No there is reason to that the best Pioneer modeler has fear. Other ships yours, should be inside. Be patient.

Many enjoy your creations, even if they are "out" . <img src="' http://spacesimcentral.com/forum/public/style_emoticons//gamer1.gi f"' class='bbc_emoticon' alt=':gamer:' />

Most important is that you enjoy creating.

[background=rgb(33, 47, 58)]"i fear that everything will be restricted to a type of models like my shitty "cobra1" that's all.[/background]

[background=rgb(33, 47, 58)]i fear to loose the freedom i had until."[/background]

It is impossible for that to happen. It would be a serious mistake.

thanks walterar, i know that (most of?) the community likes what i do, on the other hand,

i'm not so sure about how the developers think about that.

it seems to be really a big problem for me,

(i've had a nap inbetween and a weird dream which has been released by this fears, i saw a lua script rolled off before me, every line consisted in the beginning of a lot of hieroglyphs to me, slowly it turned to readable text and it turned to be comments to what i've scripted, you can imagine no good things, it was really weird, but don't give a damn about, it was just a dream. - no i'm sober <img src="' http://spacesimcentral.com/forum/public/style_emoticons//wink3.gi f"' class='bbc_emoticon' alt=';)' /> i only dream or to say it right, remember my dreams when i'm sober, i have more fears and doubts also. i hope this wil help at least to dispel some of the doubts i have).

I've worried about the future of scripted models too. I've put a lot of time and effort into them. When my upgrade of the pad stations was added to the game I was reassured that there's still a place for them. It also makes me feel good that people actually use the mods that I've uploaded. I'm still working on them because I enjoy doing it. It's a wonderful rewarding hobby and I think all the contributors must feel the same way.

Scripted models are going away - this has been discussed to death everywhere, so I'm not going to go back over it. However:

• We want a good-looking game
• We want a very low barrier to entry for anyone who wants to contribute
• We want a game that can be easily maintained, freeing us all up to make new things

If those are goals you can get behind, then you'll be happy. If you get stuck on a specific piece of technology, then you're going to be disappointed.

"[background=rgb(33, 47, 58)]Scripted models are going away"[/background]

There are many other things that will go away if this happens.

There's very good reasons for removing it, not just from a performance perspective but you've got to understand that whilst you guys have invested the hours needed to get used to it there are many other artists that want to contribute who are put right off by it and never come back. Since I joined the project there have been half a dozen people who've come along, tried to contribute, hit LMR/Lua and vanish forever more. The artists that I know professionally take one look at it and laugh their arses off and refuse to contribute unless I do all of the LRM/Lua side of things.

That's why it's going away, because it's a barrier to getting other peoples contributions. The other aspects like it performance, technical and asset pipeline limitations might be acceptable but the truth is that a more typical modelling style would arguably be more flexible too.

@robn

If there's demand for it then I could take over porting/extending the LMR exporter work you started? I could bring that into ModelViewer or another program based on it, then people could create LMR/Lua models in that and hit a big red export button to convert them to ".obj" or some other format supporting animation?

Andy

That matches our experience also. We've watched a lot of willing contributors walk because of it, and we've talked to others to ask what they need. For the most part, they just want to keep using the tools they already know and not have to faff around a bunch of scripts. And I support that completely.

To anyone who has concerns, please take the time to look at the work that's already been done on the new model system, and to ask about it, and find out where the shortcomings are. We're not being stupid about this - we're not trying force everyone into making ships that look like boxes. On the contrary, we'd like you to be able to just make whatever random thing you can think of in your modelling package and it just works.

Look at every game out there. They pretty much all use this style of asset pipeline. And they all look and feel great. Not that its magic fairy dust; there's still a hell of a lot of work to be done to achieve that. My point is more that if you're worried that some particular thing you can do and need to do (that point is crucial) in LMR won't be possible in a conventional pipeline, then speak up! But be specific about it - a vague "I don't like it" won't get you anyway. I tend to ignore those.

You're more than welcome to it. It's on my lmr-exporter branch, though it'll need to be brought up to current master. I've held off doing anything with it until the new model system lands, but if you want it, go for it <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_e_smile.gi f"' class='bbc_emoticon' alt=':)' />

And now for the actual reason I logged on this morning :/

I like a lot of the models you've done, there's no real core art style for Pioneer at the moment, it's often described as "up for grabs", in fact I think this was a large part of the motivation for the modding system.

It means that there can more easily be a "PotSmoke66" mode featuring only the ships made by you or a StarTrek mod etc.

Please do feel free to go crazy, some ships will make it into the core distribution, some won't but don't let it stop you. I'd like to see more ships like these but that's my personal taste and it will put a lot of other people off.

