Dragon1-1

https://store.steampowered.com/app/1222730/STAR_WARS_Squadrons/ Amazing visuals, VR, HOTAS support, but flight physics from Star Wars. Pretty great characters and a cool singleplayer campaign, too. Sadly, it's made by EA, and it's out of active dev. We definitely need more of those on the market. In fact, if ED were to use DCS engine for a licensed BSG sim, using physics similar to FC3 and even simpler controls, I think it'd have some appeal.

Just a quick sanity check, are the J-7s in the campaign supposed to be equipped with flare pods? The dispensers we have are specific to MiG-21bis, and I can't find good info on whether the J-7s would have had anything similar in the 90s. As far as I can tell, on any MiG-21 other than the one we have, chaff/flares were only ever carried in a pod on the centerline pylon, and this would not have been carried on most types of missions.

The thing is, you need to get the taxi directors to acknowledge your salute (otherwise the crew won't hook you up to the cat), and the command doesn't always do that. I got it to work, it seems the mission doesn't break, it's just a little annoying. Now, if only those MiGs didn't gang up on me every damn time...

Except on those which only have one radio, of which we have a few. Then it'll work normally.

In the West it's usually called "Velocity search" or some such, and Pete Bonanni had some choice words to say about it ("a joke played on the Viper pilots by Westinghouse engineers"), so I guess this is a fairly universal sentiment.

Honestly, I think a more apt description of that behavior is "thinks he knows better than anyone else just because he builds rigs for money, and is being a jerk about it". It has nothing to do with the real woke movement, and the rightwing hijacking of the term essentially boils down "a thing that we don't like", which is not terribly useful (nor is the leftwing hijacking the term to include gender/class/sex equality struggle, the real deal is about black people always getting the shaft in the US). Myself, I just want to know how much the 5090 gains in VR, and without the BS frames, please. I've seen some reviews, and it does seem to outperform my 3090 (it had better!), but I haven't seen anything conclusive on just how much of a gain there is.

No, you are insisting my assumptions can't be right, only because a few other products had tried and failed, ignoring those that tried and succeeded. That's exactly how you end up not making anything new or innovative. I never said my approach is guaranteed to work, and your "counterevidence" doesn't counter my evidence that it sometimes does. You seem to be trying to prove that a product that doesn't match or exceed the capabilities of what's already on the market can't possibly make a profit. How about that: a computer with a low res screen, wimpy CPU, a cheap knockoff OS, and the biggest innovation is its modular architecture, which doesn't really seem bring much to the table. There's no way that could sell, right? And yet, you're probably not typing this on an Amiga. My point being, a new technology that does not compare well to the cutting edge can still succeed, if the people making it have the right business idea. You're obviously unlikely to ever have such an idea, but that doesn't mean nobody else can.

It seems that after saluting by command, Paco started to taxi, I needed to put in some combination of radio menu salutes and requests for launch. The taxi directors then sent me on a somewhat convoluted path to the cat.

We need an option that gives us full sized diffuse maps and halves every other map. Text and small details are on the diffuse map, resolution of normals, speculars and glows is not critical.

No, but a single company outcompeting the others to such an extent is bad for both the market and the customers, in the long run. Especially if the product in question forms the basis of much of the national infrastructure. This is a very notable instance of free market not delivering a good result for everyone, but resulting in one company attaining a near-monopoly position. It may be "efficient", but it's an extremely brittle state, and the nature of the semiconductor industry means it's hard for a smaller competitor to get off the ground. Other countries have expertise, and efforts are being made to diversify and strengthen national chip manufacturing, due to how strategic it is (some leaned their lessons after the pandemic chip shortage), but they're slow and we've still got a long way to go. Right now, neither Nvidia nor TSMC have viable competitors, which led us to the situation we're in now.

Two problems here, which make it a bad example: 1). Quest is still a big purchase, too big for the average person. 2). Horizon Worlds sucks and nobody actually wants to use it. There's plenty of money in making the apps, as well as running a store, as Sony demonstrates with their consoles and games. The difference is, the games Sony makes are fun. Zuck's misbegotten metaverse is far from that. This sales tactic is not a guarantee of success, it's a tactic that has worked in the past. Do it with a cheaper headset (even technologically inferior to the cutting edge Quests), and make an app that people would actually want to use (this is the key part), you might get somewhere.

Marketing. It wasn't exactly green back then, either, but they wanted to attract settlers. There was a warm period at the time, but it wasn't quite that warm.

I didn't mean it's currently possible, just that it's one potential way to do it.

