Dragon1-1

Rich text file, actually, and it can be opened in Word, as well. It might not be associated with anything by default (used to be Write), but if you tell Word to open it, it will.

IMO, he really shouldn't have deleted the whole damn thing. He could have blanked out the first post to ensure roadmap isn't misleading, but the other content could have stayed. It's not like he had to do anything to maintain it. This thread was bigger than him and it wasn't cool for him to just delete it on a whim. He wants to take a break, fine, but that doesn't mean he's entitled to make the others' contributions into collateral damage. Maybe @BIGNEWY or another staffer could make an official newsfeed thread that would function the same, but not be at risk of being deleted like this. The roadmap thread was already premoderated in order to make sure only actual news were posted.

You should really consider doing a 90s reconstruction, because that's when the most interesting things happened there. If that's going to be a 2025 Balkans, I'm probably going to skip it. There's plenty of documentation for how it looked in the 90s, we really have no need for another 2025 map, especially since for most part, the only thing that got added was ugly "modern" architecture.

Libya alone is bigger than the entire South Atlantic map. Granted, most of it is quite empty, but even the coastline is very long. If that's ever to be a thing, spherical Earth is a must.

Tu-160 should be Mach 2 capable. Not sure if it is in DCS, though.

I meant you should get a GPU now and start saving up for a CPU+mobo immediately afterwards. Getting VR to run smoothly takes a very powerful rig.

It is correct, it's the A-10 picture that's unrealistically sharp. The Mav's seeker head is really not that big, its resolution is limited, which is what you see in the Viper. They should look at Mavs on other platforms, but boresighting is of unique importance to the Viper. It features TGP to Mav handoff, which compares images, not gimbal slew. While the Hornet and A-10 just point the Mav towards the same spot on the ground as the TGP, where it's locked independently afterwards, the Viper features the "handoff" mode, where after commanding point track, the image from the Mav's sensor is compared to the TGP and the missile is locked onto the same thing TGP is based on that. As such, if they aren't perfectly aligned (say, due to the distance between where the Mav and the TGP are mounted on the jet), the handoff will fail. This is why on a triple rack, you just need to boresight the first missile, since the distance between them isn't very big. Remember than D model Mavs are 80s era technology. This is not a mobile phone camera, but an IR seeker designed before Gulf War. Remember the state of your electronics back when you ran Windows 95, and then consider that the vast majority of tech that we use was made before then. The reality of the era is, pictures were blurry, displays were small, and many things that were marketed as "smart" were pretty dumb by modern standards. That's why the LMAV was such an advancement, buddy/JTAC lasing aside, not only could it use the much sharper picture from the TGP, it wouldn't be distracted by the "tactical bush" effect that affected other Mavs, as the human operator would keep track of what's being lased.

Start with the GPU, then look at CPU. VR has a voracious appetite for VRAM and AMD drivers aren't that great for it, so you stand to make gains there. However, since you're also going to be upgrading the mobo, I'd recommend you go with an X3D series Ryzen. The 14 series Intel CPUs had longevity issues, 12 series is fine, but it's also blown out of the water by more recent Ryzens. In VR, you'll benefit a lot from the expanded cache on X3D.

It's pretty crap, though, particularly when more than one person is talking (it doesn't tell you who said what). This is actually one of the few tasks AI is useful for. I haven't asked for a transcript, but I once got it to summarize an hour long video into a few concise bullet points. I'd expect it could manage a transcript, too.

I do hope they'll be better about making breaking changes to this one post-release. It made campaign builders wary of Sinai, and it'd be a shame if they were similarly slow to pick up the Balkans. Also, we could use more 90s aircraft and relevant or variants. Now there won't be any excuse for ED not to make Allied Force era hardware. Too bad about the Mudhen in particular, IRL it played a large role in the Balkans. I also hope we'll get CBU-94 someday, IRL it was only dropped from Nighthawks, but AFAIK it used a standard dispenser, so anything compatible with CBU-97 should be able to carry it. Could be very useful for attacking power plants in DC.

