yeah lets move it to the PUMA thread probably the vehicle plauged most by gaijins code
4 Likes
It’s not the worst I’ve seen, it’s perfectly readable syntax but I’m not that familiar with JSON.
Discussing the SLM code is a little different than discussing coding in general lol
1 Like
Yeah that’s true and the reason why I should have put a \s at the end of my post ^^
putting each value in an extra line just makes the code unecesarily long and harder to read, especially since each of those values are magic numbers anyways
1 Like
I spent some time looking at it, it’s just the object responsible for the double loft. When you launch the missile it first goes vertical until a certain altitude above the ground, then this block of variables will take over the guidance to “orient” the missile in the right direction, and tells it to head over the target in the weird shallow angles we have in the game. Hence, the weird “double loft”.
But I dunno man, it’s been ages since I looked at coding.
1 Like
the fact that you had to look at it and think about it for around 15 minutes just shows that it is badly readable code
big point of that are the magic numbers you tried to figure out here
I mean, it’s pretty self explanatory what’s going on in that block of variables 😭, but I wonder why Gaijin can’t code the orientation logic before launch, why do they need an orientation and heading “autopilot” after the launch, that wastes energy and time and creates this weird double loft that makes no sense in game and IRL.
So in my humble theory, what’s happening here is the following:
1st stage: Missile goes vertical until a certain hardcoded altitude is met.
2nd stage: Orientation autopilot tells the missile which heading it needs to go, levels the missile relative to the horizon and tells it to gain altitude at a shallow angle, creating the weird double loft we see in game.
3rd stage: Datalink logic??? That’s my best bet.
Feel free to add to this, as I could be totally wrong, but that is what I feel it’s going on behind the scenes of the launch logic of the SLM.
My guess is that the double lofting will be more egregious in the cases when the target is far away, as it commands the autopilot to loft harder after the missile has been leveled and set in the right heading.
1 Like
- letting the missile launch vertically out of the tube
2. missile turns towards the target
3. missile starts lofting
and the rest could be whatever
this seems to be some logic for how much it is supposed to loft
this stuff seems to be some added modifiers that dont inherently change the guidance itself
1 Like
This is a good analysis, I just don’t get why Gaijin can’t simply forego this flawed logic, the vertical launch should last longer depending on the target distance, and altitude. I guess this code works for short distance engagements but falls apart the moment the target is far away.
If anyone knows how to tinker with this in game, they should adjust the vertical launch parameters so that the autopilot doesn’t have to compensate as much for the short vertical launches.
And looking at the pictures from this report, it looks like my theory of short vertical launches might be right
https://community.gaijin.net/issues/p/warthunder/i/hO2zvBbejhAJ
Short launches → Autopilot has to compensate by regaining altitude twice → Double lofts in-game.
The missile does not loft twice. You can see it in your own screenshots
))
5 Likes
Together, we can solve the puzzle!
The lofting is its own code.
SLM loft code
People calling the orientation phase + loft a “double loft” is fundamentally false. The orientation phase tosses the missile out of the container and then forces it to pull 90deg to end up flat, the missile then gets to start guiding after that using the rest of the guidance code.
The real question imo is why does it “need” to pull the 90deg turn to start flat. We know missiles with loft code dont need to be flat to work in-game, since the 9m317 from the Buk-M3 doesnt have an orienting phase.
I also wonder how the Aster 30’s flight path compares, since there are some differences
Aster 30 orientation code
That’s kinda what I said, the “Autopilot” does not need to level the missile.
I have the feeling that Gaijin could not figure out how to make the vertical launch last longer or vary from target distance/altitude so they settled for this weird solution of making the missile go horizontal and then have it climb at a shallow angle afterward, thus, making this weird “S” shape loft on launch.
I wonder if letting the vertical launch last just a little longer, say, 150 to 300 meters altitude and then let the normal guidance carry over without the flaws of the autopilot would help a bit.
Because to me, the vertical launch altitude is hardcoded, and that’s the main issue with the missile’s flight pattern, it simply cannot properly loft when the vertical launch is so limited in altitude and the autopilot setting the horizontal orientation for an angled climb at such a short distance from the ground, instead of simply letting the missile climb and course correct without the need of the horizontal setup.
Thatd murder its short range capabilities. I was thinking in the other direction, cut down the duration of the orientation phase to something like 1s so it doesn’t pull a full 90deg turn but still gets somewhat pointed in the right direction.
3 Likes
Yeah, that could work too, maybe cutting the orientation phase to just a few degrees instead of the hard 90° pull to level itself with the horizon could help a bit?
But that still wouldn’t be a real loft, just an angled climb shortly after launch and not a true lofting pattern. Because from the footage we’ve seen, the missile climbs to high altitude vertically before it begins to turn to the target.
Hmm maybe i could fix that
4 Likes
