colmo

PMed about my entry not being included in the draw.

Thanks for that. It sounds like using FAR is the pro version of the format. Would you say stock parts are sufficient to allow enough design diversity, or are a few mods (e.g. Pwings) really beneficial?

How is FAR working out in this challenge variant? Does it make much difference to designs, relative plane performance or framerate?

As inventor of the format, my original clipping criteria was that structural parts were fine (for looks and aerodynamics mostly), but not fuel tanks into other fuel tanks, as fuel can only be so dense, requiring one or both to be partially empty. I also avoided clipping control surfaces because that looks bad (though wouldn't legislate against it, because there are limits given the parts we have) and never even occurred to me to clip engines (unless its a tiny bit, certainly never overlapping their outputs, because real jets don't do that). Ammo, guns (as they're fixed) and landing gear make sense to clip, because real planes don't have those bolted to the skin of the fuselage either. This contest was created to find the best performing designs (originally my own) which still looked authentic, not to design ghastly, unrealistic winning machines. Yes, people's ideas about what's acceptable will vary, so consensus is key. Alas, noses will occasionally end up out of joint. Btw, please confirm you got my entry?

I edited the CK-141A to have the changes above in a text editor on my phone. This may work. https://drive.google.com/file/d/0B15B3MQ2F4JTMnFadDhjODRXYmc/view?usp=docslist_api Clipped engines are lame, btw.

I won't be home in time to submit a new plane, so please reuse the CK-141A with following changes: Min alt: 400m Default alt: 1000m Steering factor: 20 (Max) Torque: off (as per rules) - - - Updated - - - Also, you'll regret having this annually as that's a very long time. There's some big exploits I can see in the rules but I'll save that for some future instaban fun...

I'm away from home this weekend. I'll submit a plane tomorrow - I could have linked an existing design but it's two points over 60.

The perfectly reproducible conditions are why I suggested just that. Having a change of scenery from the KSC is also nice.

Odd problem, that. Might be worth trying Konstructs on a bone stock install, might be another mod conflict? The idea of arenas is nice - we can use polar caps, deserts, or even mountain ranges to catch out the really low min altitudes!

I'll just leave this link about the RRS Discovery here: http://www.telegraph.co.uk/news/science/science-news/10425250/Setting-sail-on-a-voyage-of-Discovery-in-the-wake-of-Captain-Scott.html This is the kind of boat I want to make in KSP. I had the good fortune to spend a day on the much smaller RV Callista, which gave me lots of ideas for KSP - winches, probes, onboard labs etc. I also have in my head a series of stockalike crew cabins, one floor high. With hatches in the back and top, well placed ladders and a modular design, they could be stacked like layers of a cake in numerous configurations depending on how high, wide, tapering or offset you want. By being stockalike, they'd also be suitable as the bridge on deep space vessels or planetary bases. They could have varying capacity and function. E.G. labs or command pods. I know Fengist is not an experienced modeller and is focussed on the core of this pack, so I hope some of the very talented modellers here take a fancy to my idea

I don't claim ownership of the enterprise, I'm more than happy for someone else to do the legwork!

I suggested to BD that he add a 'minimum engagement range' for the AI, below which the craft would take evasive manoeuvres and run away before turning and having another go. Alas, I tried to design my little CK-141A this way (it's very light and slippery so accelerates extremely fast) but it would try to turn and engage far too close, getting chewed up in gunfire as a result. This is why cannons are so powerful in this challenge. I'm on my phone and not actually home from my trip yet, so updating OP is unrealistic! As the title appears locked, I'd prefer this remain the development thread and have a dedicated tournament thread for the fighter class.

We could require all reaction wheels to be turned off, as pods also have them.

I'm thinking the interceptor challenge would have something like a pair of heavily armed bombers flying in over the mountains to the KSC. The fighters would scramble from the Island, and thus their launch speed and acceleration would be sorely tested, timed to be touch and go whether they can hit the bombers before the bombers level KSC. This would need guard mode range increased to physics range. What's the distance from Island to KSC? The KSC would be littered with bomb targets, probe cores with a weapon manager and a RTG. Put a few on the VAB and SPH, and a few elsewhere, anything destructible. Currently bombs don't work with AI, but BD's development is coming along fast. To avoid bombers getting sidetracked they could use guard mode (for defence) + pilot assistant (hold course, altitude, speed, heading) instead of AI.

Once Kerbal Konstructs is sorted, I'm thinking about a JSF AI challenge. It would pit planes, not against each other, but against a number of scenarios, a triathlon of: - Scramble to engage in air to air against a numerically superior fleet of incoming planes, perhaps three or more. - Scramble to stop bombers before they destroy the KSC (need to figure out how to persuade bombers not to go chasing after interceptors!) - Attack a force of ground vessels (there may even be helicopters!) The principal purpose is to test the ability of the same fighter airframe to perform different roles (not just slow, fast turning gunboats as the dogfights favour). It's also a design challenge to make the scenarios work, which will be fun. Lastly it's a showcase for KSP and BD Armory AI. Scoring would be along the lines of degree of cumulative success or failure in each round.