It ultimately depends on what you're trying to do. The MiG will usually be trying to intercept something over friendly territory. That means it can afford to go low, since ground forces are unlikely to be shooting at it. So what's the Phantom doing? If it's trying to CAP (say, to prevent friendly CAS from being splashed), it's at a massive disadvantage, because MiG will be in the weeds and the Phantom isn't that great at searching, anyway. If it's escorting, then it's a matter of keeping the MiG away from whatever it's protecting. The MiG will likely eat it for breakfast, but it might run out of gas or simply be unable to catch the strikers afterwards. If it's a mud mower of some sort... it had better have some escort, or it's going to have a bad day. OTOH, if the MiG is the one doing escort and the Phantom is an interceptor, it could try blowing through the MiG screen and killing the strikers, then hightailing it home at supersonic speeds. In any case, "unload and bugout" is only good when you can afford to do so. In most cases, it'll mean you just abandoned your mission objective. If you can accomplish your objective while leveraging Phantom's advantages over the MiG, great, but there's only a handful of fighter missions which would allow you to do that.

Even the SPO-15 (which is mostly modeled correctly in our MiG-29) could detect the F-14's radar at a much longer range than the F-14 could see anything on it. The simple reason is, an RWR needs to see the radio signal that has traveled from the radar to it, while the radar needs to see the signal that traveled from the radar to the target and back. However, with no launch warning for the Phoenix, seeing the Tomcat on RWR doesn't help much. That said, lack of AWACS at sea doesn't mean the Russians would be blind. It can be expected that the bombers would coordinate with surface warfare assets that would provide radar updates to them. The attempt to sink a US CVBG would likely involve a huge, coordinated salvo of ship and air launched cruise missiles. Aircraft would be working with their surface assets on both sides, and this would mean a huge bag of EW tricks in store for both of them. And then, underneath all that, you've got submarines sneaking around. Of course, my comment applies to ships, as well. You don't have to sink the CVB, a gaping hole in the middle of the flight deck is almost as good.

Wags mentioned it in the interview, they'd like to look into it at some point, but it's way down on the list of priorities. ED knows it's a big issue with ground units. Worth noting, there's a chance that it'll still cause some randomness. For instance, if the rounds gut the troop compartment, but spare the forward part where the crew are sitting, the BMP would still shoot back. It might even remain drivable. With spall liners everywhere, an AP round might just punch a hole. SAPHEI rounds aren't particularly common on aircraft, it's usually regular a mix of AP and HEI. The former punches a hole, the latter detonates on impact.

Hits, yes, but not penetrations. The angle means the round has more armor to go through and is more likely to bounce. That is modeled in DCS. As mentioned, DCS does not account the roof is just some 15mm of aluminum alloy. In fact, I don't know if there's more than a single armor value for the BMP-3. It might be surviving things it shouldn't due to simplified modeling. Also note, a single 20mm AP going through the infantry compartment won't cause the BMP-3 to explode into a fireball (though anyone sitting inside will have a bad time). In DCS, it won't take off all its hitpoints. That stuff is simplified to the point of inadequacy, strafing with 20mm in particular gets hit hard (30mm usually kills light armor reliably).

Worth remembering, the damage model for ground units in DCS isn't that great. In particular, IRL the top armor of almost all pre-2010s armored vehicles is paper thin. They're basically protected against small arms fire, maybe .50 if it's a particularly thick skinned vehicle, but not much else. Depending on the angle, the A-10 could punch through most tanks of its day with 30mm AP, as long as it hit the top armor. DCS doesn't model any of that, I'm not sure if there's any difference at all between where you hit, but even if there is some accounting for direction, the armor model is simplified, making top armor way too strong. I suspect that's in play here, realistically putting 20mm rounds in the side of a BMP probably shouldn't work, but they should go through the top no problem.

This is realistic, the submunition has a HEAT warhead about the size of a hand grenade. If it doesn't land directly on top of an armored vehicle, it won't do a whole lot, and it doesn't even fragment all that much. The visuals are fake because actually showing 247 distinct bomblets and corresponding explosion effects per bomb (meaning a 4-bomb ripple has almost 1000 of the buggers, and a fully loaded four ship of Hornets will dispense close to 8000) would likely bog down the sim.

Depending on the mission being flown, dragging and evading could cost the bombers enough gas to be forced to abort, not to mention in a coordinated operation, they wouldn't make the planned ToT and likely wouldn't be able to maintain formation. A supersonic bomber of that era would typically run in afterburner, so it wouldn't be able to keep doing it for all that long. Turning a bomber around at supersonic speed around takes a long time, especially if you're trying to maintain your speed. The interceptor's missiles don't need to actually hit the target in order not to be wasted. All they need to do is prevent the enemy from accomplishing their mission.

We've been seeing indications of that for quite a while now. Really looking forward to it. Will feel right at home in the sky with Eurofighter, the F-35 and the JF-17, seeing as India is one of the bigger operators. We really need a Kashmir map to run Rafale vs. Jeff and MiG-21 vs. Sabre duels (yeah, they've been for a while...).

