Kin of the Stars

Please sign in to post.
Kin of the Stars

A community for the fans of Crest of the Stars, Abh culture, anime, technology, science fiction, video games, and friendly conduct.


    Shiolle's Sekai Sim

    Share
    avatar
    Almael
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2453
    Imperial Credits : 13365
    Registration date : 2009-03-18

    Shiolle's Sekai Sim

    Post by Almael on 9/21/2009, 9:32 am

    Almael wrote:

    I
    just did a quick analysis on this Simulator project to determine
    problems. For ease I used an UML tool to visualize it. As a result it
    also represents the program design.


    I color coded according to importance and possible problems. Red being highest.
    As
    can be seen the clustering around the ship class during simulation and
    the editors suggest high complexity, and hence problems.
    Discounting
    the graphic engine, I estimate the code size to be 200-300kB with
    decent documentation. This equals about at least a months work for a
    good and active freetime programmer.

    I have been looking around more, so I will post some further insights.
    First, this is not the first time our Seikai fan community has attempted this. In the past we had some programming option, but none were as convenient as or modern as we have today eg. C#, Silverlight, XNA, Angel-engine. We also have more advanced examples available today which weren't there even just two years ago..

    Now to my plan porpossal based on my findings.

    1. Graphic engine.
    This the core of the sim program, hence, it has to be made first and offer all the things needed. There is no sense to create anything else before this has been done.
    This will also decide which programming language will be used.
    (I can deal with most languages, but this is limited by availability of the compiler.)

    2.Scripting
    This refers to the program being able to read configuration and commands from a file. Normally, most program do this to a certain amount. Here I refer to the powerful ability such that scriting as nearly as much as programming.
    The advantage of this is that thinkering can be done within the scriptfile and not the program code. This saves compiling and fasten test and allows changes to be made easier.
    The disadvantage is that the program has to offer all it's abilities through the script, a complex interpreter, and a virtual machine to run the script.

    3. Ship and Mission creator
    This is the easiest part but it sets the requirement for the more important and complex part of the simulation program. So decision on ship/system abilities can have great impact.

    4. Physic engine
    This is also an easy part but it depends on decision made in #3.

    5. Entities
    This is the most complex and the part that is important to users.
    The decision on interdependenciees of graphic object and ship model will set the condition and limitation of how things will be done.
    On this decision ship management and AI programming will be heavily influenced.
    (I have done some and still do some AI sripting/modding, so I could help out.)

    6. AI
    Here a decision has to be made on what type of AI programming.
    I can assure that the kind of if-then is not going to work out well.
    My best bet is on state programming.
    Here a decision should be made on having event handling or not.
    Well, having events would be great and offers realtime, but it's quit a mess to design and program them.


    I think I pointed out all the important things and decisions, and hope this will help advance the project by avoiding problems or bad decision before hand..

    Shiolle
    4th class crewman
    4th class crewman

    Number of posts : 7
    Imperial Credits : 6176
    Registration date : 2009-08-06

