Jump to content

Box of Stardust

Members
  • Posts

    778
  • Joined

  • Last visited

Everything posted by Box of Stardust

  1. Yeah, I looked at the parts list and guessed that that's what the Modular Missile Manager was for. Tbh, I don't know how the Gunbrick reacts to that since I haven't tested it against the VS yet (and I won't be able to until Monday). I can tell you that it works kinda in favor of the PEGASys.
  2. The Brick works completely differently from the Viper. It's a hyper-Du-3, doing what that did, but better. Alternatively, it's a shorter-ranged Cookie. If you're looking for an improved iteration of something like the Viper, that's the PEGASys (which is now #3 after goduranus posted a few battles). Viper is #4 now. Only thing is, I haven't seen what the TFD 2.1 is like since I'm not at my computer to download it, so I can't actually predict what will happen since I don't know how the TFD 2.1 fights. I haven't even done a test of Gunbrick vs Vampire Squirrel MLRS (but I think it'll lose to VS).
  3. Yeah but you gotta go through Gunbrick first lol. It'll be getting at least #5 spot. Lol
  4. @goduranus So I assume that nutty MLRS device is the most effective 'initiative gaining' device you've tested. It's interesting. Goes off of the strategy that the initial missiles during the merge was always just for routing, so it doesn't matter that your aircraft can't actually get IR locks either, since the device does that already. I actually don't hate it or anything; little irritating, but not a huge bother. It completely dispenses with the 'initial merge' portion of these dogfights, which, quite frankly, is pretty welcome. I've only tested against my PEGASys-D3. Battles become more randomized; all depends on what the AI pilots do and how they decide to react. When and how they decide to turn back in to engage the Squirrels. Meanwhile, Squirrels don't seem to take full initiative and charge in straight. I think it's a side effect of them reacting to the PEGASi dispersing to evade the Sidewinders. Cool stuff. Ensuing dogfight partially depends on which side has better starting positions once the missile attack and dodging phase ends. Also seems to showcase fighters' dogfighting capabilities better. I haven't tested it against other entries in this competition, but I can tell you that my D3s aren't really fazed much and just start chucking missiles back at the Squirrels once they turn back in lol. Without an initial merge to worry about, their weakness in needing the AI to switch to Vulcans in time to do damage/survive the head-on pass doesn't manifest. Definitely adds a cool new dynamic to the dogfights. I'm adding these to my Basilisks lol. Oh, and Gunbrick, to see what effect that has.
  5. Yeah, that's a very sneaky cover pic you got on that page. well, time to test what the heck that crazy looking thing does --- huh, so that's what it does. At this rate, we're gonna go from gun spam to Macross.
  6. Yeah... I kinda just figured this out in the last half hour. If anything, it's just an additional step in aircraft setup; unlocking the fuel tank before takeoff. ? Anyways, here's another quick messing around, this time with the Lynx.
  7. Flying AI dogfight competitions. When they're approaching targets, no matter how much dry thrust an aircraft might have, it will switch over to afterburners if the max speed is set to above 300m/s. Max speed of 300m/s, it'll go as fast as it can up to 300m/s without using afterburners. Max speed of 301m/s, it'll turn on afterburners and then put throttle down to whatever low value it needs to keep a limit of 301m/s.
  8. Can the AI Pilot behavior be changed so that 300m/s isn't an automatic 'turn on the afterburners' threshold? Bit annoying that my plane has enough thrust to supercruise, but it'll turn on the burners anyways.
  9. Really cool. Here's my at-a-glance breakdown: Lynx: interesting, maybe a step up in the maneuverability department from the MF. Downgrade in firepower though. Also, as the rules are that battles are completely flown by BD AI, the drop tanks will have to be programmed such that the AI will drop them itself. Fortunately, that's as simple as hooking them up with a Modular Missile Guidance and set to GPS targeting and dumb fire once enemy is in x range, so the drop tanks drop when closing in to combat. Also, 500 drop tank fuel is excessive to begin with. (Also, after testing, they don't get fuel drained from them.) Oh, and not enough flares. Jaguar: bog standard manned fighter. Cheetah: Whiplashes are interesting idea, always wanted to test it myself but kept forgetting. However, arming with only Sidewiders will only lead nowhere: the leaderboard is currently filled with aircraft that don't get IR locked until 3km or lower. And have flares for days. Also, Whiplashes are a huge IR signature. It might be able to outrun Sidewinders though... if it flies to faster than 450m/s fast enough. Somehow I doubt the AI will do that, and it'll be in close enough range that most enemies will probably just gun run it. Lion: Uh, for the Lion, unfortunately the battles are supposed to be flown completely by the BD AI, so I don't think it'll be valid. For actual drone programming, @Eidahlil has instructions on page... uh, somewhere before page 25 on how he made Hive Princess work; it's some Modular Missile Manager magic mixed with decouplers. As for the take off... not sure. 200m/s is a bit excessive for takeoff, isn't it? That's outside of the AI's takeoff settings... and kind of outside the realm of normal takeoff speeds to begin with.