If I get my act together and make a mod with ships like those then they might not go into the core version but they'll still be out there and playable.

My advice: go nuts, then put it all into a mod <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_e_smile.gi f"' class='bbc_emoticon' alt=':)' />

Andy

Cool I took a look at it when you were working on it but that was a while ago now i think.

Uh oh, morning meeting time! gotta go <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_e_smile.gi f"' class='bbc_emoticon' alt=':)' />

edit: i'm really a slow typer, haven't finished mine 4 new was coming in meanwhile <img src="' http://spacesimcentral.com/forum/public/style_emoticons//swoon.gi f"' class='bbc_emoticon' alt=':swoon:' />

wait, i wil miss that to, but i guess i started already to accept that it will be that way, i can't decide this.

on the other hand... i guess i will miss comments like i received in forums like i.e. GLBasic, most devs their was liking the idea of scripted geometry.

some prefere even old FE2 to what is state of the art, like "i would rather like to see old style polygons flying around as all this textured stuff".

but like i said i can't decide this by myself and i have to accept what's ahead.

though i guess that's not the point, i started to accept this, i can see that it's needed (with a tear in my eyes).

neither i think that it would be a really good idea to export scripted models to wavefront, i did this already in certain cases, but to say it simply with "coolhands" words,

"i can't see any reason to invest many hours in something that can be done in a few minutes", that's true.

the result of scripted models is in general very rudimentary, and it's not enough to have a idea, you have to find a way to get this solved (besides that counts for CAD softs even, you wil have to find a way to get a idea transformed to something visible).

of course to me and some others this is the challange to find a solution for a idea.

i look at this like, you can't make with scripted geometry in certain cases what you can do with a prog. like blender and if at all it costs you a lot more time.

not to talk about texturing, to texture a scripted model is very limited, you can work only with projections, no per face texturing is possible and that's a big limit.

there is of course,

a) the difference in "pride" you get

<img src="' http://spacesimcentral.com/forum/public/style_emoticons//cool.pn g"' class='bbc_emoticon' alt='B)' /> (erm b, it turned out to be a smiley, cool) certain things can't be solved in cool way with wavefront or whatever standard, standard is the right word i guess, it's a standard and will limit certain things to this standard. besides even the LMR is a standard and has it's own limits.

but i guess exactly the limitation is what i fear, i fear to loose certain possibilities which i liked since the very beginning, not to talk about tomms own words "...no shitty matrtix animations..." now i felt with him from the very start. i liked to animate stuff with exp<b></b>ressions, it has limits to, but it's far more "sophisticated" if that is the right word for it. i guess i liked this most of all about the scripted models.

now i tricked a lot for FFED3D to get more as one (the landing gear) animation to work, that's not because it wouldn't be possible with directx meshes, it's because it's not supported fully in FFED3D (and yes standard is only one animation for them, even MS's Mesh Viewer can't show more as one animation). but yeah let's say think about that, don't limit it because you think something isn't needed. also please think about a support to animate vector groups and i'm almost pleased. to sad i haven't started yet the "Turner Class", if i had done it everybody would feel such is a "must" to have, the "Turner Class" needs a waving tail, else it's no "Turner Class", to me at least. here we come to a point where scripted geometry has it's brillance, make some bezier stuff and animate it in a "simple" way, even vector group animations can't replace that fully. not to talk about the idea i had once to replace the station personnel with a "max headroom".

i can't see a proper solution for such without to script it, (that means "i can't", perhaps there is one, but i don't see it, but the main problem isn't that i can't see this solved with vector groups, the problem is for me how can i link this dynamically? how should a vector group animated mesh read ever any text? or any else variable? but i can see it right in front of my eyes with a scripted geometry)

it would be something, i tell you, anyone will have to kneel in front of it. back to the animated vector groups, i feel that they will lack of performance as if i transform a wavefront .obj dynamically, am i wrong?

what has driven me at all to start this thread?

actually it was loss of the cockpit view, i know many here like'd the idea, whatever might have become of that.

actually we have now views in all directions, but i can do nothing with it, i have no use for it? really not, not even in the game, i don't need it.

further i can strip now all this from some models, i could be "evil" and say they wouldn't need any cockpit or pilot at all now.

personally i would say it's good for nothing, but i guess/hope this will change, i don't know.

let's stay wiht this idea, please be aware that many players simply love to see the cockpit/pilot/crew, some players like to look out of the ships window and don't care about structural weaknesses that windows might give at all. we love views like in SW, you look outside and see your own ship, that's it.

i snooped into a opensource flightsim and was fantasized, wow! i can move around my view in the whole cockpit, nearly even look under the seat to see the dust under it or a dropped chewing gum. that just to give a idea what can can be done.

i know a lot would like such, it has no real use neither yes, but it looks fantastic.

besides above described views i can't see solved with scripted models in a easy way, this is fact to. it's damned complicated to make good looking scripted geometry at all.

n, n, n... nineteen 83

[attachment=1466:Bildschirmfoto 2012-10-29 um 11.11.22.png]

it would be only half as nice if you wouldn't see the tips of the x-wing, no question.