There should be switches in editor for this, along with an engine pre-warmed by the ground crew. This was the actual practice in WWII, at least for fighters that were kept ready to scramble.

Now those are figures that are actual data that does support your point. You can now clearly see the slop was widely off base (and for some reason, it sold the Chinese variant quite short), even if the conclusion was the same. Well, dev time is in limited supply, for DCS it's in very limited supply. So the question becomes, when does making a Linux port starts being worth the trouble. For DCS, it probably won't happen anytime soon. While the idea of low overhead OS to squeeze more juice out of the hardware sounds appealing, the required software just isn't there. In particular, while Pimax announced VR support on Linux (I wonder if Chinese government's push for UOS has anything to do with that), most joystick software only runs on Windows.

False. You don't know what they're based on, just what the AI tells you. It might be using old data. It might be using wrong data. It might be using a random number that just happens to look like it belongs there. And the worst part is, you don't know, and you can't know. Not even its creators know, because AI is a black box. By pretending it's a search engine or an information source, you're saying something that's untrue. There are humans who compile those statistics, or you could have used something like Distrowatch. Does it give a complete picture? No. Is it more reliable than a random bundle of numbers made up by a bot? Hell yes. Whatever the reality is, what you posted has been dreamt up by AI, and is demonstrably false. You missed my point. I'm giving you flak because you used AI, not because I disagree with your conclusion. There's a difference between "lies" and "estimates". Anything produced by AI is firmly on the former side. The real data might well show the same trend, and had you posted a genuine estimate fact-checked by humans, there'd be no problem. But you didn't. You posted a machine-assisted lie. If it's accurate, it's by accident, and you don't even know if it is. Because you did nothing to check. It fits your preconception, but that doesn't make it evidence. My point right now is, don't us AI in a discussion with humans. Stop posting slop. Then we can get back on topic. He's posting results from a survey you yourself mentioned to be a valid source. Those results might not give the full picture, but it's well known where they come from, what are their limitations, and so on. You, OTOH, posted AI slop, which is literally worse than nothing.

I already did. Your list is off about Deepin/UOS by something like two million users. That's reason enough to believe all the other figures can be thrown in the thrash. This isn't a ballpark estimate, it's a bald-faced lie. You post autogenerated misinformation and expect others to spend time doing research to prove you wrong? Sorry, that's not how it works. Using AI to lie doesn't change the fact that you're lying. You knowingly posted something as a factual estimate, despite being fully aware it's pulled out of the bot's backside. Just FYI, I'm pushing back on this not because of whether this made-up "data" proves my point or not. I'm pushing back because you're peddling made up crap and expecting people to believe because it's AI. Anyone could type in some random numbers and present them as evidence, but for some reason, most people wouldn't do that. What makes you believe it's OK just because you enlisted AI to type them in for you? Well, it's not OK and you should be ashamed of yourself. It will never, ever be OK, so do everyone a favor and don't do it again.

If you got it from reliable websites, I'd have no reason to contest it. AI doesn't "pull up" data, it makes it up. In fact, if you compare the two columns, you'll plainly see a lot of BS. This is not proper use of AI, by doing this you're generating trash and normalizing something that is known to cause real harm.

Real data only, please. I don't know where the AI pulled those numbers from, but I treat anything it concocts as suspect. Just an example, Deepin's own data estimates its userbase as 3 million, well in excess of what the AI claims. Asking AI is usually a waste of time. That said, your post led me to an interesting tidbit, that is Deepin/UOS. Turns out, there already are OEM Linux machines... in China. They apparently have been trying to move away from Windows and Western-developed software to their own bespoke (and government-approved!) Linux distro. It's quite possible that China will end up taking a very different trajectory from the West on this. What it means for us, I don't know, but I do know it will complicate comparisons, since a big Linux surge driven by government-mandated Chinese installs won't necessarily translate into Western markets. OTOH, it might lead software companies to support Linux better, at least the ones allowed to do business with China. Interestingly, it turns out a lot of US-hostile states, including North Korea, have their own bespoke Linux distros for government use.

Impressive, but the question is whether they can sustain momentum past the first week. Apparently, the recent release made a splash. Also, FYI, it's a commercial (albeit cheap) Linux with paid features that's specifically designed to imitate Windows and in general be easy to use. Similar stories won't be happening with other distros simply because they, for most part, aren't specifically designed with that in mind.