    Re: Shiolle's Sekai Sim

    Post by Shiolle on 9/21/2009, 11:44 am

    Sorry for cross-posting this. I think if there's a separate thread now for Sekai Sim, all it's discussion should move here.Early this week (Sunday-Tuesday) it seems I am rather busy. Though some further work on the sim was done there is no new version yet.
    As for formation, either the ships form similar group, hence, similar group mass, or move all in one big bubble.
    So, ships will need to form bubbles of identical mass or they will be forced to move at different speed. Won’t there be any ways to counter this by having lighter bubbles perform small side moves?
    Our discussions on this concluded that forming a group will save fuel as only one ship has to maintain the bubble.
    What about the situation when hostile ships enter the bubble like assault ships entering cruiser’s bubble? In anime it seemed like cruiser already had the bubble large enough to host 10 assault ships and let them maneuver. It doesn’t seem very practical. If he could vary it’s bubble size, why won’t it shrink the bubble so that ASes won’t be able to maneuver or simply force them out?
    I think this topic cannot be easily solved, so the development should go on before any final decision is made.
    IMHO starting with payload and working up to heavier systems ease things up and lessen difference to needed requirement.
    I think there we will talk about it when I expand interface section in the next edition of the document.
    The difference for plane space and normal space is not that much, so I'm not concerned about trouble when trying to expand it.
    I think the normal space implementation will take place at the same time as in-bubble space implementation because they’re basically the same.
    For now, I think the only important question is: do they use the same fuel for propulsion, reactor and bubble generator? How much different types of fuel do we need to include?
    There is because the ship is still in 3D within the bubble. The problem is the translation from one universe into the other of which we have not been informed.
    The problem is those two universes have different dimensions. So if a ship performs bubble separation at the ‘zenith’ of a bubble, where would it appear in plane space?
    Infos are from proton-antiproton engine for insterstellar 'ship' dated 2003.
    If they are the same for Sekai universe, then I wonder how often do I need to recalculate ship mass during the simulation. I wasn’t intending on doing it on a regular basics. (It’s not a question.)
    Still, the weapon energy has to be reduced, if you have separate values, you will deduce from Shp part, and have some left over or deficit while
    you have hull damage whit extra Shp or deficit left. It's just not realistic. In reality you have energy which can translate from one form to the other.
    So be it. I’ll see what I can do.
    You said we needed different calculation since it seem different weapons work differently on different ship classes.
    I just meant that even without HPs a blow that would be catastrophic to light ships shouldn’t have that much impact on heavier setups.

    You may not realize it but anime tech is always well thought out since the last 25+ years.
    I gather my will to refrain from derailing this into another discussion about loud explosions in space and flight model of spacecrafts that looks like a combination of modern jet flight and cool ninja moves.
    To be clear does this mean the ship will receive full damage or is damage part of the function curve as in your mine receiving damage?
    The latter. That curve didn’t factor misses. I’ll elaborate a bit in the next edition of the document.
    Have you read my edits (at the end) from last post?
    Of course. The fact that I don’t comment some of your points means I have not much to say about it except for “Ok, I agree. Thank you. I’ll try to factor that in the next edit of design documents.” If you want to hear something on specific points please ask. These posts are large as they are and I want focus on critical topics.
    About ship conditions system. Undoubtedly, the system that accounts systems physical layout within the ship and selects damaged systems randomly is much more accurate than any ship ‘states’ or tables of offlining priority. 
    But it is also true that the purpose of the whole system is to translate weapon fire effects on ships into some conceivable form of representation. In SFB games where the number of ships in battles didn’t usually exceed 10, designers could afford display plenty of properties of individual ships. In our case where we should model the entire fleets of ships we can’t afford such luxury. If a user can’t tell how good his fleet is faring at a glance, the system needs a rework. So while the complex systems with different layers of HPs can be used for internal calculations, we need another system to translate it to some kind of representation.
    The system I proposed already offer conceivable output for large groups of ships. Basically if we have just 5 states we can make a colored line similar to that of “windows 7”’s resource monitor memory representation stripe.
    While we can implement more complex system for “inner level” calculations and means to translate it’s results into a conceivable form, the question is whether the difference in simulation results really different enough between two systems to go through all the difficulties to implement it.
    Docs you asked for.
    Thank you very much. It’s a great info. I can’t yet say what I could use within the sim, but at least crew listings are of great help already.
    I just did a quick analysis on this Simulator project to determine problems. For ease I used an UML tool to visualize it. As a result it also represents the program design.
    Here I intended to write lots comments on the diagram, but I’ll list the next few things to work on.
    1. Detailed interface description: which buttons do what and how a user can interact with it as a whole. Same for different modules.
    2. Technical aspects: files specifications for missions, ships and data tables. First tests on interface implementation (a few experiments), etc.
    3. UML models for the whole projects.
    4. And now the programming starts.
    avatar
    Almael
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2453
    Imperial Credits : 13365
    Registration date : 2009-03-18

    Re: Shiolle's Sekai Sim

    Post by Almael on 9/21/2009, 3:01 pm

    Shiolle wrote:Sorry for cross-posting this.
    That's alright. We are past the interesting theory and formulas and now on the programming, so it's all fine.