(for interested retro gaming freaks, StarWars (rev2), Atari corp. played on MAME, erm something i always disliked such big scores, you get rewarded with 1000pts for a shot tie, you need many millions to be halfaways good, what's the purpose? they could count 1 each shot tie it would be exactly the same, only lower count in pts. but i know also it was (is it still?) a criteria back then, makes me laugh)

---

@marcel, if all fails you can use GLXtractor, i installed it on my new PC and it runs "out of the box", no else softs or libraries needed, it's not to complicated to setup, just disable texture and shader export if you don't need it (which i assume and also recommend), but of course keep most other options and foremost "export UV coordinates".

you can use it to make a snapshot from the pioneer modelviewer to get a wavefront .obj you can use afterwards in i.e. blender (texture coordinates will be flipped vertically <img src="' http://spacesimcentral.com/forum/public/style_emoticons//wink3.gi f"' class='bbc_emoticon' alt=';)' /> (it has to be flipped if i have once righthanded and then lefthanded coordinate system or vice versa, if in doubt use your hands, that's not forbidden even some astronomers/mathematicians do this ("this one" as "harald lesch" stated once and he spread his fingers, lol). further if you do so, anyone can understand why they are flipped vertically). the model won't be centered on the y-axis, z and x are centered proper (use the most front y coordinates of the script to get the model centered on y then. of course the snapshot will be scaled like it's in the game, i.e. i prob. have most front y vector at 20 and scaled the model 0.5, the most front y vector is... right, 10). to get the different LOD's simply zoom out resp. in, take care not to rotate the model, else the snapshot will be rotated to (of course). sometimes it terminates the modelviewer (or the game, if you ever like to use it to snapshot a whole scene from pioneer) when you use it but the snapshot will be ok. don't be surprised it exports all the shit on the screen to wavefront which has a sort of geometry, even the sliders. to import the model to blender make sure to enable "sort by groups" (our scripted materials won't be recognized by GLXtractor) is active, else it will be a single mesh, if "groups" is active any single geometry will be imported as a single object. to export a animation (position) you will have to use the sliders, means of course only what is visible will be "exported" resp. written.

you will also recognize that the UV isn't on the "first repetition", that's because UV coordinates start in the lower left corner, we usually center a UV of a simple 1x1 quad i.e. with v(0.5,0.5,0) this moves the center to the lower left "repetition" (besides repetition isn't the right term, imagine a texture will fulfill the condition a sphere, it's a endless surface, even when it's a two dimensional rectangle in fact. but we use to look at this like a endless checkerboard). if you would use v(-0.5,-0.5,0), it would be in center of the first "repetition", again this is bound to the righthanded coordinate system of pioneer. (this only because you might have wondered perhaps why i use negative values often, but it won't matter it makes no difference, except that a snapshot's UV will be centered on first "repetition").

yet another hint, this can be used for other cases to (i.e. if you convert a model from SW-XWA). if you have such a displaced UV, like the above case, it's "displaced" on the first lower left repetition, simply use the textures size in pixels to center it in the "first repetition". means select the displaced UV and use "grab", enter the direction it should be moved (either x or y), enter now 256 if the texture is 256 pixels wide. i had this problem when i converted the "millfalc", all texture UV's of the cockpit was spread all over several repetitions, but it was rather simple to center them all proper after i was aware of this. this is important if you use later a "combine textures", because if they are spread over several repetitions blender will stupidly export all used texture repetitions (somehow logical), what would be in such a case far to much, and one repetition of a texture is enough on a combined texture of course, if it's not needed because it's a large cylinder i.e. with many repetitions of the used texture over the full length.

disclaimer,

of course "ripping" from any else as your own work will be a copyright infringement of some sort, but as long as you only "extract" your own models (or mine) i see no problems, i just said this because it's possible to extract models from any game that runs with opengl. of course if you do such because of personal interest and never spread or publish such it's neither a problem.

what can i do with this?

well, ok, guess a situation like the creation of a bezier quad, i can't make such so easy with blender i.e. like i can do it with the LMR, there are some bezier and other similar meshes possible in blender. but the cool thing is, with the LMR i don't have to create each LOD. though it's somehow far easier to create a bezier surface, lathe, cylinder, whatever LOD dependant object with the LMR and snapshot the result for a further use in blender. it will have right proportions and proper LOD easy as 1,2,3.