  10. The funny thing is, only the firepower really matters. The rest helps, but maybe not in the way you think. Did you revert to the correct previous version? And I assume your mods are still installed.
  11. Welp. @exbyde I'd like to replace the Basilisk-D5's spot in the queue with this weapon to surpass Metal Gear. TFDs, Squirrels, PEGASi- it beats all of them. All it took me was half an hour of building (well, and hours of previously done testing for my other fighters). Someone send a fighter to prove me wrong, because I really, really hate this thing. You can all thank the Du-3 (and to an extent, the yet-to-be-officially-tested X-15 [words]) for this abomination. Probably the only thing that'll beat this thing is the Cookie. I haven't tested for it, but it's the only battle where the math is much more in favor for the Cookie.
  12. WWE moves would be hilarious and awesome. You can leave the (varying scale of) seriousness to the rest of us (if you so incline).
  13. That sounds hilarious. I'm just not sure how well it would work in practice. Sounds like to properly work, it would cloud both of your IR locks. Which brings us back to winning during the gunfight merge, which is the most un-fun meta ever. I guess I could try integrating this with my goofy booster-launched aircraft idea (which is probably even less of a feasible idea).
  14. That's probably true enough. And with the drone as small as it is, BD RCS tool says it's tracked/locked at 3km, and chaff could probably do the rest in case of AMRAAMs.
  15. Yeah, I just couldn't get that 4t limit building a small drone. Had to go up to 4.5-5t to get what I wanted Ah... so it's FAR. With my regular KSP install having a ton of mods installed, I wouldn't have remembered what did which things for everything. Well, in any case, the BD AI seems to be flying fine with reverse-facing canards (with authority set to a negative value), so I guess they fixed the problem.
  16. I threw this together in about half an hour. It's a little better, some more steps in the right direction. You can study it and stuff for some pointers. I tested it against a barely tuned lightweight compact drone of my own (inspired by the Micro Fighter when it was first submitted). Unfortunately for the MF, it seems like I made my 5 ton drone still too vicious for it... and my crudely-tuned Basilisks lol. @goduranus KSC 8km starts allow the fighters to get to their set default altitude. Not sure how big of a difference that makes, but Viper never had a definitive superiority over the RS either. I usually do a 10km - 14km start if from KSC just for a bit more of fairness, since a lot of Sidewinders are set to be selected from 8km. Not sure how big of a difference that makes either.
  17. I have to admit that I downloaded the Micro Fighter when it was first submitted because it really intrigued me. It comes in under 4 tons, and yet has most of the right pieces in place that would make a good fighter. Also came to the same conclusion that it was too sluggish too turn. Needs some control surface spam, could probably get away with replacing the front wing with a canard (or a few) completely. Could also use some AI tuning; even one very specific simple change could change its lethality by a decent margin. I actually started messing around with the Micro Fighter because I was so curious about it, then tried to build a fighter like it. But it was just so ridiculously barebones that... I guess I couldn't handle not having decent aesthetics? Managed to get one under 5t though, but I need to see about that whole 'BD AI doesn't like reversed control surfaces' thing and figure out what that mod was that allows individual manipulation of pitch, roll, yaw authority. On the next battles: LOL I forgot what AMRAAM spam was like. That was definitely something.
  18. I like how we actually almost started playing Kerbal Space Program again with that last one.
  19. yeppers Also, the max speed setting does a lot of weird tangential things. But I wish pitching up at gun range was an effect of max speed above 275m/s...
  20. Tried it. Too inaccurate, can't compensate for that by carrying more because it's heavy, overheats were a bit problematic, and lugging around the weight of even two in a dogfight didn't go well. I mean, sure the point of long range gun sniping is so that you kill them before they get in a dogfight, but... it wasn't working out for me. Honestly, the best way to use the GAU-8 was how the Cookie did it. The X-14 posted earlier might be able to do it too except that it needs more engines and/or a lower max speed so that it can fly without afterburner. Maybe I'll revisit a GAU-8 platform later. I was also thinking about doing an Oerlikon Millenium Cannon build with the turret locked forward so it was basically just a really heavy gun, but I'm not sure if that's enough to make the OMC 'legal'.
  21. And yet, as a pure gunfighter, it's manages to hold onto its spot with a death grip. Not bothering for missiles, it just starts putting rounds down range as soon as 2.5km happens since it doesn't need to switch weapons, and IR locks on the Du-3 are only really made under 3km for the head-on pass. Winning against the Du-3 is much like winning against the Cookie: survive or win the initial merge. Out of the merge, the Du-3 is still a pretty damn maneuverable plane, so it's a tough dogfighter. Speaking of the Du-3, having to tune against it has annoyed me enough to play all ya'll games, so: (REPLACED BY GUNBRICK) @exbyde got another plane to throw at the board: Basilisk-D5. And it has one primary purpose: get the Du-3 off the board by throwing 20mm rounds back at it from 2.5km (assuming the Du-3 isn't off the board by the time it goes in for testing).
×
×
  • Create New...