Jump to content

Dragon1-1

Members
  • Posts

    3885
  • Joined

  • Last visited

  • Days Won

    1

Everything posted by Dragon1-1

  1. Almost all that can be applied to the SLAM-ER, except the ability to RTB. It's not a weapon for purely preplanned targets, though you need to designate the target area. It does not put the pilot in harms' way (the whole point of ER part, and why Ukraine can't do much about Russian cruise missile bombers). Also, the older SLAM is very much not a rocket. It looks like one, but it has a jet engine. It uses a rocket booster for launch. SLAM-ER doesn't even do that, being air launched. Tomahawk, admittedly a much larger missile, can even loiter and search for targets on its own, acting almost like a typical drone. Also, aircraft very much can operate solo, this mission profile is called Battlefield Air Interdiction (BAI) and it is well known to the USAF. A TGP and an advanced radar go a long way. Again, you're stating that you can get a 1000$ drone, but not addressing the fact that such a drone is basically a toy modified for combat use. That works, fair enough, but is it going to stay that way? I don't think so. They are used in Ukraine right now, because they work. This doesn't mean that there will be no technologies similar to THOR in the near future, able to rapidly burn through multiple drones. This price point will likely not hold for a tactically useful drone in future conflicts. Missile and drone technology will likely continue to converge, as we're seeing in the Shahed line, and so will costs. This is not how what you know as "AI" works at all. Modular is one thing they are not. You also seem to be conflating "game AI" with the recent developments in LLMs, (which are mostly irrelevant to drones except for image recognition). Those two are completely different terms. In short, a "game AI" is actually a purely deterministic, algorithmic decision making system, which, given conditions, will take an appropriate action. You can use machine learning to do this, but this loses you the modularity. Any ML algorithm is a black box, you train the AI, but you don't really get to look inside its "brain", because it grows organically, and there are no clear "modules" you can separate. It's still deterministic, except it also becomes unpredictable, a very undesirable component in a weapon. Outside marketing buzzwords, Loyal Wingman will use algorithms that a human can understand, although it might be called AI because that's what games use to control NPCs. Nothing wrong with that, either, a good algorithm will take you a long way, and human pilots also follow them to a large extent. They could use limited ML routines for when the (human) leader commands them to engage. As an aside, Moore's Law has failed us quite a while ago: https://www.marketwatch.com/story/moores-laws-dead-nvidia-ceo-jensen-says-in-justifying-gaming-card-price-hike-11663798618 Gordon Moore himself predicted, in 2006, that (paraphrased) "we have about 10-20 years before we hit physical limits". He was right. You can no longer double the number of transistors (not performance, that stopped being linear even earlier) on a chip as easily as you used to. You make them too small, you get quantum tunneling effects between logic gates. The "2nm" architecture currently in development doesn't actually mean transistors will shrink by 50%, there'll be only a small reduction in gate pitch, from 48 to 45nm. Moore's law is over. Keep that in mind both when talking AI and planning your next PC. You do not need to lecture me about the importance of logistics. You do, however, fail to actually consider the whole picture. However, consider this: Israel's military has a budget of 23.4 billion USD. That means that 20 interceptors, which cost 50k each, make up about 0.00004% of it. It has been estimated that Hamas had an annual military budget of about 350 million. Let's assume an S-3 rocket costs 3k, so 20 of them would cost Hamas 0.00017% of their budget. So you can see that in a 1:1 exchange, Israel easily comes out on top, despite their missile costing more. A more realistic 3:1 exchange (it takes about 3 Iron Dome interceptors to kill one rocket) makes it almost even, but still, you can't make Israel poor that way, because while Hamas is mostly its military, Israel could vote to increase its budget. Given that any rocket that gets through can make way more than 50k in capital loses (all it needs to do is take out one high end workstation, or block a major road for a day), Iron Dome is, therefore, perfectly economical, given the context it's used in. Israel needs a way to deal with masses of several thousand rockets at a time, without covering half their territory in Iron Dome launchers. Also, it's obviously better to do the interceptions cheaper, because they're not going to go away. They also need to consider other kinds of threats, because your enemy isn't going to want to keep up a fight they're certain to lose. As it stands, the limitations of the system had been laid bare. Therefore, finding a system to overcome those limitations is presumably high on Israel's list. Wars are not beancounting contests, simply because you can't shoot the money. To do so, this money has to be turned into bullets, artillery shells and fuel, which need to be delivered to the front. This takes time, and time can also be quite valuable. Attacking any part of this chain has strategic value far beyond the numbers game - it's not just about what physically got blown up, but about the capability you lose, and the time that it takes to fix or get around the damage. Hence, an "uneconomical" defense can still be worthwhile, if it prevents you from taking a hit to something you can't replace. See Russia, which can handily outproduce EU while being pummeled by sanctions, while EU is sitting on gobs of money, but they can't actually buy any artillery shells because they