NathanKell

That's not a bug. If no node_attach is specified, the part center of mass is used as the surface attach node, and for KW engines the center of mass is, well, the center. If you're forcing on surface attach on a part with no surface attach node, it's going to surface attach at its center of mass. Sternface: Doh! Clearly can't see. Sorry. It looked for some reason like an aerospike!

KSP NTRs (You don't call all hydrolox engines J-2s or SSMEs, right? NERVA was a NTR*) are more or less correct in their TWR, though a hair low in Isp. TWR is around 1.8-2.5 for small NTRs, IIRC. Also, you really should use StretchyTanks, and KJR. Then the size of the thing doesn't mean lotsa-parts or wobble, just "more mass." You need to compare the same total mass ratio in a nuclear and non-nuclear stage; only then will you get a good comparison. Comparing a rocket with a mass ratio of, what, 10:1, to one that's about 2:1, is _not_ a good comparison. *Apologies; this isn't directed at you per se. I'm just really tired of people using NERVA like it's the same thing as NTR, instead of a project name/name of a few engines (NERVA Alpha, Gamma, 1mlbf, etc).

Nuke: http://docs.unity3d.com/Documentation/ScriptReference/Texture2D.Compress.html TL;DR: Not while using that command.

I'm working on the surface attach bug right now for ST (both attaching to them, and attaching tanks/SRBs to decouplers). Sternface: Nice! You might want to try with StretchySRB: you can have solids as big as you like for that missile, with custom thrust/burntime. Also, you might want FASA: It has W88s and a MIRV bus.

Yeah, and the Vesta and the Mav-1D have a fair resemblence to the single/dual RL-10. The real problem of course is that in stock MFS we feel obligated to keep engines' mass and thrust the same, whereas in my own engine work I feel free to change ALL THE STATS so that's where I spend most of my time. But from now on Chestburster is taking over maintenance of "stock" MFS engine stats, so they'll be a bit more like they used to be. (For a taste of how I see all the engines, see KATO-Engines or its successor, coming soon, RftS-Engines) Savage117, it all comes down to mass. Do you get more dV, or less, for the same total mass (engine+tank+fuel)? The answer for hydrolox is "way more dV". Even if you're not using cryogenic tanks. If you are, then it's way, /way/ more. Note that just because a tank is the same size doesn't mean it'll mass the same; a jumbo-64 containing LH2/LOx will have lower dry mass than one containing RP-1/LOx. Even though the total tank volume is the same (LH2 tanks are lighter per unit of volume). Now you might think, jeez, all these giant tanks mean so much drag loss! Well, drag losses are very minimal, so you at most might lose 50m/s from switching your entire launcher from kerolox to hydrolox (WAG). On the other hand, you probably have 1.25x the delta V now, for the same mass. You do the math. Finally another advantage of Hydrolox (in MFS) is that while the Isps ferram cites are available early on with Hydrolox (TL2/3), it takes until TL6 or so for kerolox to get those figures, so you're better off comparing hydrolox's 380-450s to kerolox's 270-340s. Much bigger increase then. MAKC: Because MFS rounds fuel quantity, and ST rescales in itty-bitty chunks (1% at a time), rounding errors add up during scaling. When you have approximately the right size tank, go into the tank editor, and remove and re-add your MFS fuel tanks; that will correct the ratios. The way it works internally is this: MFS normally rounds fuel quantities (to 4 sig figures, or to the nearest integer if > 1000). When you start rescaling a ST, rounding is turned off, but the rounded ratio (not the correct ratio) is the starting point. When you move your cursor off the tank, quantities are rounded again. That means you get rounding errors twice, but it's the price you pay for nice numbers in MFS. I'm considering turning rounding off entirely, however, since AFAIK most people who use MFS use ST, and does it really matter if your amount of fuel isn't a nice round number? erbmur: Ah, try changing the heat multiplier in MFS/RealFuels/RealSettings.cfg So that your engines don't overheat (who designs engines that overheat!?) I turned down heat production. But the emissive texture is tied to heat production...Your edit-guess was entirely correct. (lower-TWR engines I made produce less heat proportionally.) What I should probably do, since the heat mechanic is silly, is just use heat production to tailor the graphics.

Hey cool! Suggestion: add support for a CFG file with structure TextureCompressorBlacklist { texturename1 = true texturename2 = true texturename3 = true // etc } And then before compressing a texture, check if the node HasValue(texture url) and if so skip it. That way any, say, UI textures, that need to be left uncompressed, can be.

ANWRocketMan: Don't throttle down, there _isn't_ a thick atmosphere. Also you need to turn no later than velocity=100m/s. That should be around 0.3-1.5km, depending on launch TWR. A good launch TWR is 1.2 sea level TWR BTW, and you turn start at 1.3km or so. Ferram himself commented a page or two back to someone with a similar question. Any time you're not running at 100% thrust you're wasting mass (you could have used a lighter engine, and you're suffering more gravity losses than you could/wasting the Oberth effect). Re: #2, it takes ~1800-2000m/s to land on the moon; Apollo LM budgeted 2400 for DOI, descent, and hover. If you're getting 1800, you're flying a near-optimal descent.

