sprobertson

I have used OpenSCAD not CadQuery, to me (as a vim person) the appeal is not having to do a bunch of mousework. But it doesn't hold up for long, at some level of complexity it is certainly easier to just go learn SolidWorks/OnShape/Fusion.

This in combination with making the tool call / json response itself also "batched" is a good pattern. Instead of returning a single `{english, foul}` object per record, pass in an array of records and have it return an array of `[{english, foul}]`. Adjust the inner batch size depending on your record size and spread the rest over the batched API.

Feedback:

* Font is very hard to read, needs some aliasing adjustment maybe?

* The dropdowns seem to be simple prompt injection, so sometimes just selecting "NES" as the color scheme will add an NES to the image. Similarly, the default "anime" style makes it looks like anime and not pixel art

As far as I can tell the left & right bubbles UI was first seen on (or at least first popularized by) the SMS app on the original iPhone in 2007.

As another commenter suggested it seems the point was to associate it with the placement of the right-side enter button, as the message would animate up and to the right.

Got it, I have no experience with the "resources" part of MCP but this does seem like a good use case. I could see something like job description -> LaTeX -> PDF being nice too

I like the concept, but I'm curious why MCP is better here (for something purely informational) over dumping a bunch of context in the prompt

what's the evidence that it's backdated?

The em dash is also pretty accessible on my keyboard—just option+shift+dash

which of those words are large?

FYI on the topic of scroll position, it seems inconsistent between history navigation. For example scroll to the very bottom, click your Internship blog post, and go back - it ends up somewhere towards the end but not quite. (Chrome Mac)

On mobile just clicking the other blog post takes me to the end of that post. (Chrome iOS)

Don't need any special tech or geometry, just light. If it's relatively darker in the watched-from areas it's harder to see in than out.

Based on some searches looks like all but one of the testimonials are fake

That detail and then sub-sub-page preview also hooked me, great work

Strange capitalization of atm as ATM in the HN title, but great tips in there

Hmm it still feels apt, reverse-engineering anything is effectively starting at the last page of the book.

That's how I solved #3, still trying to reverse-engineer the correct solution.

I got some confetti after an unknown number of clicks, but this doesn't answer the question of "how good is my color vision"

Great demo, straight to the point. It might be nice to have some kind of feedback mechanism if it can't find the element on the page, or if it's partially cut off. For example, I changed the GitHub profile to my own (https://github.com/spro) in the example and it doesn't scroll down far enough for the whole image. I imagine in general it would be nice to scroll to an element ID (or even element description using the vision models) instead of a hardcoded value.

Side-note: The comment for the frequency graph is wrong, it mentions stars instead.

First impressions matter, if the experience is this bad you're probably waiting a year to come back anyway.

Nice! Much better.

Couldn't find the one thing I wanted to see after seeing the preview images, what it actually looks like in context / on the screen(s).

I use both in vim - if the selection is tricky use v or V to define the selection first, but most of the time "verb object" makes sense to me since that's how english works (tho could see it being unintuitive for non-SVO languages). "delete the next word" or "change inside the parentheses" roll off the brain and into the keyboard

Wow $6000/month for a $500/month subscription is a bad deal!

The information is freely available, this just makes it more readable

https://www.sec.gov/cgi-bin/browse-edgar?action=getcurrent

I agree complex conditional LangGraph setups get pretty tedious after a certain point - though you claim to not use graphs here, but isn't that essentially what returning the "next task" does? The graph isn't explicitly defined but it still exists implicitly if you trace through all the tasks and next tasks.

Would be interesting to see a complex agent implementation in both Flow and regular LangGraph to compare maintainability.

SEEKING WORK | Remote | Seattle

Technologies: TypeScript, React, Node.js, Python, PyTorch, Go, PostgreSQL, Redis, AWS, GCP, LLMs, RAG

GitHub: https://github.com/spro

I’m Sean, a full-stack web-focused engineer with 15 years of freelance experience. I specialize in:

* Rapidly building MVPs, prototypes, and scalable greenfield projects

* Handling the full development pipeline, from ideation to design to deployment

* Scalable and maintainable software architecture for future growth

* Clear communication and a results-driven approach

Some things I've done:

* Built a platform for a solar company supporting over 1.5m sales interactions

* Built a query engine for structural engineering data serving 50k monthly requests

* Built a novel intent parsing model for home automation before LLMs were around

* Authored some of the first official PyTorch tutorials

If you’re looking to build something new, validate an idea, or integrate AI into your software, contact me via [username] at gmail with "HN Freelancer" in the subject.

Besides the tool itself that website typography is excellent. Guess I'll have to use Playfair in my next project.

Edit: One thing I'd like to see with the basic syntax example is fiddle with your default dates to make it more obvious that the span is a span. At the time scale it is now, it just looks like another dot.

Seems like a great resource but I would have hoped a site focused on UI would have proper z-indices for the main menus.

I think Swift already does this but I appreciate good type inference. Writing functions in a Pythonic/dynamic looking way (i.e. without a bunch of type declarations unless you want them) but still actually statically typed and enforced at compilation/runtime.

Very cool, how does it work? (even if just at a high level)