don't exist, and neither do factories to make that many. Sure, you can pay money to expand manufacturing capability, but that's the long game (notably, EU pledged to make one million shells, and failed to do so, despite clearly having the dough). Ukraine won't need shells next year if it gets overrun right now. Also consider that cutting the supply lines by sanctions had not, so far, worked on Russia, simply because they're able to get by without our money. They have reserves, they have material stockpiles, the have unpaid convict labor. In fact, Putin had decreed, by fiat, that workers can be made to do extra, unpaid hours, that they can't refuse, or be fired. In other words, forced labor. Sure, I think we can see what this will do to quality and worker motivation, but first, see what it does to your nice and clean dollar figures. If pressed, a nation at war can run on empty, though it will have long term consequences. The only sure way to disrupt enemy supply lines is to blow them up, even slave labor can't do much if your fuel refinery is a burning pile of scrap and the ammo factory has just been spread over half the country. As Napoleon said, an army marches on its stomach. Good luck feeding them dollar bills, but if they run across an enemy food storehouse, they'll gladly help themselves. The point is, you can't reduce military problems to bare dollar values, you have to consider how and when those dollars are used. FYI, this discussion seems to be growing exponentially, so assume any points left unaddressed were due to lack of time on my part.
  2. The rest of your points merit a longer response (and some slogging through Google Translate-mangled Hebrew), so I'll address just this one at the moment, with a simple question: how is that any different from a missile? If you have a "drone" with long range, a jet engine, an autonomous seeker and an autopilot, then you have, basically, a SLAM-ER. It will perform like one, cost like one, and probably even look a lot like one. All you can do is vary the size of the warhead, SLAM-ER is a pretty big missile, so if you shrink that, you'll shrink everything (warheads on FPV drones in Ukraine are small by missile standards). However, that will not reduce costs as much as you'd like, although there certainly is merit to an idea of huge swarms of small missiles. One of the major differences between drones and missiles is the control link. A DUKE-style jammer basically mitigates the drones' major cost advantage, which is remote control. It's overcome easily enough, by a more sophisticated MITL system, but sophisticated=expensive. This is also why onboard AI will never be used by cheap drones. In order to run anything that could be called "AI", you need a decent CPU. Which, as you know if you were looking for a computer upgrade recently, costs a pretty penny. If you want autonomous, you need to put a reasonably powerful chip on every drone, and that will run up costs, up to the point where you could ask, why not a missile swarm instead. The form of drone warfare that's most likely to revolutionize anything comes from the realization that you don't have to do that. Instead, all fancy equipment, even AI, can be located at a control station that doesn't get expanded with every shot. This, of course, has limitations of its own, notably vulnerability to jamming. Also, this single quote makes me question whether that article from which it comes from is worth anything: This is patently untrue. We are talking defensive weaponry, therefore, we have to weigh the price of defensive weaponry against the value of the attacker's intended target, not against the weapon used to attack it. If your burst of 30mm rounds costs less than the apartment building that would otherwise be hit, you're still coming out ahead, even if the drone that would have hit it is cheaper. Wars are not spending contests, although they sometimes look like they are. Spending more on defense than your attacker spends on offense is certainly suboptimal, but this is not "working against yourself" and is in fact quite tenable if you're not planning on winning the war on pure attrition (or you do, and have a vast reserve of resources you can only use for defence). Honestly, this whole quote reeks of the usual "AI glasses" and SF-like predictions about how it will revolutionize things. It is not a realistic assessment. AI is certainly going to be useful for target recognition, but that's about it. In fact, given the record in recent conflicts of human ability to tell between enemy command posts and schools or hospitals, I can see it improving on that. Well, either that, or we'll have war criminals hiding behind thoughtless machines. In fact, that is a good argument to preemptively ban autonomous weapons and unambiguously designate who are we going to drag into the Hague in case the ban gets violated. Just to make it clear that using a robot to commit war crimes doesn't absolve whoever ordered them. For missile guidance, you don't need AI. Poseidon and HGVs both have one thing in common: they're fast. Faster than most similar weapons, which accounts for the difficulty intercepting them. They do not use AI for anything. AI isn't magic (no matter how much techbros try to sell it as such), nor is it particularly superior to a well designed algorithm, or when it comes to evasion, to a simple random walk. Stick a microphone in the pitot tube and you'll have a perfectly serviceable hardware RNG to use for completely unpredictable evasion. There's no need for AI for strategic weapons, just going very fast towards the target coordinates, possibly with some random dodging before getting there. The article seems to be desperately trying to shoehorn AI into military uses that it is not suitable to. Conflating missiles and drones is just icing on the cake (and no, missiles aren't "clever", they use a straightforward, but highly optimized algorithm).
