conorbergin

Very funny to put a bibtex citation under such a small piece of work

Redefining commonly understood phrases to mean something else in your own little world make you look ignorant.

I've being trying out SQLite for a side project of mine, a virtual whiteboard, I haven't quite got my head around it, but it seems to be much less of a bother than interacting with file system APIs so far. The problem I haven't really solved is how sync and maybe collaboration is going to interact with it, so far I have:

1. Plaintext format (JSON or similar) or SQLite dump files versioned by git

2. Some sort of modern local first CRDT thing (Turso, libsql, Electric SQL)

3. Server/Client architecture that can also be run locally

Has anyone had any success in this department?

If you are a hobbyist or small business in desktop manufacturing you are basically forced to buy Chinese products.

I have never owned a Prusa, but I have owned several Creality and Bambu Labs printers, because I could get the same utility at half the cost. The same goes for soldering irons, linear actuators, oscillscopes, etc. I still buy European hand tools (Knipex, Wera, etc) because I know they won't break in a year, so they are good value in the long run.

Often the choice is whether to buy a used, last generation tool of eBay, or a brand new next-gen tool from China. The choice depends on how flawed the Chinese implementation is and the gap in utility between the generations.

The main problem with Chinese products is the lack of accountability. The same product will be sold under multiple brands, or by dropshippers, and you have no idea who actually made it, there are some strong Chinese brands that buck this trend, i.e. Bambu Labs. When you buy western tools you are buying peace of mind, something I can't currently afford.

I think everyone has this reaction until they start using it, then it makes perfect sense, especially when using editors that have multiple cursors and can operate on selections.

I would rather rely on a bunch of sanitizers and static analysis because it is more representative of the core problem I am solving: Producing machine code. If I want Rust to solve these problems for me I now have to write code in the Rust model, which is a layer of indirection that I have found more trouble than it's worth.

I think the problem with this attitude is the compiler becomes a middle manager you have to appease rather than a collaborator. Certainly there are advantages to having a manager, but if you go off the beaten track with Rust, you will not have a good time. I write most of my code in Zig these days and I think being able to segfault is a small price to pay to never have to see `Arc<RefCell<Foo<Bar<Whatever>>>` again.

Monads are the poor man's early return.

Very few people are too “advanced” to be challenged by a sufficiently difficult undergraduate degree, ridiculous thing to say imho, I went to a UK university this decade and I can give you a laundry list of issues more significant than “exceptional students being slowed down”.

Oh absolutely, I flatter myself to think I use them for "serious engineering", and I am well aware of my debt to Warhammer players who don't want to pay Games Workshop prices.

I don't think the fact that you can make fairly serious mechatronic devices with pocket money can conceivably be a bad thing for engineering as a discipline. However this does mean there are a lot of people that own a 3d printer that will never be good engineers.

Zig is becoming a real do-everything tool for me. I was learning gpu programming recently and found you could compile it to SPIR-V!

I've been using it for a while and I honestly don't even check the output until I'm done sometimes. I think it's more important to make good preliminary sketches and have a good idea of what you want to make, checking the output every time you change a dimension isn't that useful.

I see you still have to add a fudge number to stop the faces intersecting when doing a difference boolean. I'm afraid this is still strictly worse than Build123d or any other DSL than wraps OpenCascade.

I could not imagine trying to design a 3d part without fillets. I use Build123d mostly, and have even gone as far as using the Open Cascade library directly, but if I had to choose between FreeCAD or OpenSCAD/Solvespace I would rather work around FreeCADs jank than give up fillets.

I think someone says this in every HN post involving CAD, but the reason FreeCAD is "buggy" and Solvespace is small and fast is Solvespace has a fraction of the power. FreeCAD uses the Open Cascade kernel, which can do complex 3d boolean and fillet operations, not having these operations severely limits the geometry you can create, and you will run into walls very quickly using Solvespace, OpenSCAD or anything else with a hand rolled geometry kernel. Even commercial projects use an off the shelf kernel, they're just difficult to write.

Interesting that you talk about using shape optimization leading to more manufactureable results than a continuum approach. What sort of manufacturing do you envision, welded tube frames, cast parts? The weird organic results you get from the free form simulations are often pretty easy to 3d print with something like SLS, as long as they don't have voids and thin features. The constraints on other manufacturing methods are a lot more complex.

Bach's fugues are amazing, they sound great on harpsichords and organs, whereas some piano music sounds very wrong when it is payed without all the volume modulation, the fugues are carried by the interactions between the melodies, the counterpoint.

Your prompts are very long for how simple the models are, using a CAD package would be far more productive.

I can see AI being used to generate geometry, but not a text based one, it would have to be able to reason with 3d forms and do differential geometry.

You might be able to get somewhere by training an LLM to make models with a DSL for Open Cascade, or any other sufficiently powerful modelling kernel. Then you could train the AI to make query based commands, such as:

  // places a threaded hole at every corner of the top surface (maybe this is an enclosure)
  CUT hole(10mm,m3,threaded) LOCATIONS surfaces().parallel(Z).first().inset(10).outside_corners()

This has a better chance of being robust as the LLM would just have to remember common patterns, rather than manually placing holes in 3d space, which is much harder.

I agree with learning React, it pops up in all sorts of places, vscode extensions, the windows start menu apparently.

But yes, mixed-rendering, server functions, meta-frameworks, etc. You are in real danger of spending years learning all this stuff and still not being any closer to knowing how computers work.

When I made my first website I read things online like "svelte is better than react" and "solid-js will be the next big thing", and I thought this was an important part of web development. Newbies should read MDN and ignore everything else.