it's a "brainfucker" from my pov, to make a cool bezier surface in blender it needs many steps, we do it with the LMR "im handumdrehen" (like nothing).

both "systems" have their drawbacks and advantages, no question for me.

if i had to decide i would keep the LMR, first it's simply a cool tool, second you can't learn better geometry as with this, you know marcel that i always have seen this also. i tell you my teacher would be more then surprised to see that "gernötli" creates models in this way, the "gernötli" who used to look out of the classrooms window in geometry, the "gernötli" who cheated once so bad in a test that it has made him laugh out loud because he wondered how we (i) did that. i guess i will contact him if possible before the LMR vanishes, i have to show him this once. i know that most have a problem with geometry and maths because it's such a dry matter, physics and chemistry was far more interesting because something is going on, something is moving. while we boys always stated in case for geometry, "what can we do with that?" well create models for a computergame... besides such wasn't to see for us in 1983, i guess if he would have been aware of this he would have used it, he was a really clever teacher, unfortunately also his ideas didn't fittet to how teaching was done here. he moved from the "big city" to the country and always contradicted to what they expected here (sports and 1x1). it goes up to that they suspected he was gay, well i would say now "and if?..." "what's the problem?". but i know how fast they judge here and i know also that some think this of me to where i live now. "if you don't spit you must be gay..." or something similar, you know what i mean.

but how can one be gay who simply likes good proportioned "streamlined" big women? and they can't be tall enough, i tell you...

---

and "harald" is really cool, he can also tell you why klingons have this "plate" on the forehead, or why "acceleration absorbers" can't work.

very simple reasons which are comprehensible for everyone.

like to hear?

ok, to the "plate on the forehead", i'm laughing so much now it's hard to type, it's because we have a complete wrong imagination of what happens if we reach relativistic speeds, if that would be possible at all. it's what even our game shows completely wrong. to stay with this we show stars (dots) that move past us from the centre we front to, this is what we naturally assume it would look like. objects will pass us like i.e. trees on the roadside. but fact is that if we would reach relativistic speeds all light that reaches us and this includes all electromagnetical waves not only light will be bundled on one point, this will be us or our ship. now this would... ah i guess you know what it would, so klingons have this plate on the forhead because of the bundled energy else it would melt a big hole in their head, ok it's a joke, but true is that it would be bundled.

to the "acceleration absorbers", a less funny example, it's very simple "no reaction is possible before a action", even the fastest "positronic" (whatever shit) would need a little time to react, even if the difference between action and reaction is only the billionst fraction of a second, you would be smashed to "chili sauce" if you would accelerate with forces like we usually assume in sci-fi and even in our game.

this just to get a broad smile on everybodies face.

that's why i like "hyperdrives" and no FTL, a hyperdrive doesn't need a set speed at all, you "transist" from one point to the other, impossible perhaps, but at least you evade all problems of FTL or "CloseToLightspeed".

and fortunately our drives in pioneer are hyperdrives, what i only miss is a "transition shock", not especially needed to be felt by the pilot, but should theoretically at least influence the environment, conclusion would be, it's forbidden to engage a hyperdrive in a close orbit of a inhabited planet. imagine only the acoustical "bang" a desintegration of 100tons would leave in a atmosphere, add some whatever energy to this and it's obvious, no? the "view" we would have would be for sure complete different also to passing stars, i can't tell, we have no hyperdrives in reality, most reasonable for me would be a "nothing", but of course this is boring, so what about weird colors? or whatever strange effect, squeeze, bend morphing of the view(s)? i really assume you wouldn't see your ship, because how could you in a n-dimensional space. everything would be different for sure and it would be different also to all what we can imagine. a random effect? oh yes i would like. something that makes you feel "i'm in hyperspace" at least no passing stars at all, because this reflects the "wrong" idea of FTL and neither fits to hyperdrives.

it could be also something funny, why not? except you haven't enough fantasy for something funny.

also i'm not really happy with interrupting a once engaged hyperdrive, i mean yes should be possible, BUT only with a destruction of the drive, you see i like limitations. further it must be from my pov a rather stupid pilot who isn't sure where he wants to go. ok shit happens, but to interrupt a engaged hyperdrive should cost something, best to me it would be the drive itself. imagine all systems are engaged to proceed the jump, do you really think such is reversible? i didn't think so. perhaps a misjump as result would be fine also again followed by a broken hyperdrive, if you are lucky you will find yourself in a inhabited system and this could be from my pov elsewhere in the galaxy, if not, have fun with waiting for santa claus. i have been often in such situations and i won't say i disliked it, anyway if you don't play "for real" you reload the game and even if you play "for real", ok i'm dead or as good as dead, result -> start a new career. of course playing "for real" needs a little character, you reload if it's caused by the game, i mean if something went wrong and the game has terminated or stucks, but you won't reload if you get damaged or get killed. it's like when you play backgammon or chess against the computer, it's up to you to play fair against yourself.