  3. They can burst from overspeeding them while in contact with the ground.
  4. I imagine in cold weather starts it could well be important. In such cases, the heater could be turned on well before the alignment to avoid INU trying to start up when cold. Does it actually work that way in the sim?
  5. Yeah, we need FFB, but that's only recently began to even be feasible, due to patent troubles. I hope this gets more common now, IRL you get feedback through both stick and pedals, especially in a warbird. That said, the F-16 is a force sensing stick without anything like force feedback, and so is the Hornet. I imagine a lifelong F-16 driver would have similar trouble fighting in the real F-14, given that this was the case for this lifelong F-16 simmer. It takes a lot of practice to get rid of bad habits the Viper teaches you, it does so much for you that the way you operate the stick is very different from even the Hornet, nevermind an analog jet like the Tomcat. I'm better now, but the transition was hard. Also worth noting, Mover and Gonky fly enough DCS that the "real jet vs. simmer" argument doesn't really apply to them. They're familiar enough with flying without the cues that you don't get sitting in a chair. What we saw in the video can be entirely attributed to Mover not being accustomed to flying something that doesn't hold your hand.
  6. That's why Russians routinely put theirs on masts, and some Western ones come with one already attached. This is not a new problem and it has known solution. They are popular in other conflicts, too, they just don't get the same kind of coverage. Another reason is that Ukraine is the big state vs. state conflict, with both sides having considerable funding and resources. Every other currently ongoing conflict is an insurgency, where one side has to scrape by with whatever resources they can get, and the other has enough conventional forces not to need to innovate (or they don't think they need to), on top of the fact that even the state in question is typically smaller and has less money than Ukraine. Hence, the drones used there are less advanced and the press is less interested in them. Insurgents use them all the time in Myanmar, for instance, and Hamas nailed a few Israeli tanks. It did, they just won't officially admit they ignored it. For all their reputation, Israelis were about as bad as the Russians early on. This was a human failure, not that of equipment. That's because they've been designed in the pre-drone era. Now that the drones are a threat, countermeasures will be developed. The US already has the DUKE IED jammer, which could probably be improved, and they've demonstrated THOR, which is an energy weapon that can take out entire swarms. Russia is fielding a variety of anti-drone systems, some of which have already been countered by Ukraine. It's only a matter of time before drones, except the most sophisticated ones, are rendered practically useless against a peer enemy. Of course, protecting against drones has a cost of its own, but they're not the wonder weapon some people think they'll be. AI can help with target recognition and identification, but for guidance, you don't need it. In fact, a drone is remarkably close to a MITL missile system like SLAM-ER, except slower, smaller and fully dependent on its datalink, but cheaper. You can use any old guidance algorithm, such as a contrast seeker, running on the control unit. However, while this could occasionally allow you to hit a slow moving helo, it still won't turn the drone into an AA missile.
  7. Indeed, and particularly in WWII, you can sometimes get funny results if you're shooting from inside the convergence range, causing the target's wings to be peppered with bullets, while fuselage remains intact (or, from tail aspect, the whole burst goes over or under the wings). In such situations it can be beneficial to aim a little off so that you rake the fuselage with at least one set of guns (pity the Spitfire driver who doesn't have the ammo or rate of fire to spray around like that and has to open the range instead). In jets, the guns are typically somewhere close to the centerline, so that problem goes away, at least. It's things like that which make guns only dogfighting fun.
  8. Some fixed wings do that, too. The Saberjet has a light come on if you pull throttle so far back as to prevent the generator from operating. In more modern ones this is not an issue, but don't take it for granted in every aircraft.
