yes and also no ;) Think of stirling engines as driven by "heat differentials" as opposed to "heat." At first when you fire up the GPU, it's hotter than everythign else and can start driving a stirling...but eventually everything heats up to the same temp and it stops moving. To get it moving again, requires a heat differential, so part of it would need to radiate, which is blatantly difficult in a vacuum(1), such as space.
In case it's not clear: Little-St. James Wannabe Invitee, Nazi-Saluter, Musk's full of it again, but to recognize it requires being halfway through college physics to understand it, so all the elites will be glazed over thinking they're onto the next big thing. grift, grift, grift, grift.
Do the same for X! Well.. a layered addition maybe. I've always felt it's bringing swags of stuff which never gets used. A non accelerated fb or vesa binding would do for a lot of things.
I liked this piece a lot. Nice write up of how you explored the space.
> What I really loved is that XLibre X11 packages DOES NOT CONFLICT with Xorg packages. You just install xlibre instead of xorg and everything works … even better then with Xorg
How do you control which one is used? I was expecting xlibreinit or something, but the rest of the post appears to just run xinit like normal with nothing that I noticed that would select an X implementation
Reverse address lookup servers routinely see escaped attempts to resolve ULA and rfc1918. If you can tie the resolver to other valid data, you know inside state.
Public services see one way (no TCP return flow possible) from almost any source IP. If you can tie that from other corroborated data, the same: you see packets from "inside" all the time.
Darknet collection during final /8 run-down captured audio in UDP.
I don't understand how foveated tracking won't cause a sense that peripheral vision is fuzzy. Or how it will track saccades, and so avoid fringe effects.
But, the "I don't understand" is strong in this. it doesn't mean "it can't work" but I don't understand how it avoids the problems.
Maybe the size of the computed foveal coverage area is made big enough, to cover the movement? But if you move your eyes suddenly, there's got to be some lag while it computes the missing pixels. So you'd see the same as when Netflix ups the coding rate: crude render becomes clearer. Banded would become smooth transitions.
Imagine watching Netflix out of the corner of your eye. You wouldn't notice those transitions at all. Your eyes and brain are mind bogglingly good at making stuff up.
Do you know you have a big hole in your vision in each eye where the optic nerve is? It's about half the size of your fist at arm's length, and 35 degrees to the side. Your fovea happens to be roughly the same size. It's the HD part of your retina, and it's where essentially all of your vision happens. It's the only section of the retina that sees color, for instance. The periphery sees motion and that's about it.
Saccades top out at around 700 degrees per second. At 120 frames per second that's only about 6 degrees in either direction. Compared to the FOV, that's tiny. Overfill it!
I had to edit the code and change scale from 90 to 300 to see it easily in full screen landscape on my phone. I have presbyopia and often need reading glasses to see small things.
Well that's both unsettling and informative. It would explain why I seem to be a better gamer on a good sized monitor unlike the oversized one I have now. (I am ignoring the aging factor here, deliberately. Denial isn't only in Egypt.
Speaking as someone who has been using HMDs since '05, and periodically checks in on dynamic foveated rendering every year or so, the frustrating thing is that sensitivity to it varies significantly.
Some folks experience the image pretty much continuously and don't notice the edge blurring. Others see it every time they move their eyes left/right. This is on the same headset.
Part of it is driven by differences in eye geometry, and even color (as this impacts the effectiveness of the camera track of the eyes). I've seen the raw camera buffers for eye track on a couple headsets and they're.. a mess.
Honestly that the feature works at all, for anyone, is still mind boggling to me.
Sufficient additional coverage + predicting the trajectory of your eyeballs. As far as I know, all of the journalists invited to try it were unable to see the low-res periphery, despite actively trying to break it with fast eye movements.
I don’t have an answer for you, but take some applause from me for spelling this out :)
It’s very difficult for most people to intuitively understand that what they could not figure out after five minutes of thinking might not necessarily be impossible.
You're thinking about foviated rendering. They're just doing foviated streaming. So it renders at full resolution, and only streams the parts that you're looking at with full resolution on the stream.
Your eye is just another input source, if you don't feel the controller lag from streaming games otherwise, you're probably not going to feel it here either. It's not like an additional round trip or anything, your eye is here and the joystick is here can be sent at the same time, and you get back the rendered frame in return.
As for peripheral vision, any gradation being smooth probably helps, but there might be more tricks to make it look normal. I'm reminded of how jpeg images and some sound codecs only store information that we can actually perceive.
whats the impact of epigenetics on this, given we're looking at a cohort of Boomers whose parents in many cases underwent extreme dietary restriction across the years of puberty or close?
the post ww2 children are the ones I'm talking about: their parents have in large part had cataclysm events in their fertile windows. my parents were 192x babies and their parents in turn were 1890/1900 window, and so dodged a lot of things because of a peace bonus. But my parents began a family in the 1950s after stress, and since neither fought nor were in the ETO or Asia, I suspect impact on me is minor but for dutch, or german, or french, or polish or chinese ...
I very much hope this doesn't descend into licence wars but I would think all of the BSD, MIT, ISC, hold-harmless, RAND and GNU licences qualified. If that's true and it was understood the public/commons got an outcome, I'd be in favour.
If the code is under restrictive clauses, or gets tokenistic input and the quotient of time and money is spent doing something else, then I think this is a licence to cheapen out contracting rates for-profit.
they are loosers.
they spent decades and billions and billions
and billions of dollars , sort of rebuilding used shuttle parts(SLS= shuttle leftover systems), and cant fill the fucking tanks, cant get things to seal, there thruster valves(old shuttle stuff) sieze up, and they are a bunch of bumbling grifter loosers.(sorry! to the many good people caught up in this debaucle)
A lot of productive thinking happens when asleep, in the shower, in flow walking or cycling or rowing.
It's hard to rationalise this as billable time, but they pay for outcome even if they act like they pay for 9-5 and so if I'm thinking why I like a particular abstraction, or see analogies to another problem, or begin to construct dialogues with mysel(ves|f) about this, and it happens I'm scrubbing my back (or worse) I kind of "go with the flow" so to speak.
Definitely thinking about the problem can be a lot better than actually having to produce it.
There's a reason the Martin B57 which NASA crash landed recently was not made by English-Electric despite being a Canberra Bomber, and the Harrier Jumpjet is known in america as the McDonnell Douglas AV-8B. National-Strategic concerns dictate production happen on-shore, and has done for a very long time.
Comac will make better and better aircraft. It won't stop Airbus or Boeing being viable companies any more than it will stop Bombardier, or Embraer. All airlines make political compromises. Wings are assembled inside one economy, so the aircraft can be sold inside that economy. Or, engines are maintained inside that economy. Or, a new JV is spun up which blends both economies interests somehow.
If you look under the covers, Bombardier has had many acquisitions, sell-offs, "manufactured under licence" in it's time. Not bad for a snomobile company!
Aircraft are not bought solely on price, despite what people might say about fleet bidding. operating costs might be a stronger reason than initial fleet acquisition cost. Or cost of compliance. Or pilot re-training. or, in the post MAX crash era, reputation.
(not in the industry, don't drive an aircraft for a living)
If there was an acceptable efficiency to concentrate the heat into mass, you wind up at ion engines and shed heat physically for thrust.
I think he hasn't done his sums, or maybe we need a "show your workings" moment over the Fermi numbers on this.
I'm sceptical.
reply