again hyperspace, i guess it would be cool that it would by possible to do a jump even if i selected the system i'm in, it would simply lead to a misjump, this will happen very often, i swear, you forgot to select a target system and... "oh no, what has happened? ahh, im a idiot!" also such would give the "interrupt" possibility a real value, in this case no misjump would have to happen for interrupting a engaged jump, but a broken drive for sure. only refuelling isn't punishing enough for such a stupidity like to select the wrong taget or worse to select the "home" system as target.

simple stupid traps, yes, but makes it far more interesting. back to the idea of entering hyperspace to early, it shouldn't be blocked, no, you will notice it only when entering the target system, when you get fined for doing so (probably? means less policed systems won't fine you and you get away lucky, or you will be fined only when returning to the system you proceeded the "unlawful jump", both i see as a good treat) .

i swear you will over and over step in the same traps, at least i can remember such well... it's easy to forget a little thing, very easy.

i (we?) played FE2/FFE for i don't know how long, but still... "melted to a piece of junk, because i forgot to fit a atmospheric shielding, do i have lost my mind?"

for the acceleration "issues" (chili sauce), we have it in our hands to lower this to acceptable forces, which i'm still up to and my newest revisions are lowered in ship specs and i probably will lower them more, depends also a bit on what you like.

but let's say a "fully ladden ship" and it should have then not more as prob. 12g, that's still to much but i feel this as "acceptable".

of course with "empty" tanks and empty cargo bays you will reach far higher values, but some extension to reality is allowed i guess.

you will hardly fly a "eagle" empty, this also because i lowered all cargo capacity of the ships by the amount of fuel they can carrie.

reason for me is mainly to get to this FE2 feel, you never will be satisfied with your ship, something is always missing, either cargo capacity, performance or amount of fuel. finally fighters will be fighters and very fast, but with each tonne extra capacity you will loose in general a lot of performance, at least this is my idea i have of this.

besides i feel with clever tuned specs we need no "sophisticated" autopilot (varying fuel use) or other nitpicking stuff.

the ship designer has it in his hands to make it more realistic already, i think.

what i miss a bit is the distance to the central star, resp. especially for binary or tertiary( + ) systems, in good old FE2, everybody knows you have entry points very far from the central stars, depending on the layout of the system, very very far.

i guess some dislike this, i liked it always.

also a bit a reason why i won't like "sophisticated" autopilots, some might think 3200 we have... (whatever)

and of course such is to solve in the game i can see this to,

but imo boring, i don't have to think as a pilot then, i like when i'm forced to think, so i like how it is and if fuel consumption is set high enough it will propbably get you in big trouble with the autopilot. so "turn off" the "shitty autopilot" and approach manually. e.g. like it was needed to reach "alpha centauri" (sorry, toliman) in FE2, lay in course with the help of the autopilot, switch to "computer speed controlled", set the proper speed, as soon as you reach that, turn of engines and wait til you are close enough to re-engage the autopilot, in this manner i was able to do most missions to "alpha centauri" in the given timespan withoout running out of fuel.

besides i see NO problem to "cheat" this for the AI ships. but i like to force the pilot to use the autopilot as less as possible.

and i dislike the idea of a autopilot who solves all for you, that's probably realistic, but no "damned..." you will ever hear of the player.

if the game engine does all thinking for me... i wouldn't have to play i guess.

a "one button" control would strongly dislike, of course one could argument now that it's up to me if i don't like to fit a autopilot, well, we humans are lazy and if we have a help we will use it... so this is no argument for me.

but if this help is unreliable or weak, i will have to find a better way of using the help i have.

pehaps that's no "fun" to code or no challange i don't know, but i can't also do all i want, i told that once before, i would like perhaps a pilot going to the toilet, such could prob. be made, but exceeds totally what is wanted and needed for pioneer, even when it's a big challange for me to animate this.

it's called "space simulation" but not to forget "game"....

i would like to see in future a thread here called,

"TYPICAL PIONEER QUOTES"

and i guess you know how they should sound like.

certainly not: "alles in butter"

more like: "damned, again i missed this occasion" or "why do i always forget to...", "what have i done? where am i?"

this i like to hear of pioneers pilots!

---

@robn

and i write "i like", "i feel" and "i think" because this is the way you should post in a forum, no?

i know that i left this trail far to often in the past, this shouldn't happen anymore.