  9. I don't think DCS supports that, however I would very much appreciate if HB included some special code to recreate Pardo's push. Currently, fireballs ensue anytime anything touches in the air. And yes, IRL two Sabres could also do this by sticking the tip of one jet's nose into the other's tailpipe.
  10. That looks just like a G2, except a little worse, and its drivers are said to be a pain (but hey, at least it's not WMR). I've actually looked at it, but does it actually offer anything notable over the G2? I'd like my next headset to be an actual upgrade, I can hold out on Windows 10 otherwise.
  11. Also, if you want to do some solo practice, there's an excellent training campaign on NTTR, Speed and Angels, that takes you through a realistic F-14 training course, and also Iron Flag, which is an in-depth training campaign for the A-10C. Also, NTTR has a better price when on sale. Sinai has no campaigns and worse discounts, and it's still a bit of a WIP.
  12. Yeah, but at least it uses dedicated Varjo software, so it's not going to be bricked by the update. Pimax Crystal looks awesome, but it actually costs more than getting a glider pilot's license. Of course, when you add in the base stations, the Varjo is pretty damn close, too.
  13. Well, Vulcan has the "shot string" effect built in, simply by shooting on bullet after another. The hard part, of course, is placing your nose so that the cloud of shells coincides with the bandit's flight path.
  14. What are the options if I one doesn't want to touch anything by Meta ever again? Only high end headsets like Varjo Aero or Pimax Crystal?
  15. Drones have limitations, though. Notice what kind of targets they hit. They've proven their worth against static or near static ships and vehicles, not against mobile targets like helicopters or even ships under sail. A drone is guided by a human, meaning the target not only has to be much slower than the drone, but also that the drone can't be moving too fast. Of course, solutions exist, but such drones are more properly called "missiles". Sure enough, Moskva was sunk by Neptune missiles, not drones. Also note that at least at sea, these drone tactics are, technically, not new. Italian torpedo boats worked more or less the same during WWII, swarming expensive battleships with small boats carrying powerful explosives. Drones are simply a less risky take on the same concept. Accordingly, ships are already equipped with rapid fire cannons capable of countering speedboat swarms. Places like the Black Sea were always a challenge for large combatants optimized for high seas operations. Also, thus far, we're hardly seeing "drone swarms". Yes, there are drones doing observation duties, but that's not the only innovation. Tracking helicopters was never the problem thanks to a surveillance radar. What drones did so far was expand the scope of combat. It's getting difficult to set up a safe area, and even positions behind deep behind the lines can be attacked with considerable frequency. This means that concentrating defenses along a front line no longer works, and you have to assume a drone attack could be targeted anywhere. They also represent a low cost way of destroying high value assets with insufficient anti-drone protection. Blanking all drone control signals in an area would require a ridiculous amount of power. You'd basically have to jam all frequencies in a fairly wide range all around the jammer. It would also make regular comms difficult. It's probably better to build an ECM system capable of jamming a lot of specific drones, coupled to a radar capable of tracking that many. This type of warfare will no doubt evolve, and we can expect to see countermeasures make the drones much less economical than they are now, due to a need for expensive ECCM systems.
  16. By rifle or tank cannon standards it's actually horribly inaccurate (1/6 of a degree is 10 MOA), but for an aircraft cannon, it's just fine. Remember, it doesn't take many 20mm rounds to damage something in a modern aircraft. The big concern is scoring a hit at all.
  17. That does nothing to address the fact that we used to have a feature that stood in for DTC, and it got severely restricted. Of course DTC is a huge and complex feature, that's why the interim solution should have been allowed to stand.
  18. No, but HP, or the appropriate store, might refund you something that was sold as working and turned out to be unsupported and not working with Windows 11, an information that should have been on the store's page (because they claim it works with Win11). A 2011 HP printer was sold as working with Windows 7 (11 not being on the radar then), and it was working for the expected lifetime of the product. If you install Windows 7, it'll work. A Reverb G2 still under warranty, bricked by a Windows update that was already announced when the sale was made, would absolutely qualify for a warranty refund, unless the customer was explicitly informed of the update before making a purchase and decided to do so anyway. At least, in the European country it probably would. Microsoft is simply the wrong target for a suit, because they didn't sell you anything VR-related, unless WMR was a selling point of Windows 11, which it wasn't. You wouldn't sue them unless you were making WMR headsets and they failed to inform you of the change. Again, it seems that you don't fully understand your rights (or you live somewhere where you don't have any) and how to enforce them. Reverb G2 is a HP product, so it's HP's responsibility to make it work for as long as they're supposed to be working.
