Dragon1-1

Looks like HB reinvented the wheel with their in-sim web browser. This is not the greatest UI I've seen, but maybe we should remind ED it's a thing. They could develop this into something more substantial.

It's reasonable for us to ask which modules are at risk should anything like that happen and which aren't. That much shouldn't fall under the NDAs. Someone could, for instance, avoid buying specific modules that aren't covered, as opposed to avoiding buying anything new for DCS, period. It is not unreasonable for the customers to ask for this information, as it is in their interest to know whether their module is at risk of ceasing to work in case the devs have a falling out with ED.

In my experience, people who think AI is hot stuff had either never tried to use it for anything serious, or their mental capacity is not enough to produce anything better. For instance, it can write sitcom episodes that aren't that different from the human-written fare. Which says more about sitcoms and their writers than about the AI... It can be a tool used to alter existing imagery and text. Used with care, some people had good results with it. It's not a revolution some people expect it to be, but it's not completely useless. It just shouldn't be used to generate "original" text.

Just hold the stick full back and give it some differential aileron if you're drifting. If your landing technique was good you'll slow down to taxi speed in no time.

Show me an AI that can generate a textured 3D model that's usable in a game. This task is orders of magnitude harder than making a 2D drawing, and training data is much more sparse, not to mention not easily extracted from free content. In fact, this is the very reason why 3D models are more often sold than offered for free. There's also no easy way to steal 3D work, since 3D previews are a rarity, and most models are showcased via 2D images. That's before you factor in very specific requirements for models to be used in a game engine. As such, I don't see LLMs ever being able to do much useful work with 3D. Sure, an artist could figured out some way to use them, but a lot of legwork will remain to be done by hand. There seem to be some AI model generators, but I'm not sure how they work in real world use cases. Even if they create passable models for static renders, they're unlikely to be easy to convert to something that can be used in a game. I don't call that top tier content. That humans have made similar slop doesn't mean it's worth anything. It may make it possible to churn those out quicker, but it's a long way from the top. As of recently, it's being used for SFX in a proper movie, and it supposedly shortened some work that's tricky to do the traditional way, but that doesn't mean an indie director is going to be churning out the likes of Avatar. It can be a good tool, but only in hands of an already good artist, used for speeding up some tedious technical bits.

That's hitting just about every genre. Kerbal Space Program community was kickstarted by incredibly simple art style of the early versions, allowing new parts to be made quick and easy, on top of a relatively simple game setup. Every time the quality standards went up, the number of new part mods shrank. Nowadays only a few are making top quality PBR assets. Art for modern games aiming for realistic visuals is extremely time consuming to create. Some UE5 games still manage to have a thriving modding community making assets for them, but even then, it's nothing like the massive modding communities of old. Now, we've seen a proliferation of games which deliberately forgo fancy visuals and use something more stylized, often making it easier to make assets, but it's not really an option for flight sims, which attempt to get as close to reality as they can, and graphics are part of it, too (see the endless discussions about spotting). And then, you get highly elaborate coding on top of that.

I use Kingston Fury Beast and it's fine. Currently running 4x16GB (had 32GB originally, decided to do a quick upgrade to 64GB), no problems with it. Reasonably fast and not too expensive. The nice thing about RAM is, if you ever run out, if your initial build had two cards, you can add another two of the same model and double your capacity with minimum fuss. The performance hit was negligible when I did so. Go for 64GB now, you can upgrade if it ever stops being more than enough.

I think it was mentioned in the initial announcement that it'll be an option.

Why don't someone ask a real Mossie pilot? There are flyable ones around, someone who flew the real thing would know.

This is something that has to be fixed in a mission, and it's fixed by tediously blocking parking spots to make you spawn in a place where you'd get directed to an unblocked cat. Each parking spot only has a single cat to which the directors will send aircraft. If that happens to be blocked, they don't care.

I don't think I've ever seen a high fidelity sim do a proper Zero. There was exactly one that did anything near PTO, and that was way back. Japanese aircraft are noted to be hard to get documentation for (Imperial Japan having destroyed much of it at the end of WWII), so that will probably be the first time anyone has seriously attempted such a thing. Epic indeed. Worth noting that as far as '44 ETO goes, some aircraft would be shared between that and PTO, especially if we got multiple versions. The P-38 is a good example.

They're probably aware, but it's not just Winwing. It seems to be an issue with how Windows handles rapid DirectInput inputs. For Winwing to fix it, they'd have to code some sort of solution when turning an encoder quickly enough registers as a continuous press instead, but this would have to be on the device side, before the input is sent. I don't know if other encoders from other brands act the same, but IIRC my Arduino-based control panel had to handle a similar problem.

This is actually a problem with how encoders work. Since they send individual presses and not continuous input, this results in weird behavior. The mouse wheel seems to have special handling on Windows level that means it doesn't have those problems, but encoders on the Orion 2 are seen by the input system as rapidly tapping a button. My recommendation is to bind those encoders to radio channel knobs, not to heading and course. Rotating a control that moves in steps is a far better use of them.

Well, that's early access for you. Myself, I'm not buying anything until I work through my considerable singleplayer backlog. Given that I'm on a break from DCS lately, this will likely be a while. After that, I will catch up on campaigns for modules that I already own, and then, I will start thinking about buying something new. I'll probably miss out on the 30% discount for the MiG-29, but I'm on Steam, so if I have it at all it'll be a very narrow window. If I happen to have enough money freed up I might go for it, but in this economy I probably won't.

Yeah, but that's why the current LITENING was based on, at least where it wasn't more like LANTIRN. If that pod returns to the Viper, it should be a proper LITENING II, without digital zoom.