Holy f**k gernot that's a lot of text!

Ok I'll pick up on the thing I noticed first and treat your Green text as points to address:

• [background=rgb(33, 47, 58)]"i can't see any reason to invest many hours in something that can be done in a few minutes", that's true. [/background]- this is one of the main reasons LMR must go away, it's a hardest route to getting things done.
• [background=rgb(33, 47, 58)]back to the animated vector groups, i feel that they will lack of performance as if i transform a wavefront .obj dynamically, am i wrong? [/background]- this I can't answer, I don't know how the new model system works but I don't think it will mean animating things using vector frame style animation.

This specific case and others like it you'd usually do using vertex skinning animation. That name makes it sound like it's only for "skin" on people but it's not.

What it does is define a set of "bones" to form a "skeleton" inside of an object, you then animate the "skeleton" by moving the bones around. Each of the bones affects a group of vertices in the model - this is the skin part. You can have vertices affected by multiple bones with varying weights given to the contribution of each bone so that you get smooth blending along the skin.

Take a look at this, it shows a wagging tail on a girl... for... some reason, anyway the white section down the middle are the "bones" the purple green is the "skin", you animate the bones and voila the skin moves and bends correctly.

[media]

You make the animation, it moves the tail from beginning to end and back to the start, you make this a looping animation or we handle that in code. We play it over and over and you've got a wagging tail on a Turner ship.

This sounds like it would be more expensive to run but actually it's dirt cheap to process because you only really move the "bones" so there's very little animation data, the "skin" is usually handled in a vertex shader on the GPU, although you can also do it on the CPU with some simple vector mathematics which can be nicely optimised by the compiler or by hand. So it's actually a lot more data being processed, but it's much much much cheaper to do than animating individual vertices or rebuilding the mesh repeatedly like LMR.

Yeah I liked the cockpit views but I understand why they've been removed for now. They just weren't the right way of doing them.

If you want to discuss them more then you can message me privately but the _only_ way to do a good cockpit view like those in the forthcoming "Star Citizen" requires a lot of extra work. At least an entire separate, hi-res, cockpit LOD. Then some rendering code work plus all of the input support AND that's before you get to having any of those little screens actually showing anything.

I'd really really really like to see it in the game eventually so don't discount it as a possibility but it means a lot more work on every single model just to get it looking "ok".

I must admit I didn't read all of your post, too much text!!! So If there's something important that I missed just say... briefly <img src="' http://spacesimcentral.com/forum/public/style_emoticons//wink3.gi f"' class='bbc_emoticon' alt=';)' /> what it is.

Andy

@potsmoke, Thanks for all the info. It will take me a while to digest it, like the lua stuff you taught me.

@fluffy, That's how I thought it worked. It's kind of sexy for some reason, even though the animator doesn't seem to know where her coccyx is supposed to be. <img src="' http://spacesimcentral.com/forum/public/style_emoticons//wink3.gi f"' class='bbc_emoticon' alt=';)' />

no, buddy you cached the most important points, others of my ideas i can repeat on a other place/thread.

sorry for calling it "animated vector groups" it's a bit ancient, skinning animations, yes is the more recent term for it. (a max term?) such exists since long in one or the other way, i think i.e. of NFS4's drivers resp. their arms, "bones" as helpers was unknown then, but the principles are the same i guess, you animate a specified vector group (resp. in case for ancient NFS4 you assigned only the vectors which was to animate), with the help of bones and if i'm not wrong the exported mesh won't contain the bones, no? at least this was (is still?) the case for directx meshes. not that i couldn't imagine this to, exporting the bones and leave the rest to the engine, sounds even more advanced)

in blender i guess it's still called a "vector group animation", but i guess it wont matter how i call it at all.

the process how to work it out at all is "common" to me (even if i would say i'm absolutely unexperienced), but good anyway you posted the vid, because it might not be present to all what it is. also if one can handle a rigged animation proper it should be no problem to setup bones and assign vector groups to them.

to the cockpit view, yes i expected this answer already, or was longing for it?

i know that it was "clumsy" what i did and it didn't turned out to be really good (in disrespect of how bad they looked probably, it was more a experiment), even when i restricted the cockpit for some models (and this are really only a few) already to LOD4, respectively restricted the ship to the lower LOD's, but i got to the problem that it will differ bound to resolution, so it might have worked in the highest resolution but not in a lower one.

but yeah, let's see what the future brings and i'm already afraid of the challange it will bring to me <img src="' http://spacesimcentral.com/forum/public/style_emoticons//wink3.gi f"' class='bbc_emoticon' alt=';)' />

@marcel

and those feet (feet? no fins) let me fear... but at least they help her not to fall on her coccyx, even in stormy weather.

coccyx is a "cruel" word, how do you spell this? without to cough (i know "Steissbein" isn't really better, it sounds... well dont' say a word now i know it sounds almost like "Scheissbein", which comes very close to what it is) couldn they have thought of something more pleasing for the ear? "coccyx" "steissbein" argh.

some critics allowed?

i watched the vid 3 times i guess

not a really smooth animation, i feel, could that be made better? or is such the result only in a preview?

looks a bit staggered to me.