  19. Where did I mention drones? Helicopters are being given trouble by effective MANPADS, ground fire and tactical SAMs. Drones have nothing to do with it, and are hardly the only lesson to be learned from Ukraine. Speed and range are good because they let you dodge SAMs better and make you harder to hit while on the ground (because your base is further back), respectively. Nothing whatsoever to do with drone warfare. They don't, however you do need to make it fast. Suicide drone are not a magic wand, in fact, they're quite slow. If the enemy doesn't have a drone launcher right next to the LZ, the drone would take its sweet time getting there. If a VTOL can get in, drop its payload and get out before a drone can get near it, it's got nothing to worry about. You do need to be careful choosing the LZ, though, because there are other threats that helos are vulnerable to.
  20. I think Iraq will come first, given that they already gave it a forum. I would not expect Afghan before that comes out.
  21. It's absolutely going to be an option. Those "agreements" don't mean jack. EULAs have frequently been ruled as unenforceable, and they're not allowed to make you sign off your rights granted to you by law. In other words, if an EULA tells you a lawsuit not an option, you can sue for that. Way too many people are cowed into thinking they signed away their rights by corporate propaganda, just saying "we're not responsible for our software causing you to lose hundreds of dollars" doesn't means they can't be sued if that happens. At least, that's the way things are in Europe. If US judges bend over backwards to say companies are above the law, you're kind of screwed. Then again, they didn't like McDonalds' using shady delaying tactics to deny paying one woman for treatment, and slapped them with a humongous fine. Mind you, that wasn't a class action, that was a single person who simply went all the way to court. So even in the US, you can win if you're dogged enough and actually go to court, as opposed to being intimidated by their army of lawyers. In other words, know your rights and enforce them. No matter how expensive a lawyer the have, or how big a legal team they turn out, if the law says you're in the right, then there's nothing they can do about it. Companies rely on arbitration, settling out of court and other tricks so that the case doesn't go before a judge. Because the moment it does, it's out of their hands, and they can only watch as the judge hands the plaintiff a good chunk of their revenue. Provided you live in a country where the justice system is worth a damn, and have enough money and patience to weather their delaying tactics (which judges often aren't fond of, either), it is quite possible to win. Cable problems are a well known issue with the Reverb. I haven't had any (besides a well known problem with the USB controller, solved by buying a USB3.0 expansion card), but quite a few other people did. And yes, replacement cables are ridiculously expensive.
  22. That would be best left to someone who actually uses VAICOM regularly, and is comfortable fiddling with it. I'm not. One would think that since he was charging for it, he'd have been more proactive than "open to suggestions" in ensuring it's actually correct. This was actually a factor in my decision not to buy it.
  23. That is true for many Russian fighters, but ironically not for Su-25, which has Western-style toe brakes. The Soviets tested the F-5 they got from Vietnam around the time it was developed and the brakes was one thing they really liked. The real Su-25T retained the system, though ours does not have separate toe brakes. The paid Su-25 does. Amusingly, the previous Russian aircraft to feature toe brakes was the I-16, where they were not hydraulic but purely mechanical, like a bicycle brake (the thing was notorious for having weak brakes, not that this was a bad thing in a taildragger). They switched to pneumatic brakes around WWII and have kept them for a long time, with Sukhoi and Tupolev eventually going back to toe brakes with hydraulics. MiGs use pneumatics with a shared lever to this day.
  24. This would be easier if a larger command list was provided. For instance, it won't understand "Chief, pull chocks." You have to say "Chief, remove the wheelchocks," which is ridiculous. If the commands were matched to how you'd actually say those things, as opposed to DCS voiceovers, it'd be easier to work with.
  25. Where in Nevada can you fly that? Afghanistan is not desert. It's rocks, mountains, there are quite a few arid areas, but even that depends on the seasons. By all accounts, its beautiful. While it is more brown than green most of the time, enough that desert camo was issued, the tan/gray UCP was more suited to it. Indeed, Afghan and tacky couches were the only environments in which it worked well. The point is, Afghan will be a real treat, and rather challenging to fly helos in due to high elevation and hot temperatures. Nevada could possibly stand for the rather sandy, flat Iraq. Not for Afghan. Even then, Iraq has mountains, too, in an area which we're getting, even.
×
×
  • Create New...