comex

There is this old project, but it’s stagnant: http://openhmd.net 

Oh, and I just learned that OSVR is just a wrapper around Oculus, Vive, and others’ proprietary SDKs rather than fully open source.

If I’m right, it’s not possible to override the IPD properly without a patch, which is a bit of a disappointment.

so I think I’m done figuring out the IPD issue with the new Oculus driver. maybe. hard to distinguish bad code from code I don’t understand

(Because the CV1 lets you physically adjust the horizontal lens distance. Which the DK2 doesn’t.)

So apparently the release Oculus software works with the DK2 but doesn’t allow adjusting IPD. Wonderful.

(Can simulate multiple returns by loading globals and such, of course, but that is much more expensive.)

I wanted to use it as a “slow path” for code JITted to JS. I guess a regular branch works, but functions can only return one value.

Does V8 still not even optimize functions with catch statements? Sheesh.

In other news, it’s annoying that asm.js doesn’t support catching exceptions. C++ can do it and zero cost exceptions have some use cases…

I looked at the previous post and it’s a good example of using naive libertarian reasoning to reach a nonsensical conclusion.

how did this colonialist crap make it onto HN http://www.overcomingbias.com/2016/04/school-is-to-submit.html …

.@crashplan is deleting backups older than 6 months with only a 5 day notice to customers. They even quote their EULA to justify it. :(

(…yet another period of minute or two spent sitting and staring at the Fitbit app because it syncs so goddamn slowly.)

Or something like that. After spending a shower thinking about it I’m still not sure what the algorithm looks like.

Thinking about how you’d make a game AI whose actions are subject to input lag. Have to replace states with prob distributions on states…