HN user

DesertVarnish

132 karma
Posts3
Comments28
View on HN
Nano Banana Pro 8 months ago

Largely but not entirely a data problem; specifically poor captioning. High quality captioning makes such a big difference.

Difficult engineering problem but working from first principles suggests that the energy requirememts are not insurmountable. The roundtrip efficiency is worse than batteries but much better than photosynthesis.

Terraform Industries (and others, like Synhelion) has a plausible if slightly optimistic target to be price competitive with fossil fuels for methane in the early 2030s.

Some places with very cheap to extract hydrocarbons like Saudi Arabia may be able to compete for a very long time, but there are many futures where most of humanity's hydrocarbon consumption (including the ones used for the chemical industry, plastics, etc) derives from atmospheric carbon.

And this can happen fast, the world (mostly China) has developed a truly massive manufacturing capacity for PV.

My wife has been working on a pixel accurate recreation of MacPaint on the web, and one of the things that I've learned from on-and-off helping with it was just how much of the magic in it is actually just handled by QuickDraw.

Also, the text styling has a lot more complexity than you would expect.

My understanding is that fluoridation of municipal water supplies serves a different purpose than fluoride in toothpaste. The fluoride in the water supply is low concentration and is intended to cause teeth to develop with stronger cavity resistance, toothpaste is a much more superficial and immediate measure.

They really aren't the same.

Crypto didn't and still doesn't have the same immediate utility. The value proposition just wasn't there to justify the money and attention it was getting. Bitcoin in particular was a prototype that got mythologized into being "digital gold" despite it's many, many technical limitations.

Diffusion models and LLMs work today and make possible things that were science fiction five years ago, and have shown tremendous and exciting progress in the past 18 months.

Haskell is purely functional, but if you make ample use of do-notation it's very easy to write in a familiar imperative(ish) style for day-to-day tasks. Most learning materials won't emphasize this but large production Haskell codebases that do a lot of IO usually end up looking a lot more familiar than you might expect.

If you have interest in both OCaml and Haskell, you might also consider just learning Rust which borrows aspects from both of them. Rust traits aren't quite as powerful as Haskell typeclasses and its module system is definitely more limited than OCaml's, but if you write it in a more functional style you'll get exposed to a lot of similar patterns.

A lot of things in Haskell never really clicked for me until I spent some time getting comfortable with Rust.

For instance, using the features of the language to describe a nested structure of components, each of which has different dependencies that are a subset of its parents dependencies

I may be misunderstanding you, but this seems tractable with React.memo. Merely encode the dependency or lack thereof in the propsAreEqual predicate.

Most of the time an application built using web technologies is slow, it's not because the web's UI layer (HTML/CSS) is intrinsically slow.

A desktop app written in Tauri that punts most of the heavy data processing to Rust with a UI layer written in well-optimized React or Solid.js is really plenty fast. In practice what we get with most Electron apps is not especially well written React with frequent coarse grained re-renders, inefficient data processing, frequent allocations causing GC pressure, and excessive JSON RPC. The problems here aren't HTML/CSS + JS for basic event handling, it's everything else about how the software is developed.

Is it actually standard practice to use nitrogen fertilizers with soybeans? Soy is a nitrogen fixer and I've read that nitrogen fertilizers often reduce yields for soybeans because it interferes with nodulation and undermines that plant's nitrogen fixing capacities.

I'm another Mercury employee. When I joined there wasn't much of a specific onboarding process at all, beyond being given a few simple tickets to work on and pointed to some documentation.

I had written a very small amount of Haskell before, but IIRC correctly I think I made my first commit only a few weeks after joining. In general even people without an existing strong FP background seemed to become productive pretty quickly.

It's interesting to see companies showing up in this space (e.g., https://hash.ai/), many of them based out of London. I'd consider applying if I was based in the UK, but I'm curious about what inspired the founding team to start the company. Can you share an origin story?

Rust 2021: GUI 6 years ago

Packaging gtk for systems other than Linux can be a bit of a nightmare, and the experience on platforms other than Linux and desktop BSD leaves a lot to be desired.

Rust 1.45 6 years ago

println is a macro in Rust because 1) the language currently lacks support for variadic functions and 2) being a macro allows the format string to be parsed at compile time.

At one point, Roam supported importing org-mode files (my team actually asked them to implement it back in spring of 2018). The product vision has changed a lot since then and I'm not sure if they still support it, but if there's interest I wouldn't be surprised if they brought it back.

I may be wrong, but insolation probably isn't a substantial factor in keeping cores molten because of how small the energy input is. Venus does have extremely active volcanoes. There's suspicion that it might have a stronger magnetic field if it were spinning faster (it's rotational period is currently longer than its year).