Jump to content

Ssnake

Members
  • Content Count

    21,755
  • Joined

  • Last visited

Everything posted by Ssnake

  1. I run a couple dozen shots at ranges 2,500...3,000m, and pretty much all of them were first round hits (except for the cases where I misjudged the range). At first glance the M900 seems to perform just fine. Do you shoot? Or do you let the computer crew shoot? At what ranges to you shoot? If you do it personally, do you shoot from the thermal sight, or from the auxiliary sight? How do you establish the range; in the M60 only the commander has the range readout. Can you rule out crosswind as a factor? Is your vehicle in a level position, or at a cant angle? Do you hit with HEAT rounds from the same tank? Do you have the same type of miss all the time (rounds going over/falling short/left/right)?
  2. The M900 is type qualified (only) for the M68A1 gun (due to higher chamber pressure), which is mounted in the Stryker MGS, the M60A3, and the M1 models with 105mm gun (Jane's Ammunition Handbook 2009, 2011 (pg 379)). So, in principle, this round should be available. Why it misses, I don't know. I guess I'll have to run a few tests.
  3. If in doubt, go with the SB Wiki: http://www.steelbeasts.com/sbwiki/index.php/Main_Page There are no tuturials for the Leclerc because the Leclerc has no crew positions. But for the M60, at least there's detailed descriptions in the Wiki.
  4. But it fills ready ammo first.
  5. It's a bit more than simple scripting (air quotes or not). Steel Beasts is about as complex as a state machine can get before becoming unserviceable. Downside: Limited creativity. Upside: Predictable behavior. Often enough in training, you don't want the computer-controlled units to be too creative and rather err on the side of predictability.
  6. Congratulations, and thanks for confirming what I suspected. Now consider your chances, next time you raise an issue.
  7. Thanks. If possible, a tabular overview about what combinations of explosives you tried on which bridge with what kind of outcome would be super helpful.
  8. If some of you want to misread what I'm writing, there is nothing that I can do, and I'm not going to waste more time on this. The AAR is reasonably accurate within limits. Impact vectors are visiualized as thin prisms, approximating the shape of cylinders, so you can see them better than a simple line of sub pixel width. That a cylinder as a 3D object is something different from a vector line with but a single dimension is basic geometry knowledge. If you don't know the difference it makes no sense to continue the debate until you do. Also, the AAR visualizes what's been recorded. a. not everything gets recorded, to trim the AAR file size b. in network sessions where packet loss occurs, not all data may reach the computer doing the recording c. data are stored with a certain resolution - like 32 bit floating numbers, which are most precise for values near zero, and rather imprecise for large figures (due to the way how binary values are being stored) the four-second time resolution the "between frames" interpolation of pretty much every impact event (usually a hit is about to occur, or already has occurred before the current frame time-step was completed) character animation key frames and the interpolation of poses between them There is nothing that can be done about these things. That's what we mean when we say that the AAR is not 100% correct 100% of the time. Within these limitations, it is absolutely correct all of the time. But as engineers we feel compelled to point out that there are fundamental limitations, so people don't take our model for reality. Apparently that fact cannot be stressed enough. Now, WRT to crew served weapons, we can assign only a limited number of damages that can occur to the weapon itself. Only such defined damages may happen in the first place, and only to the weapon itself. In reality the scope might disintegrate into a gazillion of tiny fragments blowing into the operator's face at a thousand meters per second, probably killing him that way or disfiguring his face for life. The KE dart's overpressure wave would probably kill him. But we don't simulate those effects because Steel Beasts is a real-time application. We can only devote so many CPU cycles per frame.
  9. My take on the M829A3 vs M829A4 is that the new features that make the A4 model interesting are going to require some programmer involvement. For the cases discussed here - M829A3 in a flank shot situation, the A3 has so much power overmatch that you'd never notice a difference outside of a quantitative statistical analysis with a large sample size. I'd rather hold off with the introduction of the A4 because, right now, you wouldn't see much of a performance difference. It'd feel like fixing imagined bugs with placebos.
  10. The impact beam has a radius, for better visibility; the projectile's trajectory vector has but a single dimension, direction (and implicitly by frame time, length). The AAR is but a visual representation of a model, it's not the model itself.
  11. This is now bug #8060
  12. I honestly don't know what to say. If you choose not to read and try to understand what I'm writing in rather plain and direct text, that is your choice, but not my failure. I made clear that "more than nothing" would happen in real life. If you want to harp on that point, enjoy your strawman, but I'm notgoing to participate in such a phantom debate. We're modelling reality. Every model has its limitations and abstractions. This is one of them. Where we can, we try to err on the side that's supporting/reinforcing doctrine - in this case, that APFSDS rounds aren't the best choice against infantry. The vector line of the shot missed the soldier, so nothing happened. End of story. This isn't a bug. It's not a case of an insufficiently documented suspected bug. It's an (undisputed) model limitation.
  13. As was demonstrated with the tank commander's ear that was nicked by a sabot fin, we don't simulate the atmospheric supersonic overpressure wave. Our projectiles are points flying through the virtual space, a necessary simplification to conserve precious CPU cycles. If that sabot point object struck a location for which no damage was assigned - and we are somewhat limited in the kind of damages that we can assign to crew served weapons in the first place - well, then nothing will happen. I'm not losing sleep over this. You aren't supposed to shoot trucks or infantry with Sabots in the first place because they are not very effective. Here we are: No effect.
  14. Arguably, they shouldn't.
  15. These are now bugs #8054 and #8055, respectively. That said, a scenario demonstrating the reluctance to use coax would still be nice, so we can find out which trucks specifically don't get engaged by coax and why.
  16. We will certainly have a critical view at the component damage likelihoods in the Armata. I'm not saying that everything is perfectly fine as it is. 24% of shots passing through the vehicle and not causing any damage at all is not exactly what I'd like to see either. But we have to ask all these questions to make sure that there really is something that needs to be done. Wouldn't be the first time that someone is frustrated by a single game session and then starts blaming game design rather than accepting bad luck.
  17. Chances are, you have a version 4.0 license on your stick; SB Pro PE 4.161 requires a "version 4.1" license, which is available in the eSim Games web shop: https://www.esimgames.com/?page_id=1530
  18. I don't think so. Other ambulances are still being ignored. I think it's a problem of this ambulance model, specifically.
  19. No. It's a good point, but our AI isn't that sophisticated. They would know that they're dealing with a non-target. It's just, probably, that the magic bit hasn't been set. Besides, in real life you're usually expected to apply proper ID procedures (including an optical comparison to the thermal signature).
  20. There is a sensor for cant angle that should make the fire control system compensate for such a situation. This may be a refactoring issue. Can you provide a test scenario?
  21. It's somewhere in Bugzilla as an enhancement request. One day...
×
×
  • Create New...