    So, ships will need to form bubbles of identical mass or they will be forced to move at different speed.
    Won’t there be any ways to counter this by having lighter bubbles perform small side moves?
    -Same ship type usually are in one unit, they are likely to form a unit bubble.
    -As we have seen the assault ships lining up in a line-formation we can be sure that faster ships/bubble can slow down.
    So it's not a problem anymore.
    (I'm drafting AI now)
    So we just tell the faster ships/bubble to slow down or adjust speed.


    What about the situation when hostile ships enter the bubble like assault ships entering cruiser’s bubble?
    If he could vary it’s bubble size, why won’t it shrink the bubble so that ASes won’t be able to maneuver or simply force them out?
    Good point.
    IMHO the bubble field is dependent on life time of spatial particle and power or size of the ship. The generator creates a field such that a safe zone covers the whole ship, while the field gradient gets less with distance.
    At a certain distance fluctuation between the dimensions disperse the field and create the bubble shell. Anywhere inside the distortion is too low to disrupt 3D, so 3D object can still exist.
    The assault ships still use their generators, but their mass is insignificant to the big ship. If I remember right the bubble does grow in size when the basroil entered it.

    Here the rule of 'if you can't hurt or kill the enemy there is no point fighting' applies, so shutting out an enemy who can survive with his own bubble makes no sense. You have to stand your man and fight! Very Happy


    For now, I think the only important question is: do they use the same fuel for propulsion, reactor and bubble generator? \
    How much different types of fuel do we need to include?
    Yes. I do not see any sense in having differrent fuel. So all from same antimatter and water fuel tanks.
    We only need antimatter and water. The problem is the ratio of consumption. In theory for propulsion the ratio of water is higher than for the reactor, because thrust is provided by expelled water.
    reactor: 1:4
    propulsion 1:16-20


    So if a ship performs bubble separation at the ‘zenith’ of a bubble, where would it appear in plane space?
    It doesn't matter.
    Only the direction matters. The altitude is random. (I think it's said in the Cots novel.)
    Maybe there is a max. possible difference to the plane of the ship inside.
    Just make it simple on a plane. Wink


    If they are the same for Sekai universe, then I wonder how often do I need to recalculate ship mass during the simulation.
    I wasn’t intending on doing it on a regular basics. (It’s not a question.)
    Simple.
    -Ships can only be faster as they get lighter.
    -If the ships don't fly at max possible speed but at a set speed then there won't ever be a problem.
    Speed will be decided by the fleet formation (slow down) and everytime the ship changes action. (AI)
    So all we need is max speed under max mass. and set it as permanent constant parameter.
    Still we need a normal speed parameter for calculation and as option for certain mission. Wink


    I just meant that even without HPs a blow that would be catastrophic to light ships shouldn’t have that much impact on heavier setups.
    Well, we wil see after we get to test the systems. (Maybe my layer model will solve this. bigger ships=thicker skin)


    I gather my will to refrain from derailing this into another discussion about loud explosions in space and flight model of spacecrafts that looks like a combination of modern jet flight and cool ninja moves.
    That's artistry. lol
    Luke, otherwise you would not hear R2D2 telling you how he feels while reparing your fighter. LOL


    I’ll try to factor that in the next edit of design documents.” If you want to hear something on specific points please ask.
    Just want to make sure you get all the imput, since I thought it was interesting. I often find new infos after I post so I do edits regularly.


    So while the complex systems with different layers of HPs can be used for internal calculations,
    we need another system to translate it to some kind of representation.
    the question is whether the difference in simulation results really different enough between two systems to go through all the difficulties
    to implement it.
    Fine. Once the sim is done, given that it's well designed, it shouldn't such a problem to make the change.
    Well, I can assure you that a slight difference eg how shield damage is applied you can get very different results.
    (I wrote once a performance simulator for ship designs for a certain game. Little differences make great differences because of many weapons and many shots, so they sum up big.)