No, but you can make a mass market VR app and sell that, which is the scenario I mentioned before. Just like MS and Sony making a loss on selling consoles to make (a lot of) money on games for them. I probably should have used that as an analogy instead, the printers just came to mind first. And yes, it's abusive of both the customers and the market at large (because the competition either follows suit or is priced out). When did that ever stop anyone? No, the one where you don't outright ignore what I said to make a cheap shot at a subtly different statement. There's a clear difference between the razor and blades sales model and economics of scale, and I believe I made the distinction clear enough. Easy mass production enables both, but only one involves actually losing some money. Only because there's no easy way to restrict a car to driving on oil and gas from your company. As a matter of fact, Rockerfeller did hand out kerosene lamps to the Chinese for exactly this reason, but that was when his company was the only game in the country. Funnily enough, it turns out nuclear power of all things does work that way in commercial settings. They'll sell you the reactor at cost, and then bilk you for the fuel.

Neoliberal (and it is politically conservative, Reagan and Thatcher were its champions). If you can't even get that right, then you should really stop pretending you know anything about "basic economic reality" until you can at least recognize the basic terms. So was mine, and you failed to see that, too. Looks like I have to spell it out. What if there isn't anything more to accomplish? What if we're about to hit the plateau? These are the questions the investors are asking. We're still being promised the next great leap, but here comes a Chinese guy and says "how about efficiency instead?" That's got to get some people thinking, what if we make the leap and faceplant into a wall? What if progress, from here on, will mean doing slightly more cool things with fewer chips every time? Deep Seek runs on NVIDIA chips, but the investors aren't thinking about Deep Seek. They're thinking about is successor. Will you put money on it running on NVIDIA chips? Will you put money on it requiring more of them? A large amount of people just didn't. The fact that it's Chinese, not American, means that it will likely to be looking at using Chinese chips in the future, the international situation being what it is. ...you do realize you're actually mangling a quote from a socialist? Your version is just as wrong as the original one, BTW. Churchill was good at witty quotes, but I'm not sure if you're aware of what happened once he had to campaign on his policies. Both times.

Because it might be that one would need the one that loses money to use the one that makes money? Say, sell printers at a loss and then make money on selling ink for them (real example, believe it or not). If your margin on ink is high enough, and people use it up fast enough, you'll make up for the loss on printers, but nobody will buy an ink cartridge if they don't have a printer to put it into, no? That's why you need a strategy to make them want it. You also seem to willingly ignore that I'm not talking about selling at a loss here, I'm talking selling at a small margin. Stop arguing a strawman and stick to the point. A product that is usable, but merely not as good as the competition, can still sell if you can undercut the competition on price by a big enough margin. Mass production is what enables you to sell stuff at a margin of a few cents per unit and still come out ahead. Yeah, because chips were a was completely different tech from tubes. Small designs are cheaper to make, up to a point, that point being where you need to go into heavy duty nanotechnology in order to achieve densities that you want. The smaller design is, the more precise your manufacturing process has to be, and precision costs money. A chunk of silicon is cheap, a machine to make a chip out of it is not, and the denser the chip, the more expensive the machine becomes. Again, material costs in electronics are typically negligible unless you're doing something exotic.

You should do training missions before say things like that, because you're quite wrong. The ones that are limited don't take months (and don't break on every update, either), and the ones that do take months are very clever and flexible. The only thing AI could do better there is, perhaps, recording the voice lines. DCS scripting sucks, that's another problem. The real reason that mission development can take months is that tools we get are difficult to use. No AI will ever be able to bypass inherent brittleness of DCS scripts, simply because it can't be bypassed without a scripting overhaul on the engine side. ED seems to be trying to make ME less painful to use, but it's a slow process. AI is not magic, nor is it a person. In the end, it's just a bunch of equations and a lot of brute force. As a tool, this can be useful, but it won't do the thinking for you. Those "things only human can spot" only seem hard to guess, but that doesn't necessarily mean they are. You can use it to rephrase existing information in a way that suits you better, but be careful not to assume too much about what it can do.