---

to certain things i posted before about the LOD, yes i know it could be made even with the CAD, but i have to say this i understand far less as to animate with bones, phew.

also i was interested to get educated in CAD, but well i didn't earn in a year enough money to subscribe to such a training,

i knew this also in advance and i always thought "why does all have to be limited to rich poeple? max cost's already more as i earn in one year with legal work, the school costs far more as i can ever earn with my job in one year, this means all the money i can earn, let's say 40'000 bucks easy for all (school and software), unfortunately you will need max to get educated, you will have to pay this price, tell me who can such?"

makes me sometimes a little angry to see such is limited only to the rich.

i guess i return to chinese ink, that's far better in price...

you can say now usually it's payed by a company, oh well now i know why everything gets more and more expensive...

besides i'm no company, even worse i'm unemployed.

its a sort of treadmill i feel. something i wanted to get rid of, no more treadmill, no more earning more and more just to pay the price...

or should i really go and murder for money? at least i would earn enough money then...

selling weapons? drugs? do something else criminal?

OUT OF THE QUESTION!

no "gernötli" don't get angry,

we still have blender, i guess it would be worth to get some literature to get deeper knowledge.

at least that won't cost as much, perhaps a education as a desktop publisher, even if the market is tight, i guess i have qualities.

for this my inherited little spare capital will be enough, probably.

but first bu-jitsu, i have to "animate" my own bones first a little bit...

and HEY!

THANKS i noticed a few days ago i'm on the list, that makes me very proud, let's hope not to proud...

"mit stolz geschwellter brust" lol.

can such be translated?

nah, google translation makes no sense, but "with a swollen chest" comes close to it i guess.

let's hope i don't stumble if i didn't see my feet anymore.

[background=rgb(33, 47, 58)]"with a swollen chest"[/background] - yeah that still makes sense in English <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_e_smile.gi f"' class='bbc_emoticon' alt=':)' />

About the animation in that video not looking smooth. It's partly a control issues, they've not set it up very well. There's a lot of thing like this that I'm actively avoiding until the new model system goes in and we see what we've got to work with. I know that everyone has been working really hard on it, they're not cutting corners they're working on something that will give artists the flexibility they've been after.

Last night I took Robn's work on exporting LMR meshes to ".obj" format and updated it to the latest master so that's being looked into too.

i guess i was just a little insecure.

really your answers has helped already a lot, i see now better where we are heading to, that's fine.

i think (didn't i said such before?), it's not easy when you only have the machine to communicate.

it would be completely different if a contact would be possible, at least for me.

i neither like to telephone instead of meeting because of this, one could say i fear the telephone.

I still don't quite get the way models are handled in Pioneer, though to be fair that's mostly because I haven't had time to really dig into it. I must say, though, that even if it IS hard to get right and scares off a few modellers that scripted and/or procedural models should stay. I won't claim that the current system is the best way, but the way assets are produced for most games will not necessarily be the right way for a game like Pioneer. Most games these days have significantly less play area than an entire Galaxy... you can't compare a game the size of Skyrim or Just Cause 2 with a game the size of Pioneer! Thousands of light years with trillions of potential NPCs compared to a few dozen square kilometers? Not even in the same class!! If anything I'd like to see an attempt at procedural ships, I don't think the problem is as large as everyone thinks... at least if you focus on "realistic" spacecraft designs. It would be more work up front but it would potentially pay off BIG time in the long run since you could focus on refining ship parts and styles instead individual ship designs.

Of course, I'm biased strongly towards a "Rocketpunk" aesthetic and gameplay style, with elements of Transhuman and Anime styles... frankly most of the designs in my head would be (in pincipal) easier to script than to model... Nanotech Fractal Squid Mecha FTW!!! <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_razz.gi f"' class='bbc_emoticon' alt=':-P' />

Umm... just to be clear it's the mecha part that's anime inspired.. not the squid!! <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_razz.gi f"' class='bbc_emoticon' alt=':P' />

@NeuralKernel

You need to look into stuff like this http://www.blendernation.com/2011/07/01/the-shipwright/ and http://ship.shapewright.com/?name=BlenderNation

You can generate a lot of random rubbish looking ships with a few good, sometimes even great, looking ones popping out by random chance.

it's doing it well consistently, then texturing them, giving them detail, making them feel real. Hard, exceptionally hard to do that part.