    Thank you very much. It’s a great info. I can’t yet say what I could use within the sim, but at least crew listings are of great help already.
    You're welcome. The crew list is genuine. Note that not all positions are manned by 3 people in 3 shifts, hence total crew number is less than 3x.
    The scientific docs are now for sale and not free anymore. Sad


    Here I intended to write lots comments on the diagram, but I’ll list the next few things to work on.
    It's just a simple one I made on the fly to see complexity and problem. Now I would model the simulation part differently.

    Shiolle
    4th class crewman
    4th class crewman

    Number of posts : 7
    Imperial Credits : 6176
    Registration date : 2009-08-06

    Re: Shiolle's Sekai Sim

    Post by Shiolle on 9/21/2009, 3:56 pm

    For this post I will focus on main technical aspects outlined by Almael earlier.

    As I’m writing the interface part of the design document, I can’t help but to reevaluate some of my options.

    Programming language

    I would really love to go with C# because of wide range of tools already built in and convenience of some of its concepts. Also after I graduated from Uni I never wrought a line in C++, although I’m confident I can get back on it pretty fast. I know very well that C++ is de-facto standard in game industry, but choosing it would force me to either do more work on programming or seek 3rd party solutions.

    Graphics engine

    All applications except mission editor and the sim itself won’t need any special graphics and can be done with either MFC, win forms or WPF. All the advanced graphics lie in drawing the plane space.

    For that we will need:

    1. Sprites that can be colored in code (like units in starcraft have parts of different color depending on plyer).

    2. Lines and ellipses to represent bubbles borders.

    3. Advanced splines to draw outlines of groups. (Optional).

    4. Free for non-commercial use.

    What we do not need:

    1. Lighting.

    2. Sprite animation.

    3. Intricate content pipeline (there won’t be really a lot of graphic content anyway).

    4. 3d graphics.

    I have never worked for gamedev so far and until recently had little experience in graphics programming. I read DirectX API documentation back in the day (and concluded I couldn’t really hope to put together a game by myself), and lots of theory on the subject as well as outdated course on it back in the Uni. I also did some projects involving 3d graphics but my own part in them only consisted of scripting. I was always interested in this field but never had actual experience. Recently though, my work required me to use a 3d graphics engine and a bit of WPF vector graphics, so I got into it a little.

    I don’t intend to learn how to do low-level graphics programming on this project. From my experience such initiatives never got finished or at least drag on for years. So I won’t be writing my own 2d engine using DirectX or OpenGL.



    At first I intended to use WPF built-in capabilities in vector graphics (they are even hardware accelerated). But as I write further it becomes more and more evident they won’t be sufficient. Microsoft never intended WPF vector graphics to be used for extensive game (or game-like) graphics. The greatest drawback is that WPF uses event driven model. It doesn’t work very well when you need to update content as fast as you can. It’s good for UI though, and I intended to make a good use of it.

    XNA is also looks promising. I didn’t figure out if it has the capabilities for vector graphics I need.

    About other 2d engines, a few of them support C# (or other managed languages), and I don’t want to write wrapper, cause it’s also a lot of work and unsteady ground for me. I will see if I can find a fitting one.



    Scripting

    The easiest way to implement serious scripting is to integrate python or other similar language using 3rd party libraries. They must be for non-commercial use too. It is in itself a lot of work, and I don’t plan to do it until the simulator reaches version 1.0 for sure.

    So for now I think editing constants via config files should be enough.



    Ship and Mission creator

    For that we need basic classes of the universe, but we don’t need their dynamic interaction which makes it the easiest part of the work.



    Physic engine, Entities

    I think I will talk more about this when I start writing about project’s technical details (after interface description).



    AI

    The AI should be minimal for version one. This is a preliminary list of task it should be able to handle.

