other_platypus [she/her, pup/pup's]

right, yeah I ended up figuring it out. This stuff has multiple names, weirdness is ridges and PV is ridges_folded

So apparently there’s not really a secret sauce to Minecraft worldgen, it’s just a lot of really big json files, a ton of splines, and there’s … Tolerable documentation on the wiki about it.

Or so I thought, until I got to the bottom of the files, and saw a function called old_blended_noise… Even the wiki doesn’t know how it works, so the trail is cold. Another algorithm gets away scot-free.

Edit: also I haven’t been able to trace weirdness or the resulting PV value it generates either.

beautiful powetry

it’s getting easier for me, slowly… I found reading a bit of Effective Rust (which is free online) helped solidify a lot of my knowledge, at least I think that was the one. Now I’m all deep into iterators and returning impl traits (what the hell is a future), but lifetimes, they do not scare me, for I know of .clone() (unless I’m trying to speed up an iterator and using a Cursor pattern and trying to store an internal reference to the object we’re iterating over and it hates me riviwkfhithenfoswi)

on the plus side, if you know both python and rust you can probably make pretty much anything

learning rust has been such a wild ride, at first it was “oh cool cargo is neat” and “oh I can make games with this” and then it became “wtf is a borrow checker or an elided lifetime” and now I’ve discovered profiling and benchmarking and wondering why the flamegraph doesn’t go deep enough to see why my functions that are being called 100 times in a single frame are taking 5ms each.

Its been years since I’ve worked with truly hot code, so I’m out of practice, and I’ve never done it with rust before, so it’s a lot of learning and scratching my head about the different tools. Like, it looks like tracy will only measure spans you declare yourself, cargo flamegraph will only go so deep for some reason, maybe there’s other ones that can tell me why this stuff is taking so long. And every time I use a different tool I need to recompile all 488 dependencies lol yay

There’s a Rust performance book which has been helpful, but it’s not intuitive for me to look at a complex function and say “oh yes this will be slow because of X”

glad to hear it went well, the news paints a really grim picture, and I do wonder what the variables are that increase the risk

I’ve been reading Devon Price’s “Laziness Does Not Exist” and it has been challenging and life changing. I am now refusing to feel guilty for taking breaks at work, even hours at a time, for feeling like if I’m not moving my mouse I’m not being productive, and for blaming myself when solutions don’t appear in front of my face when I want them to.

It helps that work has been unbelievably dysfunctional the past month or so (before that too, but there’s definitely been a recent rampup), since it pushed me to the point where something needed to change.

for the first time ever I’m actually using the usb-c port on my pc to plug into a usb-4 dock. I fully expected it to work since most of these dock devices can degrade down to usb-3 etc but I was still happy when my keyboard, mouse, audio, etc. worked

of course the display doesn’t work but I solved that with an HDMI cable, now at least I can switch to home shit by moving one cable from work laptop to PC, since my laptop is too slow to run my poorly optimized game code :(

yes its too hot

less than an hour in to work and already shit is very fucked

i think it took me a while to come to the same conclusion, last project i worked on (for learning) was a zelda: link to the past clone and i grabbed one of the alt-character sprite sheets from lttp randomizer (which are compatible with the base game) to work from from that, which saved a lot of time for learning and was at least a little fun for the slog that was the entirely random (to me) indexing of the sprite sheets and building framed animation.

some of the free assets i’ve found online are missing core things (like tilesets that can only make the example image and are missing corners/etc for anything else), or not aligned or sized correctly for actual gamedev, and since i’d never release with a free asset pack anyway i ended up just grabbing files from some other game a couple times for reference as its all placeholder anyway.

fun fact i learned last week: the grass_block texture in minecraft is uncoloured but there’s a wiki page on the colour values per-biome you can use (although i ripped from the jar file, apparently the textures are also on github, intended for modders, maybe the grass is pre-coloured there), if you are crazy like me and want to trace the entirety of minecraft to learn about deterministic noise-based generation and all of the performance challenges that come from generating, rendering, and lighting voxel worlds.

of course, my dream would be to do it all myself from scratch, or split the work with someone i trust, but maybe a year from now i’ll be releasing games with synty packs or w/e, if indeed i ever release anything at all.

People online talk about how trying to get seen by “the algorithm” is like trying to get a win out of a slot machine.

It’s easy to be negative about creating art of any kind, especially lately, but I feel more compelled to create than ever. Either we lose art as a medium to slop generators or we fight back and prove intentionality matters and that we can change people’s views and perspectives through art.

An image is much more than an 18 word prompt, heck I would argue it’s much more than even 1000 words. It’s a series of hundreds or thousands of small decisions made with intentionality and skill behind them.

Making a platformer with some tilesets you can get online is relatively easy, but making a player feel like they’re visiting a different world, tuning the controls so it “feels good” but also is unique, music and sound design that fits and builds a through line through the world, reusing themes, motifs, and instrumentation to connect it all together? All of that requires intentionality and cohesion that you can’t get out of a slop generator or an asset pack, and even if you could, you don’t get the source files or the understanding of the underlying systems that is required to tweak or to fill gaps.

I’ve been programming my whole life, I started by creating bots and chat apps with network protocols, and moved on to working with whatever free game engines I could find. I’ve never been able to finish anything worth releasing because I didn’t have that burning desire to keep going when, for instance, I had to write my own damn mouse picking algorithm (curse you, random dll files I found on the internet). I also never really believed I had anything worth saying or that my ideas had any merit whatsoever. I still don’t really believe that, but I am learning anyway. Maybe one day I’ll move from learning by copying to creating something that feels original and worth sharing.

can you share the book title? asking for a friend (who is me)

thank you, needed this today

happens a lot with Bevy since I think it uses macros to bend the type system for it’s crazy dependency injection magic

still waiting for the part of my transition where I can understand dyn or how to use lifetime parameters properly

edit: also my favourite is when the compiler is like “dw fam, the type that caused this error is too long for the tooltip so I wrote the full type name to a file. You’re welcome, good luck reading that”

ok but I only just saw this and it’s morning so

no I have to or else I could maybe be up all night

Am I crazy or is allergy season getting longer every year holy crap I want to call in sick and I’m on antihistamines