marlon445

What you’re describing here is AI-assisted work—something that has, I believe, already become a firmly established part of software development. However, I would point out that all the major providers rely on LLMs; while these can ingest vast amounts of data, they frequently get things wrong. Furthermore, if you point out errors to these models, they have a tendency to get stuck in a feedback loop where they keep trying to "correct" their incorrect information with even more incorrect information. It is also incredibly easy to trick LLMs by presenting actually correct information as if it were wrong. So, there is nothing wrong with offloading the tedious, exhausting, and time-consuming parts of scenery development—tasks that are easily automated—to AI. But handing over the entire development process to AI isn't the way to go. You could try it, of course, but then you’d end up with results like Microsoft’s Windows localization in my native language; it contains so many appalling textual and grammatical errors that it was almost certainly generated by AI without anyone ever reviewing it afterward. I also generally consider the development of scenery to be a less creative process than creating video games themselves. When you recreate an airport, the real-world model already exists; your task isn't to creatively design a new airport, but rather to replicate the existing one as precisely as possible.

The odd thing is that BATC didn't have these issues that are often mentioned for a very long time. But, for some reason, the programme is steadily regressing, becoming increasingly unstable and unreliable, and getting cluttered with unnecessary features, such as GSX integration. When I bought Beyond two years ago, I was really happy with it. These days, however, every flight just annoys me. This is because traffic is getting in the way again, controllers are issuing stupid vectors, and traffic still costs me 20–25 FPS, etc. I know it's Early Access, but that usually implies development, not continuous breakdown.

Exactly. I’ve heard the same thing from experienced A380 and A350 captains—that both aircraft handle with incredible agility, and you often find yourself surprised by it. I trust their feedback more than that of simulator pilots who say, “The plane is heavy; it has to handle sluggishly.”

I consider the flightmodel feedback from the community to be rather irrelevant, given that a large proportion of respondents to a popular survey assume they can land a real A320 lmao

It looks terrible. The exterior design is completely wrong, and the cockpit is a total mess. It's as if they just experimented with the wrong proportions and fonts. In this regard, no one can beat the ini A350; the exterior model and cockpit are among the most beautiful in the simulator. However, there are issues under the hood, particularly with stability (WASM, etc.), which is why I'm more excited about Rhodiumcode's system-level capabilities.

Is that because of Beyond or the Fenix? The last time I flew it, the VNAV was completely useless, and I couldn't make a single descent without using speed brakes the entire way down. I actually have a nice screenshot where, on final approach with flaps 2 out and a 12-knot headwind, I’m gaining speed at 900 FPM instead of losing it. Needless to say, I’ll never fly that thing again. In the A350, at least, I haven’t had any issues with the descent anywhere; everything went smoothly in Frankfurt yesterday. And the new instruction to fly over a STAR point at a certain altitude and then continue descending makes it even easier. The only annoying thing is that Beyond always makes you fly through the entire transition instead of using sensible shortcuts.

I once asked on Discord why Beyond always uses approaches that don't seem to make sense. In Singapore and Bangkok, for example, it always uses RNAV approaches instead of ILS ones. The response I received was that Beyond prioritises approaches that connect to the final waypoint of the STAR, and that many of these are not ILS approaches. Of course, it still doesn't make sense for Beyond to rely on a visual approach in the US (the flawed logic you mentioned), but it would probably be a different matter if the STAR led to the final approach. At KSFO, a visual approach is always used, whereas when approaching from the east at KLAX, I always get the ILS.

For me, TC leaks memory like the Titanic at every airport. I'd love to use it because Beyond Traffic has a significant impact on performance and Live Traffic isn't suitable for me as I don't fly in real time in the simulator.

I tried it out at Narita this afternoon and was actually very satisfied, except that, unfortunately, there were cargo planes at the terminal and many of the 747s in the cargo area had the standard livery, with both AIG and FSTL installed. Might hive FS Hud a try. It uses AIG plans directly and it's vectoring and spacing is excellent. I immediately deleted AIG Traffic Controller because it uses an absurd amount of RAM, which I will never tolerate (in Narita it was 14 (!) GB).

It wouldn't surprise me at all if they had illegally accessed Flightradar's data. They were stealing the AIG models with impunity until AIG rightly put a stop to it.

I just had my first WASM crash in over 8 months with the A350. Five steps backwards, again.

Exactly! We had SODE (SimObject Display Engine) for that, which could handle multiple jetways. Unfortunately, the developer didn't see any added value for MSFS and has taken his site offline and made it password-protected. Frankly, I'm also frustrated by the fact that boarding an A380 can take up to an hour when there's only one jetway available.

The A350 is one of the easiest aircraft to fly. In a positive sense. Airbus has really thought about how to make the aircraft as intuitive and comfortable as possible for pilots, and once you get the hang of it, it's like riding a bike. Some of my procedures only have 2-3 steps, the entire engine start-up, for example. But you have to like it, it's definitely not for everyone.

Since 1.1.7, I haven't had any problems with the aircraft at all. Sometimes GSX acts up a bit by only letting a single passenger board, but a quick Simbrief reimport and boarding usually solves that. In terms of pure flight, I didn't notice any problems in over 1000 hours last year. Not a single WASM crash either, PMDG's 777 is the new king in this regard for me.

I agree completely. That's why I'm always extremely suspicious, even of Microsoft. If they lose interest and abandon the project, we'll be left looking pretty stupid. The fact that almost all flight planning centres on Simbrief doesn't make it any better. PFPX was better in that respect.