    1. Assign targets to individual weapons using priorities for different weapon types.

    2. Mine AI

    3. Which system to offline based on behavior settings.

    4. Small-scale maneuvering. (like heading for nearby enemy bubble or separation on damage if behavior permits).



    In the meanwhile can I ask you about fleet structure in sekai universe? For interface part I need to 3 level of command. Lower is reserved for individual bubbles, but to name the upper levels I need to know how sekai fleet of different sizes are composed.
    avatar
    Almael
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2453
    Imperial Credits : 13365
    Registration date : 2009-03-18

    Re: Shiolle's Sekai Sim

    Post by Almael on 9/21/2009, 4:47 pm

    I agree on C# because C# express is free and I have experienced on it just more than a year ago, so my memory of it is better than any other language.

    I don't care much about the graphic engine, since most free ones can do it nowaday. I dunno why you need vector abilities that much, though. I only see the use of lines for them. Anything else is just math.

    Well, the final AI design will look somewhat different, we will discuss this later.

    I really don't think this simulator will be slow as long as the ship number doesn't reach 5000+ ships. But logging into file will certainly slow it down. (which I do wish to see, though.)

    ----------------------------------------
    Fleet Command structure
    I will just draw it all out so we can make a choice.
    1. single ship
    2. squadron: 6 ships
    3. group/strike force: 1 command sqr.,
    (3 mission type specific[mine,assault etc.] sqr, 1 mine sqr., 1 supply sqr, 1 guard sqr)
    or
    (5 recon sqr, 1 supply sqr)
    4. Fleet: 1 command group
    (3 mission type specific gr, 1 mine gr., 1 supply gr, 1 guard gr, 1 recon gr)
    5. Contingency/Corps

    Well, anyway I think we only need some kind of group modelling. From there it should be possible to have applied recursively.
    For example, a group entity which gives commands down to another group entity or ship entity.

    Here are some battle logs I dig up from a certain game + my text based performance simulator.

    --------------------
    Edit:
    There seem to be not much about Vector graphic engine on the net. IMHO it's just a reorganized 2D or 3D engine with different stored data format, + interpreter etc.
    SVG NET
    SVG Paint
    SVG library (C#), C++

    Edit:
    After all 3D engines are vector engines, just arranged the use of their functions for SVG.
    Anyway, I redrew the design for simulation important part (in pseudo UML)...

    As you should see, there are a bunch of problem areas and descisions to make.
    1. 3DObject, should the entities inherit it or are assigned an object?
    2. Ideally group entity and shipentity should be compatible so we can process them with the same overall algorithm. There would be no worries about command level, if we can make group entity work recursively.
    (random though, if a ship can be assigned as leader, then there is no need for group entity.
    Only group bubble management will be handle by physic engine or speed-calc of the commandship)
    3. Events: I hope you got enough experience here, because mine is lacking
    all communications are going to be based on events


    So this is roughly how simulation objects and AI could be.
    I separated powermanagement from AI to make things faster.
    damagecontrol&powermanagement are only used when receiving damage and damage to energy system.
    The ship classes will be big! and the AI will have many many simple classes.
    These are not really difficult to do, but are very complex. If done professionally..I see lots of possibilities.
    avatar
    lymhfeubdach
    Rear Flyer
    Rear Flyer

    Number of posts : 116
    Imperial Credits : 6825
    Registration date : 2009-03-24

    Re: Shiolle's Sekai Sim

    Post by lymhfeubdach on 9/26/2009, 1:01 am

    Hmmm..... if this about build up the mod and all for it. I like how you lay out your game plan get up running.
    avatar
    Almael
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2453
    Imperial Credits : 13365
    Registration date : 2009-03-18

    Re: Shiolle's Sekai Sim

    Post by Almael on 9/26/2009, 6:20 am

    No, this is not about the mod. For the mod we need specific tools and more time.
    This is a simple simulator, that allows us to design or build ships and then see how they fare.
    It's not quite a game, because the AI will be simple.