Instead of clumping random "chunks" together I reckon you'd have better luck using shape grammars and L-systems to provide the rules to make the basic structure, then several stages of geometry refinement using whatever process you like to generate the texturing. Then you'd have to generate levels of detail for the geometry using some automatic refinement algorithm... or maybe that could be done along the way as the shape grammar/L-system is run?

I dunno, I've thought about it a lot but I've never had the sheer volume of time it would take to do it.

Until there is an established aesthetic it's hard to say what kind of asset production method would work best. For "mainstream" science fiction style ships (stylized aircraft in space) procedural generation would indeed be very difficult to get working consistently well but it's far less of a problem for Rocketpunk or Fractal style ships. Rocketpunk spacecraft would all be Spheroids, cylinders and discs... the same way pretty much every car looks like a big box with smaller boxes fore and aft and most planes look like tubes with wings... there are really only a couple overall shapes that work well for near-future combat spacecraft. In any case... I don't know enough about the guts of the engine to really make any bold claims about what I think should be done <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_e_smile.gi f"' class='bbc_emoticon' alt=':)' />

Ah now we're talking two different things if it's production vs aesthetic <img src="' http://spacesimcentral.com/forum/public/style_emoticons//wink3.gi f"' class='bbc_emoticon' alt=';)' />

In my view, and this means it's not necessarily shared by all team members let alone the other coders, is that by the time an asset reaches the engine it should be in a compressed binary final format.

All the creation, editing, modification etc should have already been done in Blender/3DsMax/Maya/Photoshop/GIMP/MSPaint/etc. It can have animations, decals, texture painting etc done in game and things like sub models for turrets/radar/etc with their own animations can all be attached too it but the actual base mesh, its basic textures and animations should all be done as that single chunk of data.

So take your modelling thread, you'd do all those in Blender and export them with their textures and animations. You can use whatever tool you like to create funky / impractical / pop-punk / rocketpunk / steampunk ships and their insane wibbly-wobbly animations (this is criticism btw it's just hyperbole as example! <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_e_smile.gi f"' class='bbc_emoticon' alt=':)' />) but then you bake them down so that those animations are set in stone, that mesh is final and the hardpoint locations for sub-models are chosen.

Lots of reasons for doing this, hard learnt from previous projects but here's a few:

• Rendering performance - LOD generation, mesh welding & optimisation, texture merging all take time, are better done by hand and shouldn't be optional,
• Loading times - Pioneer takes an age to load a lot of this is spent simply parsing the model and getting them into the games internally rendered format, the cache helps but it's still pointless and a waste of time, I should have written a tool that does this which people could run to produce a "game ship" formatted object but... I've been lazy and we have the new-model system on the way now.
• ...damn run out of time gotta dash now!

As for the games aesthetic not being set... well mostly no, but you don't have to worry about what it's not-set-as, you can define that aesthetic yourself, you might have to modify it to meet other people part way if you hope to get people on board with your preferred style but don't view it as a hinderance instead view it as your chance to guide the aesthetic! <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_e_smile.gi f"' class='bbc_emoticon' alt=':)' />

Right, really must run!

Andy

Don't worry about scaring me off with criticism, I go by intent and not specific wording when I decide whether to get upset <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_e_smile.gi f"' class='bbc_emoticon' alt=':)' />

What I meant with my comments about aesthetics was that a focus on fuction over form would be easier to generate procedurally than a more stylized interpretation of Interstellar technology. A plain spheroid with a bit of procedural detail for RCS thrusters, sensor pods, weapon ports, etc... would be a lot easier to get consistently good looking than trying to randomly spew out stylized spaceplanes.

In any event, I'm not going anywhere, even if Pioneer never turns into exactly the game I've always wanted it's a damn cool project I'm lkeen to help on in any way I can... seems I'm not being particularly helpful at this point, though. Hopefully I'm not actually hindering anyone's efforts with my ramblings! <img src="' http://spacesimcentral.com/forum/public/style_emoticons//icon_e_smile.gi f"' class='bbc_emoticon' alt=':)' />

When it comes to aesthetics, don't worry too hard about it. It's a big galaxy, and will have several factions, each of which might well have a very different look. As long as the ship looks physically plausible (we probably won't accept a tall ship modified for space travel without a very good explanation) we will be able to find a place for the ship somewhere.

In this respect, I'm a fan of diversity. It also makes the game better - working hard to prove yourself to a hostile faction because they have the ships that you like the best, or being able to judge the allegiance of a ship because it looks like their style, even if it's a ship you have not seen before. This is all good.

(edit: Big, not bog.)