-
Posts
652 -
Joined
-
Last visited
Content Type
Profiles
Forums
Developer Articles
KSP2 Release Notes
Bug Reports
Everything posted by hoioh
-
I think I got it, I was previously simply using the PM system of the site, which works fine for me. But I guess it's easier to have several threads on a single channel It may be prudent to add the channel details to the first page of this thread, it's currently 2 pages back, but that will change with time
-
For those of you that have already downloaded the SK-14 Razor Mk6: I've updated the design a little bit because I noticed some crucial parts were too easy to destroy. So now they're much harder to hit because I've moved them into the protective shelter made out of straight wing pieces that also houses the engines and most of everything else
-
Gave up on waiting and plugged the old 780 ti back in to make this: Download link: https://kerbalx.com/hoioh/SK-14-Razor-MK-6 Which I discovered can beat anything on the roster, including HSC and Missilebrick (reasonably reliably) Built it like this: And then added some extra missiles, landing gear, swapped out some parts, did some tweaking, etc. Then ran some battles off of the KSC grounds, lazy mans way:Against missilebrick: And against HSC: I ran more, but this post is long enough and the results speak for themselves I guess
-
The formulas are worked into this spreadsheet: https://docs.google.com/spreadsheets/d/1d7sQ29krOUyyFNCWxbpo-jJK_16te12qtuQh52RbN4M/edit#gid=1862528954 Start with a plane that already jas a review, read that review and fly the plane to get a feel for what we think is good, or bad, or anything im between. Then start on your test review
-
Regardless of being able to beat the record it's a nice excesize in building as well as piloting, so by all means go for broke and challenge your skills If nothing else it'll prepare you for a next challenge and it might give you a good feeling about your accomplishment
-
As far as I can tell: 1. Attach all parts to your core, then use the move tool to position each part 2. Have many parts, use them to make at least 1 spare for each function 3. Use part clipping to cram a great many parts into a very small core volume and protect that with more parts 4. Make the total volume small and flat, this allows for the best control of the AI, basically a 2 dimensional slab I set my yaw stabilizers to 0 input on newer models because they tend to induce roll more than allow the AI any control, or I set them to roll only so it works that way on purpose except for vertical stabilizers that are in the central plane and won't induce roll, for example between the engines Finally set all attach rigid and autostrut to heaviest, although not everyone agrees on the autostruts, it works great for me The trick is that when you lose a part you lose only that one part because there's nothing else attached to it and having spares allows you to keep control after losing a part. By protecting the core with basically excess parts, such as air intakes, you can lose a few and keep flying without too much trouble
-
While testing a very promising new desing I unfortunately discovered I'm one of those people that has bought a faulty RTX 2070 card, so I can't continue development untill my RMA gets approved and a replacement arrives Good reading btw! Makes tweaking the AI easier Still experiencing F9 issues though, after I launch a new fleet and set up a battle scenario to test everything work great, but after each F9-load KSP starts misbehaving more and more, which is totally weird. Sometimes one of the fighters becomes invincible (manually flown it into all the ksc buildings without losing a single part) or one or more of the planes just fly off into the sunset or start suicide runs into the ground, totally wack (Yes I am using the latest updates) At first it may look like savegame corruption, but it's a runtime thing because restarting KSP fixes it Anyway, long story, I got a design in the works that can beat the Zircon UT-7-B4 without fail, 4 battles in a row, couldn't test against the PEGAsys D7 unfortunately, maybe more next week!
-
Okay, so lose the radar and the radar guided missiles, they're dead weight. The drone will fire them early in the match and chaff will deal with them quickly. The trick is to make something mostly flat with some vertical stabilizers for balance. Canards on the front and control surfaces on the back will make it maneuverable. Make sure your vulcans are firing directly forward, any angle will be detrimental and put any missiles under the craft. More smaller eninges and TWR above 1 to make sure the afterburners cut out at about 300ms, which I found is a good max speed. Larger engines attract heat seeking missiles The rest is fine tuning of the AI to make it control the plane smoothly, but also quickly and make sharp maneuvers when needed, but be stable on the tail. Start with the settings of a similarly shaped plane and work it from there
-
I suggest downloading some other craft that do flip the way you like and see how the AI is configured, then tweak that untill you have a result you like The SK-22 is basically made out of control surfaces, so it needs the settings tuned down a little, but it also depends on the stability of your plane (how far the COM is removed from the COL) because the AI mostly has limiting settings. Also, getting it to flip too easily can cause the aI to do so all the time, which slows down your plane and makes it easy pickings, so it's important to test, tune, test, tune, etc fighting against the #1, once you can beat that, try the others in the queue to see if you can reach the #1 spot
-
Good to know that even though the Zircon is a bridge too far, the SK-22 can ward off most everything else I've upgraded the video card this week, so no more stuttering with 6 100 part planes in the air, which allows me to better test the MK-3 version I'll have to see what causes the planes to sometimes just run away, happened in a test battle betweeb the MK-2 and MK-3 the other day as well, even though the plane still has missiles and guns with ammo, sometimes they just choose to leg it for the hills Weird behaviour I also had a self-strike of a MK-2 a couple days ago, but so far the MK-3 has no such problems
-
Just a random thought on the design: You could put a fuel tank on the back of the shuttle with a separator and attach some engines to that, then attach the external tank to it, if need be with more engines and then jettison the whole lot together when you reach orbital velocities, or even after you bring the whole thing up to orbit Shaped sort of like a spear-throwing hook, or a (smoking) pipe
-
It's legal to use, but the OP has not decided the Kerbal content yet. As it takes a while for a plane to get reviewed, by the time your's is up for review I'm sure we will know what size they really are
-
Odd, didn't experience that myself. I've been tweaking the design last night and have a stronger model already, but I have a lot of AI issues. I'm not sure if anybody else has this, but after reloading a whole bunch of times I start to get errors up to the point where things stop working properly. I've found that by that time I have to close KSP and remove the 4 files that module manager creates, only leave the dll and then reboot KSP. This appears to fix a lot of issues. It may be that the stuck missiles are one of the symptoms
-
A number 2 spot it is then, I'm going to work on my next model some more and now know what to do to get an edge against the zircon In my own tests it was a coin flip to see which would win and I won't submit another craft until it's capable of reliably beating both the SK-22 and the Zircon Thanks for the cool footage!
-
That would be a case of pick and choose, we can compare the performance to both during review to see which category it fits best
-
I've been looking around for that info too, as far as I can tell KI stands for Key Input, which I guess is used to determine how long the steer input for maneuvers is held before recalculating. A higher setting will result in smoother maneuvers, a lower setting will result in more jittery maneuvers. I found that tweaking the value can change a craft's behaviour significantly and recommend you test differences of 0.01 between runs to find the best value for a specific craft