100% this!

…you’re gonna have a really hard time keeping that cool. Most radiators in space try really hard to avoid collecting energy. I’m not saying it isn’t possible to cool something in a vacuum, just that it’s not trivial!

True, objects will radiate heat without the need for conduction or convection. The trouble is that heat dissipation through what is essentially black body radiation is massively inefficient compared to conduction through another material. If you have a body continually producing heat…

The problem is that this heat literally has nowhere to radiate to though. Heat transfer in vacuum is a really hard problem to solve. A satellite generates very little heat, so you can vent it as gas/liquid as it condenses. A datacentre on the other hand…

I mean, yeah, he’s crazy. But the thing I really don’t understand is the “datacenters on the moon” bit. I mean, the whole problem with these is cooling… how do you do that in zero atmosphere??

Haha, ain’t no amount of money that would be worth me selling 19 years worth of games! 😆

“I want this transparent thing to appear in front of this transparent thing” is my personal favourite…

Haha, I’m British so also don’t know! We just call it coriander here…

So, what, Americans call the leaves cilantro and the seeds of the same plant coriander?? That’s wild! 😄

Out of interest, what do you call the seeds? Are they cilantro seeds too? Or do they get to be coriander?

I use Nobara for both gaming and dev. Definitely my favourite distro so far!

I’ve used Magicavoxel in the past and found it to be great. There are .vox importers for Unity (and Godot I think) that let you pull out the raw indices as an array. The .vox format is also pretty open: here’s a vox reader for c# that should do what you need: github.com/sandrofigo/V...

The best kind of evening!

This is so cool!

Copilot to help relieve RSI caused by typing the code you’re writing. I personally run a distilled llm on my local machine and compared to standard autocomplete it’s incomparable. I definitely don’t have an llm “code for me” and the vast majority of devs that use things like copilot don’t either!

No. I’m saying that copilot is used by a large number of developers in the industry, as a tool that essentially works to autocomplete *your* code as you write it. I’m also saying that there is a big difference between having an llm generate code for you (a la Cursor) and using…

Copilot or other fancy autocomplete is pretty much standard practise now. Having copilot suggest the third line in a series of boilerplate code lines is vastly different from vibe coding with cursor for example. But hey, each to their own.

<chef’s kiss emoji>