This is exactly what would happen if someone were to invent a new fuel, or a better engine. Moreover, this example is particularly bad, because rockets are designed to run on a specific type of fuel. Sure, people would be going to Mars in fewer launches, but not on SpaceX rockets. That is why they'd take a stock hit, and also why other AI companies are taking the hit now. It's not their AIs that are in demand. Likewise on the semiconductor market, if it can work with fewer, lower end NVIDIA chips, who's to guarantee the next version won't run on, say, a few 2070 from the bargain bin? Stock market bets on promises and expectations, and right now, it took a hit because there's a very reasonable fear that the AI market will not continue its current explosive growth trajectory, but will instead plateau, with companies moving towards efficiency instead. Except the consumers want to go to Mars, not to a giant ball of gas shrouded in radiation. Maybe, for instance, parts for Mars bases do not get cheaper or faster to produce just because rockets did, and bases designed on Mars will not work on Jupiter. In that case, even if SpaceX reengineers their rockets to work on the new fuel (itself a big investment that will eat into profits), the result will be that fewer rockets will fly through no fault of their own. Understand that, and you'll understand why companies sometimes choose a FUD campaign instead of getting on with the times. Sometimes, it's enough to outright kill off an innovation. Another example: if you build a better mousetrap in a mice-infested country, you'll certainly make good money. But what if it's so good that it outright extirpates the mice? Would you be able to sell an even better trap in this situation? I don't think so. In some markets, it's very much possible to go out of business because your product was too good. For further reading on why your theory is bunk, look up Phoebus cartel. Yes, which is why most of of what you wrote is neoliberal balderdash. Your thinking is exactly what made the markets, and the world as it is now, such a mess. There is no such thing as infinite demand, the only reason NVIDIA can say things that it does is that they're a near-monopoly when it comes to AI chips, and thus are free to manipulate the supply. It's not that demand is infinite, it's that the people making the stuff made damn sure to never make enough chips for everyone, even though they could. Markets are a complex thing. They're never completely free, they depend on finite resources, finite logistics, finite market size, and are often interconnected (for instance, a shortage of latest socket CPUs will bog down sales of mobos with that socket, no matter how good they are). Real markets, as opposed to simplified models used to bamboozle the general public, are never as well behaved as you'd like to think. I'd suggest you educate yourself on this before you pretend to know anything about economics (and I really wish politicians heeded that advice...).

Note that the only thing you need an AI in this scenario is natural language processing, in other words, talking to you like a human. The rest is, in essence, a glorified flight manual and basic flying handbook reader, cued up by cockpit actions. Effectively the same thing DCS training missions already do, autogenerated in real time. I admit, Deep Seek makes a pretty good job at the talking part. Seeing as it's lighter on the hardware than other models, it could be a good starting point.

You can set them so that the key does not command full deflection of the stick. However, a realistic sim will be unplayable without analog controls almost by definition.

That was my point. My dad was very much in the race for a long time, in fact, he'd been at it since he swapped his second Atari for a PC. As a kid, hand-me-downs from him were sufficiently powerful that I seldom complained about not being able to run something, except when things like SSE2 appeared, at the time I had the last high end AMD CPU that didn't support it, which kind of sucked when it became a hard requirement. At one point, he just decided further upgrades weren't worth the money, the 20xx series RTX being a bit of a dud had a lot to do with it, and the 30 series launch pricing sealed the deal. The "target consumers" are getting tired of ballooning prices for incremental performance gains. While I haven't yet declared the 3090 will be the last GPU I buy, unless something truly revolutionary is added (more than spamming fake frames, anyway), I expect it to suffice for the foreseeable future. When I do upgrade my rig, it'll likely be because something like retina-level VR came out.

My dad actually did. It's been a while since he got himself a 1080-based rig. Nearly perfectly silent when idle or close to idle, can run almost everything but the very latest stuff on his modestly sized screen, and unlike my own rig, it doesn't double as a space heater. He barely uses its full capabilities, anyway. In fact, I suspect that quite a few people dropped out of the performance race, which is causing a lot of grief when they try playing the latest games on a rig that was cool when it was built (like my dad's), but the technological progress passed them by. In fact, I'd expect the majority of people to not be too eager to constantly update their PCs. I think it's way past time for that. UE5 games, in particular, seem to be suffering from an epidemic of bad optimization and poor coding that can be powered through, but people shouldn't have to buy a top tier GPU just to play the latest releases. Even lowering the graphics settings often doesn't help.

Whatever the beancounters think makes them the most money. Idle them, store the output in a warehouse, or retask them. I don't know what they'll do, but lowering prices is unlikely to be part of the plan.

What NVIDIA is saying can also be said in other words: "pay and cry, we know full well we haven't squeezed every possible dollar out of you yet, and we'll just keep throttling supply if you try any of that do more with less nonsense". With Trump tariffs, expect prices to only get worse. TMSC will likely not come to the US, since that would mean paying US taxes and US wages. It could be cheaper for the company to leave its customers paying tariffs, and even if they do build a US factory, chips from that will be expensive. Samsung does make chips, too, by the way, in addition to Chinese manufacturers who are slowly catching up, so it's not that all our eggs are in one basket, but either way, the market is not very competitive, and any hiccup from TMSC will result in others jacking up their prices. This "business" is all about the reason why the pricing on the 50 series is as insane as it is, specifically how NVIDIA is blatantly abusing the mechanism. The other discussions are not that far off topic, either, since they all circle back to two fundamental questions about the 50 series: "why can't I afford the one I want?" and "do I need that overpriced crap at all?".

Here's a thought: someone should make a modular GPU board. Instead of selling everything in one pack, let us buy a board, the GPU chip itself, however much VRAM we care to insert, and a cooler for all that. It works that way with the rest of the PC, why not the GPU? I don't think inserting a CPU-style socket would affect performance too badly.