Anybody is also welcome to try this hotfix (without Starwaster's fix) and see if that works v5.2 prerelease. https://www.dropbox.com/s/s4ainrqvwlxe2q7/RealSolarSystem_v5_2.zip

That's what L/L+ vs U/U+ _is_. A U/U+ engine isn't really an orbital maneuvering engine like the Poodle or LV-909; it's a (relatively) high-TWR engine with a vacuum nozzle and some chamber pressure increase. Like the LR-91 or J-2, or the NK-43. The poodle doesn't fit the bill because its TWR suggests it's a pressure-fed many-restart engine, like Apollo SPS. Actually the Skipper is a good example of a 2.5m U. But it would be nice to be able to toggle an engine's type, at least until we can have fully procedural engines.

Razorcane: Just find the total mass (in tons!) of each fuel, and divide by the "KSP density" of each listed here: https://docs.google.com/spreadsheet/ccc?key=0AvHneDAy4k99dHlhdktvZW1NS1lndlhNNnRwd3FEblE&usp=drive_web#gid=2 That is your units of fuel.

Yes, Unity will page if it can't allocate to its entire potential address space. If it can, however, it (buggily) decides not to page. So yes, it's best to make sure (somehow) that KSP can only access ~3.2GB memory

Ooh, sorry, thought you had merged that.

Its nozzle is closer to vacuum-optimized than most (and actually looks more like a double Vesta than a double Wildcat-5 IMO). And we actually do need larger upper stage engines, since we actually have none.* Almost all the "upper stage" engines released are actually tiny OMS engines with pathetic TWR; we need engines that can be used in the second stage, not the last stage. *Name a single upper stage engine in the 75-200 thrust range. Let alone one with decent TWR.

Zander: play with the SSF and PQSfade settings in RealSolarSystem.cfg. That's what controls the swap distance between PQS and scaled space. You want your SSFEnd to be the same distance as PQSfadeStart for obvious reasons. dimovski: usual figure is 9.3km/s. You can get by with a bit less, however, and it depends on how much you lose by firing in atmosphere where you have lower Isp.

On the semi-contrary: low power engines would be nice, but please don't make them short. The real SPS was twice the size of the KW one, but delivered only 91kN of thrust; vacuum-optimized nozzles are BIG.

And MFSC / Stretchy Tanks. You probably also want the various realism tweaks which I have sadly scattered all over and will be posting once I finally fix the bugs in ST to go with it.

theBlind: I'll look into it. Although KSP made it much harder to play sounds after .19 so people have been using hacks to get around it; that's why DRE plays the sound directly (and thus ignores KSPvolume). Frederf: Yes, RSS is "Real Solar System." And no, it's not going to split, there will just be a module-manager patch for the heatshields for RSS-level reentry. At that point, for your own purposes, what you should do is use the RSS-level heat shields but just cut the resource amounts on them to like 1/4. Then you get exactly what you want.

Yeah. Although AG was getting unreadables--we'd have to track down whether any unreadables are not already compressed and look into why that breaks.

Bump for an interesting find: At offset 0C of an mbm, if the byte is 00, it will compress (to DXT1/5 depending on alpha); if the byte is 01, it will leave as RGB24/RGBA32.

That sounds about right.

Of course.

According to MOARdV's tests, TGA-RLE was actually worse. That must be what I was remembering about the leak--it looks like when it uncompresses the TGA even that doesn't get freed. Sorry MOARdV, missed your link the first round. Regarding compressing: http://forum.kerbalspaceprogram.com/threads/49011-Testing-Compressing-Textures-in-Memory

Sorry! I totally missed this yesterday. Also, Dragon01, hope you feel better! The main idea is this: KSP operates off volume ratios. Each, tick it calculates how much of each propellant to consume based on the weighted (by ratio) density of all propellants divided by Isp and multiplied by thrust (basically, the opposite of the burn time equation, solving for mass rather than time, with time fixed at tick length). What you need to do is calculate the total impulse of the stage (thrust * burn time) and make sure it stays constant. If you add an additional propellant, and leave Isp as it is, burn time will lengthen since there's slightly more propellants to be burned. So. You first need the total volume of kerosene and the total volume of LOx. Set your engine PROPELLANT ratios to those for the two propellants (the exact volume as the ratio, not some percent like 60/40, but 30000/20000 or whatever). Now add a third PROPELLANT block, Oxidizer, and set its ratio to the total volume of H2O2 in the stage (H2O2 mass divided by Oxidizer's density). Now, you have to correct Isp. Multiply Isp by (total mass of kerosene and LOx) / (total mass of kerosene and LOx and H2O2) for both sea level and Vacuum Isp. That should do it, I think.

iVG: Need the F3 log. Do they just explode, or do they overheat/overG? The latter is a known issue with DRE, which I'm working on, but the best fix yet is F9 or revert. Sparker: Yes I plan to move KSC, or rather make many KSCs. I mean, my actual plan is to make Kerbin Earth, so obviously KSC will have to move. But I plan to have KSC, Kourou, Baikonour, Woomera, V-berg, etc. metaphor: Cool! Also, you don't need UR: since Planet Factory is a mod, you can just manually replace the textures yourself. The only reason we need UR is because Squad locked up their assets in KSP_Data.

Senshi, yeah, the power consumption / range balance for antennae needs tweaking. Working on it.