Tristam
@trist.am
Recovering from BigTech. Procgen and 3D rendering enthusiast. Restauranteur
This is all the config I use for that sort of thing: github.com/swiftcoder/f... and then I leave the wasm-pack side of things up to trunkrs.dev
November 15, 2025 at 3:41 PM
This is all the config I use for that sort of thing: github.com/swiftcoder/f... and then I leave the wasm-pack side of things up to trunkrs.dev
It works well enough, but you do have to pick some options carefully (rand needs some custom config for WASM, logging and panic handling both need customisation, etc)
November 15, 2025 at 3:14 PM
It works well enough, but you do have to pick some options carefully (rand needs some custom config for WASM, logging and panic handling both need customisation, etc)
This is maybe my biggest pet peeve of the way modern UI toolkits are going. Why aren't you ready to handle input by the time the UI draws to the screen? What else are you doing? Layout is finished!
November 14, 2025 at 10:40 AM
This is maybe my biggest pet peeve of the way modern UI toolkits are going. Why aren't you ready to handle input by the time the UI draws to the screen? What else are you doing? Layout is finished!
To Tom’s point, the patterns that make GC work well are about exercising strict control over who owns what data. And those are the same patterns that make non-GC languages work
November 13, 2025 at 9:33 PM
To Tom’s point, the patterns that make GC work well are about exercising strict control over who owns what data. And those are the same patterns that make non-GC languages work
But, most of the time the problem is in the letting go. Stored a reference in a cache? No GC till eviction. Another thread stored a reference? Passed a reference across the JNI boundary to native code? Good luck.
November 13, 2025 at 9:33 PM
But, most of the time the problem is in the letting go. Stored a reference in a cache? No GC till eviction. Another thread stored a reference? Passed a reference across the JNI boundary to native code? Good luck.
Can’t really do mark and sweep of a 100GB heap, if you want the app to stay responsive. End up needing a generational GC with a bunch of strategies at different scales
November 13, 2025 at 9:33 PM
Can’t really do mark and sweep of a 100GB heap, if you want the app to stay responsive. End up needing a generational GC with a bunch of strategies at different scales
The JVM’s many GCs were also designed to handle cycles. They just start to struggle when your app gets too complex (and on a big server, 100GB of heap and a few tens of thousand of threads is not uncommon). And of course, that all assumes one didn’t reference the cycle somewhere to start with…
November 13, 2025 at 9:13 PM
The JVM’s many GCs were also designed to handle cycles. They just start to struggle when your app gets too complex (and on a big server, 100GB of heap and a few tens of thousand of threads is not uncommon). And of course, that all assumes one didn’t reference the cycle somewhere to start with…
In theory, yes, in practice… ymmv. The JVM has a pretty good selection of enterprise-grade GC implementations, and yet somehow half my job was tracking down all the ways that memory had failed to be freed
November 13, 2025 at 8:46 PM
In theory, yes, in practice… ymmv. The JVM has a pretty good selection of enterprise-grade GC implementations, and yet somehow half my job was tracking down all the ways that memory had failed to be freed
Reference cycles are really hard to break, any stray reference can keep a big reference cycle alive indefinitely. When I worked on Java cloud services, wasn’t uncommon to have to track down a stray gigabyte or do the GC was failing to detect could be reclaimed 🤷♀️
November 13, 2025 at 8:36 PM
Reference cycles are really hard to break, any stray reference can keep a big reference cycle alive indefinitely. When I worked on Java cloud services, wasn’t uncommon to have to track down a stray gigabyte or do the GC was failing to detect could be reclaimed 🤷♀️
Worth checking if your device is in the OpenRGB support list: openrgb.org/devices_0.9....
OpenRGB
Open source RGB lighting control that doesn't depend on manufacturer software
openrgb.org
November 13, 2025 at 11:08 AM
Worth checking if your device is in the OpenRGB support list: openrgb.org/devices_0.9....
Specs-wise it’s a bunch of very minor improvements over the Quest 3, but the linux-based SteamOS is a *lot* more appealing than Meta’s very locked down Android fork…
November 12, 2025 at 8:27 PM
Specs-wise it’s a bunch of very minor improvements over the Quest 3, but the linux-based SteamOS is a *lot* more appealing than Meta’s very locked down Android fork…
MDN, thankfully, is correct
developer.mozilla.org/en-US/docs/W...
developer.mozilla.org/en-US/docs/W...
Indexed collections - JavaScript | MDN
This chapter introduces collections of data which are ordered by an index value. This includes arrays and array-like constructs such as Array objects and TypedArray objects.
developer.mozilla.org
November 11, 2025 at 6:49 PM
MDN, thankfully, is correct
developer.mozilla.org/en-US/docs/W...
developer.mozilla.org/en-US/docs/W...
Yeah, I just poked several interpreters with a stick, and AFAICT, every reference accessible by google is subtly wrong™
for...of calls Array.prototype.[Symbol.iterator] under the hood, and that iterator definitely does not skip missing elements.
for...of calls Array.prototype.[Symbol.iterator] under the hood, and that iterator definitely does not skip missing elements.
November 11, 2025 at 6:47 PM
Yeah, I just poked several interpreters with a stick, and AFAICT, every reference accessible by google is subtly wrong™
for...of calls Array.prototype.[Symbol.iterator] under the hood, and that iterator definitely does not skip missing elements.
for...of calls Array.prototype.[Symbol.iterator] under the hood, and that iterator definitely does not skip missing elements.
Does your array actually have a hole, or is item 29 set to undefined? (I’m not convinced the debugger displays these two cases differently)
November 11, 2025 at 6:35 PM
Does your array actually have a hole, or is item 29 set to undefined? (I’m not convinced the debugger displays these two cases differently)
Yeah, I tend to think it's a shame that blizzard's writers lean on "world ending calamity" to drive their plots so very often. WoW is just decades of world-ending-nonsense stacked on top of itself at this point
November 10, 2025 at 3:31 PM
Yeah, I tend to think it's a shame that blizzard's writers lean on "world ending calamity" to drive their plots so very often. WoW is just decades of world-ending-nonsense stacked on top of itself at this point