again.
Hi, author here. The app itself (Ilograph[0]) is fairly old; its first ShowHN was in 2019.
These particular yt-dlp diagrams were an experiment in getting meaningful sequence diagrams from a code repo using Claude. If you open the black tab on the left, you can see the source of the diagrams. Claude was able to generate most of this, albeit with a good amount of back-and-forth to get the details right, plus some manual tweaking in the end.
Calculator / Rubik's cube / game of life apps should be very close to 100% identical, right? I don't see the point of asking an AI for one of these when there are dozens (hundreds?) of repos that all have exactly what you want.
It's a bit like a (simplified) version of Ilograph[0]. Are these editable, or strictly LLM-generated?
Consider trying to follow along with this app [0] without the transitions. Without them you'd experience much more cognitive load to staying oriented.
5 years later...
People who wanted to ask a specific question now won't have that option. Instead, they'll simply be shown whatever Google thinks is most relevant to them at that moment. The "Chat" UI we've grown so accustomed to is on its way out.
Gotcha. So this is more for diagrams of new systems than existing ones. Our maybe a blend of new and old.
I’ve found that Claude can make decent mermaid diagrams pretty reliably.
When doing this, do you give it a codebase and ask for a diagram, or is it more of an iterative back-and-forth where you describe the system to it.
"OAuth 2.0 authorization code flow with PKCE as a sequence diagram — user, browser, app server, auth server, API"
If you do a Google image search for "OAuth 2.0 PKCE sequence diagram" you get good results also. Maybe if you ask for something more esoteric this becomes valuable? Of course, that also makes hallucinations more likely.
Interested people will be following maybe even expecting to see them.
But they won't. That isn't how modern social networks work, and X definitely isn't an exception. The chronological feed of people you follow is long gone.
You are defining the goal of Ilograph to a T.
I mean HN modifies headlines all the time. Sometimes hours after the fact. News sites themselves A/B test headlines constantly. I don't really think there is any "contract" to speak of.
I can see this being a net benefit if it's limited to re-writing clickbait headlines.
I'm very curious as someone who also offers Mac software but not on the app store... why is not being on there a dealbreaker?
There was a time around 2016 where you weren't allowed to write a React application without also writing a "Getting Started with React" blog post. Having trained on all of that, the AI probably thinks React is web development.
You'd be doing way more harm than good. The battle between ad networks and unscrupulous website owners using bots to fake ad clicks has been going on forever.
You would probably just start seeing worse and worse ads [0]. Legitimate ad accounts would stop bidding on your profile so you'd be left with only scam ads.
[0] https://www.theawl.com/2015/06/a-complete-taxonomy-of-intern...
Might as well go all the way and call it "Microsoft 365 Copilot: Best App for Office"
I'm not sure, but starting with the ads that appear with most popular searches isn't a bad idea per se. It's a bit like sending law enforcement to protect popular areas.
Reuters: Restaurant hides unsanitary waste from food inspectors by hiding it in dumpster.
“not findable” for “regulators, investigators and journalists.”
but it's unclear from the article what the source or context of the quotations are.
Good point, this quote could just be painting their actions in the poorest possible light.
Writing a Typescript program that takes external input but has no runtime error checking is already a mistake, though.
If it's a service, yes, and that's true no matter what technology the service is using. If it's a library, no, because...
and if you write type assertions without ensuring that the assertions are accurate, then that's on you, not Typescript.
That's on whoever is using the library, not the library author. If the library author provides type definitions, and you as the consumer choose to ignore them, then it's on you.
I guess I'm trying to emphasize the distinction between information in the repo (code) vs. information elsewhere (discussions) that the model looks at.
The lack of a pipeline operator in JS (and JS libraries like lodash) has also been discussed online a lot.
From a year ago: AI can't diagram codebases, and it isn't even close [0]
[0] https://www.ilograph.com/blog/posts/diagrams-ai-can-and-cann...