    A similar simulator can be found here: http://fcomm.sourceforge.net/
    This one allows you to build ship AI only. fcommx is an older version but more advanced in code and 3.7 faster, at max speed it's about 256x faster than normal.

    ---------------------------------------------
    Edit:
    Further on design
    Normally, game speed depends on graphic, number of objects and AI.
    While testing fcomm, I was unable to find bottlenecks. This program was too well designed for speed.

    Anyway, I'm struggling with the event design.
    It just seemed too impractical if you have to subscribe all the possible events for each object.
    Windows surely has a more practical solution.
    In any case, a message class is needed to send the information needed, since we can only send a certain amount of information.
    It would need to have the information about the sender, receiver, the type of message, the various possible message content.

    Another speed problem would be distance calculation. For every cycle this has to be done at least once for every object and counter object.
    This would lead to a N^N-1 problem. This could be reduced by group calculation only. And when needed later for single objects.
    I'm all ears if there is a better solution.

    AI design:
    fcomm allowed up to 12 simple AI commands to be process per cycle per object. And it was able to run it through in a second at max speed no mater how many objects were created.
    I think it just skipped the graphic part while tying it to real time, hence, gaining speed for the actual core simulation.
    We should be able to do the same while keeping all the features we have been talking about.
    Another major speed problem depends on the complexity of the AI.
    If the AI is too flexible designed then it will lose speed due to class jumps or calls or initiation.
    If it's monolithic within the entity then it needs lots of memory.
    In any case, we only need the most simpliest needed abilities.
    If have been thinking about AI singleton classes, but this is no good for many objects. They would become bottlenecks.
    If ever singletons might be usefull for the smallest group of objects.
    In any case, necessary information should be passed to reduce calls. But still all information access functions should still be available just in case.

    Now, to how it might work.
    Damage and power management are only called when needed, and can take priorities into consideration.
    Some event might be needed in order to cange ship condition status and strategies. (will see below)
    (group and single ship decisions may be similar)
    Single ship:
    So the first thing the AI will do is look at the condition, the situation and decide which strategy to take:
    Patrol or Combat or maybe Retreat.
    Then it should decide on the action:
    just move or move(closing on/retreat) & attack/defend
    That's it.

    In detail, the ship could report enemy contact found or being attacked or call help while in the situation decision making.
    A lead ship could issue commands in this stage. Commands will just override decision with information.

    Too bad there aren't any AI sims that would allow to test an AI design.

    Maybe a simple fuzzy logic function can be used for weapon priority based on range.
    Or it could be use for threat priority, also based on range. Range is important here as flexible parameter because of changing ranges for plane space and normal space.
    avatar
    JGZinv
    Admin
    Admin

    Number of posts : 667
    Imperial Credits : 7900
    Registration date : 2009-03-17
    Age : 31
    Location : 9468 Parallel World Dr. Artifact USA Earth

    Re: Shiolle's Sekai Sim

    Post by JGZinv on 10/6/2009, 9:42 am

    Perhaps I missed something, and granted I only skimmed the posts (wall o text you have there)
    but instead of taking years to write a new game engine... why not mod something that already exists?

    I mean, I run a space combat fighter mod... so I've got some experience in that area....
    Building something from scratch... we'd have never got off the ground. We're 3 years in, although
    we've had several restarts... and we've just been doing light mods to the engine we chose.

    If you're more interested is seeing what can be accomplished, let me know.


    _________________
    True power comes not from strength, but from the soul and imagination
    avatar
    Almael
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2453
    Imperial Credits : 13365
    Registration date : 2009-03-18

    Re: Shiolle's Sekai Sim

    Post by Almael on 10/6/2009, 1:16 pm

    Well, obviously Shiolle is interested in learning vector graphic skill, while I'm more interested design and theory.
    I'm sure he can create a simple Win UI in a day. I could, too, but I'm out of shape and need some read-in again. There is no doubt this project needs an existing graphic engine, no sense wasting time on doing this one ourselves.
    The rest is simply coming up with a working design to be put into code. In theory the most part are no more than using files and if-then, the complex part lies in events and messaging.
    Everything else is some simple math methode to create ship parameters.

