Current status: vImageConvert_420Yp8_CbCr8ToARGB8888

Not going to make any decisions at ~3AM after just thinking this up and rambling online about it, but curious to hear what others think. Do you use Linux+VR? If I port the next version of the game to Linux, would you try it? Devs, should I run away screaming from this idea?

The real risk is that the VR+Linux userbase stays small or shrinks. Free content or not, if there are no users, there's no added reach. Another risk is so much content gets ported that Soundboxing can't stand out. I'm not so worried about that, I'll keep making it better.

Longer term, if a significant number of users use the free Linux version, won't I go out of business? Prob not—in the future I plan to make revenue by helping users sell content to each other, so the added reach can only help add to the top of funnel for those users (and me 😁)

But importantly, we can grow the pond. If it's free and easy to install, maybe people will try it and realize the game plays the same on any operating system. There's likely a Metcalfe/power law correlation with each game ported, so the earlier you port, the greater your impact.

Obvious downsides are whatever work it takes to port it, list it, and maintain any code that needs to be written—with no boost to revenue short term.
There are upsides, however, to being a normal-sized fish in a tiny pond. More reach, more content & game activity; it's marketing.

Would it be a genius or idiot move to port Soundboxing to Linux natively and make it free—for Linux users only? I'd work to distribute it as widely as possible (Snap[craft], PPA, AppImage, the whole shebang.) Goal being to be so easy to install, you can easily test a system with it.

I keep thinking about something like AWS Lambda, but a game engine. Instead of registering functions to respond in HTTP to URL events, you'd register functions to respond in SceneGraph deltas to game events. With a WASM spec for that function signature, you could use ~any lang!

Of all the signups so far from the instance launching today, literally none of them followed anyone. Onboarding is a colossal, behemoth, showstopper level unsolved problem in the fediverse at the moment.

Also tomorrow morning I plan to "launch" tinyfed! For now this basically just amounts to me posting the following explainer video on my currently dormant social media profiles and seeing if anyone is interested. LMK if you have any feedback on the video or the concept for the instance

I am of the opinion that Swift extensions are generally a bad idea, just like open classes in Ruby. Composition should be the answer way more often than extension/inheritance. Fitts's law for code - functionality should be colocated, not scattered in different places.

I really hope Valve has a team working on inside-out tracking, or eyes on an acquisition

You can get so far with path-based blob storage and prefix matching. I'm hoping to be able to drop the (operationally expensive) relational db for this service I'm rewriting. So far so good!

Unfulfilled bucket list item: build a platform (in the Bill Gatesian sense.) It aligns my skills, politics, and personal ambitions. We live in a time of consolidated and truly awesome global platforms, so anything new seems quixotic. On the contrary, creating new value is our only option!

I really like the IETF Web Packaging spec, but they need to find a way to disassociate from AMP, because people are reflexively against AMP-adjacent things (and rightfully so, given Google's bad-faith reasons for pursuing AMP)

Show more

Generalistic and moderated instance. All opinions are welcome, but hate speeches are prohibited. Users who don't respect rules will be silenced or suspended, depending on the violation severity.