Here's an example of what this might look like: gist.github.com/brendanzab/3...
You could even consider dropping some local type annotations if they might be easily inferred from context - there's no reason to stick to the limitations of C for this kind of thing!
Tbh I’d probably prefer an ML/Pascal type annotation syntax (as seen in Rust, Typescript etc), as I find it more readable, giving emphasis on the identifiers vs the types.
I find types really helpful for understanding tbh! That way I don’t need to run type inference in my head to figure out what data structures are being used.
The trouble is that once you do that, it makes mutually recursive datatypes really challenging/annoying.
I do wish OCaml was less weird about it though - raising an ambiguity error rather than picking the last defined constructor, for example. I also like it when you have option to disambiguate based on the type name (like in Lean).
I like it, because I hate it in Haskell and Rocq when you have to come up with random, adhoc prefixes to disambiguate stuff.
What I gather is that most people will answer unreliably if you ask them directly if they've understood, either to not disappoint you or to save embarrassment, so it's a very unreliable measure.
I learned from a teacher that it's sometimes better to ask questions that tests their knowledge, to ensure the ideas have in properly. Then you can adjust your approach based on their response. (alas, easier said than done)
a spread of 16x16, 2 color characters from the post-jam version of my 7drl project. various oracles and mystical beings, some people disguised as devils, some skeletons, a grave digger.
if you appreciate my art, design, procedural, or experimental work, or design analysis, please consider supporting me or reposting! i'm /extremely/ good at living frugally but i would like to eat better and get a new computer. i'm aiming to get to $1000 a month.
www.patreon.com/vacuumflowers
Yeah, a lot of people who care about the environment are now pushing for urbanism, walkable cities, public transport, electrification etc. This is not just good for the environment, it’s good for public health, housing affordability, the economy, energy security, and people’s wellbeing in general.
I didn't know you had a patreon! I had considered asking you about it in the past, because I find your posts really inspiring. If I was going to donate, it would mainly be because I'd want to support your creative endeavors and curiosity, not necessarily to get a specific monthly thing out of it.
added a page with scans of the original sketches
Never get tired of seeing people’s sketches and creative process like this. So good! Very inspiring!
As a reader I like to think of them at their best as signposts… or treasure maps… or a friendly guide pointing you to somewhere where you can learn more, or to find alternate perspectives. Not sure how to illustrate that but yeah.
Those icons are so lovely
FWIW, the jury is still out on Gen AI being good at text summaries. Because it shortens & emphasizes text based on distribution, not meaning, it’s less of a summary (ie, salient points highlighted) than a compression (ie, the most frequently asserted terms in the distribution). Not the same thing!
This was in my recommendations but I am worried it will contain lots of ahistorical ideas about the development of Rust.
Lynda Barry (2016)
Not sure if it helps at all (its tangential), but there’s some questions over how a similar word choice originally entered into programming languages from natural language: github.com/e-n-f/if-the...
I wrote a piece on LLMs, citation metrics, and the institution of science. It's been up for a couple of days, but it is officially "out" now.
Very cool to see these progress updates! And nice to see Runebender getting some love!
You need to then do another step with some other provider to verify it I think.
Welp, I got the age verification thing. Not sure if I'll end up sharing my data with this third-party provider. Feel free to follow me on the fediverse via the link on my profile.
It’s so frustrating to see scientists using image models for this kind of thing, the results are really bad compared to what actual artists could do. So thank you for your service!
Bringing artists and technical people together is really important work, even if you can’t be all things to all people. This brings me so much hope.
It’s widely known (and, I think, pretty uncontroversial) that learning requires effort — specifically, if you don’t have to work at getting the knowledge, it won’t stick.
Even if an LLM could be trusted to give you correct information 100% of the time, it would be an inferior method of learning it.
at any rate it's very pretty and thanks for sharing! commiserations about the wet feet and other inconveniences 😅
context: I still find cold rain weird
me as an australian
oooh, very exotic