Eh, it’s just a start of development. It only goes 300 meters. Blue Origin goes higher, but even they aren’t in orbit.
Japan also has some odd limitations on their rockets as part of their self defense only constitution. They don’t build a rocket that could potentially be used to strike mainland Asia.
Big Tech has done everything they can to convince people that they absolutely cannot be trusted with this.
Just so we’re all aware, GPS jammers will invite the interest of the FCC. Now, protests aren’t about being well-behaved, but just know that there’s an entirely different federal agency being brought in.
Pixel 6 is the minimum that can load up GrapheneOS. Those are like $130 on eBay.
Pixel 8/9 does have some CPU features that help separate memory, which can be useful for keeping apps from creeping on memory they shouldn’t.
This is the sort of nuts and bolts of protesting that Americans are learning the hard and fast way.
Since you’re posting this all over the thread, I’ll also have to repeat the information that gyro/accelerometers are not capable of doing that. Small measurement errors stack up and throw it completely off.
Gyro/accelerometer data isn’t accurate enough to do that. Small errors in the data add up and will quickly drift away from the actual location. You can use it for video game controllers, but not tracking over large distances. Edit: there’s a reason the best VR tracking often uses external methods, not controllers alone.
But most phones have GPS and that’s where the real problem is.
For raw computation, yes. Most programs aren’t raw computation. They run in and out of memory a lot, or are tapping their feet while waiting 2ms for the SSD to get back to them. When we do have raw computation, it tends to be passed off to a C library, anyway, or else something that runs on a GPU.
We’re not going to significantly reduce datacenter energy use just by rewriting everything in C.
Every one of these only makes me say “wouldn’t it be great if we did everything with RPN”?
Have you ever chained three Cisco 2600 routers together and then successfully ping’d clients on each end? Do you know what BGP is? OSPF? Do you know the difference between routing and routed protocols?
I know you don’t, because people who do don’t make the claims you’re making.
No they fucking don’t, that’s not what routers do. You don’t know what you’re talking about.
And don’t fucking tell me NAT is for security, either.
Skype won’t be supporting anything at all very soon.
What happened with Vonage is something that could happen with any kind of instant messaging, including things like Discord.
With everything directly addressable (not just static addresses, but directly addressable), an IM/VoIP service can simply connect to the recipient. No servers are necessary in between, only routers. That doesn’t work with NAT (CG or otherwise), so what you have to do is create a server that everyone connects into, and then that forwards messages to the endpoint. This is:
This is largely invisible to end users until free services get enshittified or something goes wrong.
Yes, it’s only tangentially related to static addresses, but it’s all part of the package. This is not the Internet we should have had.
And at least in the US (in single family homes) its crazy unlikely that your router is behind any NAT
Your router has NAT. That’s the problem. CGNAT is another problem. My C&C: Generals issues did not have CGNAT.
. . . nobody at home actually runs VOIP . . .
Plenty of people used Skype and Vonage. Both were subverted because they have to assume NAT is there.
. . . quick game servers don’t need static . . .
But they do work better without NAT. That’s somewhat separate from static addresses.
My old roommate and I had tons of problems back in the day when we tried to host an Internet game of C&C: Generals behind the same NAT. I couldn’t connect to him. He couldn’t connect to me. We could connect to each other but nobody outside could. It’s a real problem that’s only been “solved” because a lot of games have moved to publisher-hosted servers. Which has its own issues with longevity.
You can get IPv6 addresses. What you can’t get, in many cases, is a static IPv6 prefix assignment.
CGNAT is not fine. Its problems are simply hidden from most people. ISPs have to have more equipment that’s less reliable, increases latency, and is potentially a bandwidth bottleneck.
The reason they have no use for a static address is because applications haven’t evolved to work that way. Roll back the clock 30 years, do IPv6 seriously so that everyone has static assignments by the time the Y2k problem has come and gone, and you have a very different Internet.
In fact, many applications, like VoIP and game hosting, have to go through all sorts of hoops to work around NAT.
“Hashed emails”. Besides the fact that they can match up a hash from one source to a hash from another source to link them to the same person (they never said they’d salt them), emails often have enough predictability to break the hash. Assuming they all end in “@gmail.com”, “@outlook.com”, or “@yahoo.com” will get you the vast majority of emails out there. Unlike a good password scheme, people don’t shove a lot of random data into their email addresses.
The US is allergic to it, but needs to get over it.
Aluminum wire was tried in the 1970s due to a spike in copper prices. The problem was that they just tried to swap it right in. Aluminum and copper have different rates of expansion. Over time, that would slowly loosen the connectors, and the wires would pop right out and cause a fire.
You can design connectors to handle both, and you’ll see many electrical things today specify that they’re good for aluminum or copper wire. It still has a bad reputation among electricians; they haven’t unlearned the problem yet.
Now, one place it’s more of a problem is in things like transformer windings. There are kilometers of wiring in any of them, so the higher resistance of aluminum is a problem.
From old electrical connections that weren’t designed for the different rates of expansion of aluminum and copper. Today, most of them are.
There’s a thread of thought that pops up in pro-AI posters from time to time: technology can’t go backwards. The implication being that the current state of AI can only improve, and is here to stay.
This is wrong. Companies are spending multitudes of piles of cash to make AI work, and they could easily take their ball and go home. Extending copyright over the training data would likely trigger that, by the industry’s own admission.
No, self-hosted models are not going to change this. A bunch of people running around with their own little agents aren’t going to sustain a mass market phenomenon. You’re not going to have integration in Windows or VisualStudio or the top of Google search results. You’re not going to have people posting many pics on Facebook of Godzilla doing silly things.
The tech can go backwards, and we’re likely to see it.
The Estes Corporation makes rockets that will do 600 meters.
It’s great that Honda is doing this. We really need other companies in this area, because SpaceX is dominating it. Even if Elon weren’t a walking disaster, we don’t want one company so badly outclassing everyone else.