A few other things to ban in the context of a fighter challenge: - oxidiser (I do want to allow RAPIER in air breathing mode - tempting to have a little oxy to get a rapid takeoff, but not very fair!) - unmannable planes (not having to pack a cockpit is a 1T weight advantage).

I'm on the way home and it looks like I have to bring out my best plane! It looks like the WW2ers have found a new home and normal jet dogfighting can resume. There's still a few things we need to tweak - reproducible starting positions, and some weapon costing. I've mentioned Kerbal Konstructs as a means of building custom arenas. I haven't yet had a chance to have a go at it. It looks like guns are the ticket, so they need to be valued accordingly. Missiles seem utterly ineffective in the majority of contests - is that a fair assessment? If so, either increase countermeasure values or have a hard cap on them.

Well done everyone, seems that contest went well. I hope I can find some WiFi this evening to watch the matchups (my 141 got whitewashed, ouch!) Has anyone looked into Kerbal Constructs to create 2v2 starting arenas? Doesn't have to be near the KSC or water, which might improve framerate. Noted on Tweakscale and Firespitter for WW2 variant btw.

Click Reply with Quote on the top post, and copy and paste the BBcode. I think it's just image tags round the URL Woot, a whitewash in the first round! It isn't my best plane either...

That would make an excellent variant 'WW2 fighter' class, possibly with Tweakscale and Firespitter too. I'd recommend toning down the hidden Vulcan with a reduced fire rate (in the weapon cfg), more consistent with WW2.Those old Hurricanes still flew when shot to ribbons!

I don't have time to test this, but this is a variant of my light single-engine deltawing fighter which, to my sums, comes out at 57 points. Colmo-Korp 141A: https://drive.google.com/open?id=0B15B3MQ2F4JTUWV2cC1Falk5WlU I'm pretty sure the M230 is undercosted, the Hellion only has two and is already pretty lethal, and is parsimonious with ammo so only needs one can per two guns maybe.

Bear in mind we're going to have 30m and 40m mirror ground scopes soon, v JWST's 6.5m - that's 20 times+ the equivalent light gathering capability. If making observations of a time sensitive nature e.g. exoplanet transits, that matters. JWST cost iirc $8 billion. The EELT has been costed at €1.3 billion. We could build several huge ground scopes, giving greater coverage, for one space telescope, and they're serviceable, so likely will last longer. The big scopes like EELT and TMT are in Hawaii and Chile, quite stable places. The worse trouble they've had was that issue in Hawaii with the locals being agitated, fearing the mountain god would be angry, but Chile is fine, and has a load of infrastructure already in place. I did say visible light. Well aware of the value of e.g. HST observing in the UV to complement ground surveys. I very much hope to get time on some of these toys in my planned future career (astrobiology), they all have their place.

It's also more cost effective to build ground telescopes for viewing in the visible spectrum. They're bigger, and adaptive optics make the atmospheric perturbations disappear. Space telecopes, like JWST, specialise in the IR and UV, which the atmosphere absorbs strongly.

I asked BD before if he'd be supporting VTOL with the AI. He said no, understandably. I've just spent a fun hour trying to get a helicopter to work and I think, with a few tweakable options, it would be possible. These could be combined with a docking port jr. (control from here) and/or a slightly angled main engine to, I think, make them possible. 1. Yaw control. He turned it off for planes, but VTOLs need it. 2. Pitch and bank limiters. Configurable limits on how far the vehicle will pitch or bank, because helicopters flying at 500m really aren't good at barrel rolls. Default would likely be 45 degrees, but varies depending on TWR and other factors. Set to 0, and it would only turn using yaw and gain speed due to thrust vector (slightly forward, probably). This would also be handy to planes with very limited control authority and TWR. 3. Deleted, rubbish idea, see addendum instead. Optional: 4. Speed limit. Planes cut throttle, but that's not a good idea with a VTOL. Instead, change pitch. Clarified below. 5. Firespitter interoperability. All the rotors use Firespitter, which has hover control. Min and default altitude could directly influence this (add or lose height until altitude criteria are satisfied) instead of altering pitch. I hope BD finds this post useful! - - - Updated - - - Ah, that took a little thinking - planes control altitude with pitch and speed with throttle. Helicopters swap those around - speed is a function of pitch and throttle controls altitude.

OK, thanks for the tips. I'm heading off for a few days (back Wednesday) so by all means keep testing and offering ideas for refinements. Have fun with the YouTube playlist. The King of the Hill idea is a useful testbed for the tournament, allowing some fun in an unstructured, ongoing basis. Depending on who starts it, it doesn't need to use the same rules either, and people can mod to their heart's content as the next person will know exactly what mods are required - I ache to spring my ultimate weapon on you...my Death Zeppelin!