    Normal modding isn't much different, though, here you also use some graphic engine.
    Then you write some script which is admitedly simpler because you don't have to design it.
    Making models etc. is similar to the programming part, easier maybe, but not necessary faster to be done.

    I'm always interested, but mostly I just don't have the tools or time. Prefereable something quick.
    I wouldn't mind discussing solutions or problems etc. Just shoot away. Unlike hacking&cracking japanese games nothing else is scary or overwehlming.
    avatar
    JGZinv
    Admin
    Admin

    Number of posts : 667
    Imperial Credits : 7900
    Registration date : 2009-03-17
    Age : 31
    Location : 9468 Parallel World Dr. Artifact USA Earth

    Re: Shiolle's Sekai Sim

    Post by JGZinv on 10/6/2009, 1:36 pm

    Well I've always thought that a BOTS mod would be possible in the Freespace 2 engine.
    It's quite well situated to handle the design of the ships, weapons, graphics, and so on.

    There is one cavat, and that is that largely capital ships cannot be piloted. There is one guy who
    made everything except two items work... and he requested what he needed but people basically brushed him off. If you have some experience in coding or scripting, then you might well be able to bridge the last gap.

    I would thumb through here, see what you've got to work with.
    http://www.hard-light.net/wiki/index.php/Portal:Modding

    Just look at the options you have for setting the attributes of ships:
    http://www.hard-light.net/wiki/index.php/Ships.tbl

    There's already a community behind the engine, there's at least one guy at the FS2 forums that used to be at Abh Nation, and liked it.. You already have programs to edit the files with. I wrote the last guide on how to get models in game to Freespace 2 from 3ds max.

    What you need then is textured models mainly, and you can get vary far as is, if you're doing a basroil simulator. Controlling caps, would as I said, likely need some work on the sections the one guy was missing... but you'd be a lot closer to something usable rather than starting from scratch.


    _________________
    True power comes not from strength, but from the soul and imagination
    avatar
    Almael
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2453
    Imperial Credits : 13365
    Registration date : 2009-03-18

    Re: Shiolle's Sekai Sim

    Post by Almael on 10/6/2009, 2:36 pm

    I will take a look. If it's not tinkering with the engine itself I should find out what can be done or not. If FS2 can indeed allow control cap ship, I'm all for a mod.

    Currently, i don't have any working 3D program installed or running. So I can't do even if I want to.


    Edit:
    If I understand it right fighters and capship are handled differently eg scoring etc.
    I don't see anything in the tables description hinting speticifctly at capships, so I believe
    the problem lies probably in the core code of either the engine or other important code files.
    To find the problem a trace would be necessary from the table script code and the declaration of ships and the use of the AI.
    The easiest test would be changing out code for fighters and capship to see what happens....
    I would need to see the actual codes to have a better analysis.
    avatar
    JGZinv
    Admin
    Admin

    Number of posts : 667
    Imperial Credits : 7900
    Registration date : 2009-03-17
    Age : 31
    Location : 9468 Parallel World Dr. Artifact USA Earth

    Re: Shiolle's Sekai Sim

    Post by JGZinv on 10/6/2009, 4:30 pm

    Cap and fighter tables are largely handled the same, but of course there are differences.

    The original thread was here: http://www.hard-light.net/forums/index.php/topic,60834.0/topicseen.html

    You could probably start with this guy for engine questions, and he can direct you from there to a better person. I already know he's not extremely deep in code, but he handles some areas. http://www.hard-light.net/forums/index.php?action=profile;u=340

    I suppose the best code forum would be here, http://www.hard-light.net/forums/index.php?board=50.0 although modding is handled on another section.


    ...as to 3D do you have a problem with Maya or Blender?


    _________________
    True power comes not from strength, but from the soul and imagination
    avatar
    Almael
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2453
    Imperial Credits : 13365
    Registration date : 2009-03-18

    Re: Shiolle's Sekai Sim

    Post by Almael on 10/6/2009, 5:54 pm

    I haven't (couldn't) used Maya, otherwise our HW2 mode would be done to some degree.
    I haven't used blender either, but I have used Truespace v2-4.

    I have some suspicion that in order to get capships control able there are more things to do then just put it on the list of player controlable ships. It seems only fighters are special while everything is an object. There are probably differences in processing orders, i dunno if capship even handles cockpits. This may even be a necessary part for controlable.....
    It would help if I have sets of cap and fighter files, as well as all program tables.
    The problem is probably all over the place.
    Well, I will look into the forum.

    Eidt: I just poseted my proposals. I was surprised by the simple problem needed fixing.
    Looks like the game had been badly designed. Instead of creating coomonality, and expand by creating abilities useable for all, it's based on ever growing definitions, but without a based to process them properly.
    avatar
    Almael
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2453
    Imperial Credits : 13365
    Registration date : 2009-03-18

    Re: Shiolle's Sekai Sim

    Post by Almael on 2/26/2010, 11:29 am

    I might get someone from another simulator project on Star Trek, TOS era, to cooperate with us. He wants to make a newer version of an existing old version. It's a simulator game kind of like RPG or so.

    So how is the current status of the project?
    avatar
    mitsuki lover
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2322
    Imperial Credits : 10703
    Registration date : 2009-04-10

    Re: Shiolle's Sekai Sim

    Post by mitsuki lover on 3/1/2010, 3:00 pm

    I noticed the ads for the new Star Trek Sim,but that one seems to be for Next Gen. through Voyager as it deals with more of the "modern era" Trek.Wonder if anyone has come up with an Enterprise Sim yet.
    Anyhows,my question is,when this new Seikai is finished will it allow you to play whichever side of the conflict you want.For example if you so desire could you play as a UMK soldier as well as an Abh?
    avatar
    Almael
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2453
    Imperial Credits : 13365
    Registration date : 2009-03-18

    Re: Shiolle's Sekai Sim

    Post by Almael on 3/2/2010, 9:15 am

    ST Online: well, in contrast to reviews given by them it's pretty much a normal Mmorpg. I had that feeling when reading theirs (since soe details where left out)and after digging.
    I was especially suspicious when they offered lifetime gamership worth two years of subscription. Also it would only be offered before the game actually gets officially started. IMHO they knew it may not last that long....


    First of this simulator is a two part 'game':
    one is for playing around creating ships
    then simulating ship battles

    I don't see any problem to play different sides since we could just use different flags for the sides anew in every simulator session. It's not a campaign driven game that we need to keep consistancy, but it wouldn't hurt, though..
    Don't expect too much regarding mission objectives. It's likely to just use normal command options like 'do patrol', 'search', 'follow waypoints'. Afterall the aim is to see how ship design fares. Expect to see insights you might not have thought of before and gain better understanding of (Seikai) space warfare..
    avatar
    mitsuki lover
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2322
    Imperial Credits : 10703
    Registration date : 2009-04-10

    Re: Shiolle's Sekai Sim

    Post by mitsuki lover on 3/3/2010, 2:05 pm

    Well that's good.I think if you allow variety of some type in the game it will make it better.
    avatar
    Almael
    Imperial Admiral
    Imperial Admiral

    Number of posts : 2453
    Imperial Credits : 13365
    Registration date : 2009-03-18

    Re: Shiolle's Sekai Sim

    Post by Almael on 3/7/2010, 2:41 pm

    Well, mission creation (in general) isn't as much a problem as shooter level designs. Everything bases on the command options available, so it's kind of advantageous for space sims to concentrate on controls. But it's still too early to say how far we can go.

    Sponsored content

    Re: Shiolle's Sekai Sim

    Post by Sponsored content


      Current date/time is 12/15/2017, 10:16 am