Page 2 of 3 FirstFirst 123 LastLast
Results 51 to 100 of 114

How do fighters and corvettes get fixed?

  1. Homeworld Senior Member  #51
    Hiding out on LM-27 Norsehound's Avatar
    Join Date
    Jul 2003
    Location
    Space.
    Also some of us like the squadron solution.

  2. #52
    drunken1
    Guest
    Just confirmed through testing that torps are 1 shotting heavy vettes almost every time unless some munitions miss
    1 torp vs 3 heavy vettes (about equal RU)= torp kills them in 4 shots, suffers ~20% damage
    3 torps vs 9 vettes: 0 losses for torps
    5 torps vs 19 vettes: 1 torp lost

    lol vettes

  3. #53
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    @Langy: if you have experience working with systems like that fair enough, but i know rule 1 of games design is "it's never as simple as you think". so i tend to get pretty bothered by people asserting it'll be simple if they don;t indicaite actual experiance with somthing.

    @Siber: Sure you can, DD's do it all the time but based on my experience of watching them if something moving like that turns the vector of movement turns with them. You can't have the ship rotate independently of the vector of movement, and it does it by completely ignoring the direction the engine is pointing in, whereas if you wanted to change vector in HW1 i believe you had to swing the engine around first in the right direction).

    @Drunken: After some work i figured out whats going on with the cluster torpedoes. My fault for missing a flag in a file previous time around. Don't think it's used in any other weapon though so i just overlooked it. As far as the rest goes. Bear in mind you probably can't produce equal numbers of ships under most circumstances.

    Would also be nice to see numbers of Assault frigs, (HW1 races and Vaygr), vs HW2 corvette's and fighters. Also would me interested in HW1 attack bombers vs flak frigates. They have more health so not sure how that effects the balance.

    And yeah squads are the best way to balance this mess on the SC end my far, that and switch corvette's tot he same flyby attack style as inty's.
    I don't know what i'm talking about, ignore me.

    Thousands of years ago, Egyptians worshipped what would become our ordinary housecat. The cats have never forgotten this.

  4. #54
    Member FriendlyFire's Avatar
    Join Date
    Jul 2006
    Location
    On the imaginary axis.
    @Carl: Why do you assume that Langy has less experience than you do, exactly? Did you make a game without telling us? Did you program a Newtonian physics engine or something? I'm genuinely curious now, though it feels strange to come out and claim that others don't provide what their experience is when you have not done that either.

    I agree with Langy that it's most likely nowhere near as complicated as you make it up to be. Doing all the graphics upgrades they've done is guaranteed to be far more complicated.

  5. #55
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    @Carl: Why do you assume that Langy has less experience than you do, exactly? Did you make a game without telling us? Did you program a Newtonian physics engine or something? I'm genuinely curious now, though it feels strange to come out and claim that others don't provide what their experience is when you have not done that either.
    I never claimed i had more experience. What i do know from modding a lot of games is that even at a very low level some things can be amazingly involved. And games developers always when interviewed about it like point out how complicated stuff is.

    So unless i'm dealing with someone i know has games design experience i'll assume they're no more knowledgeable than me and we should bow to the voice of experience that is actual interviews with actual developers. I'll allows some fudge on that if the game allready comes close to he required feature on the mechanics end, but that's a very subjective judgment to make.

  6. General Discussions Senior Member The Studio Senior Member Boardwars Senior Member  #56
    Player Hater Langy's Avatar
    Join Date
    Sep 2001
    Location
    CA
    Modding and actual source-code alterations are generally completely separate experiences in terms of difficulty to do something; something that's almost impossible via modding can be extremely easy if you have access to (and understand/familiarity with) the source code.

    (Also, to be clear on my experience: I don't do games programming aside from modding, but my day job is essentially as a programmer for high-fidelity simulation systems, which can be more or less involved than games in many ways)

    But yeah, my point was that while changing the movement system to include inertia or vectored thrust might be complicated, it's probably not as complicated as a lot of other parts of the game engine. It would require multiple changes to different parts of the game - primarily AI and the simulation progression engine itself, but the main difficulty would probably be in the AI changes required rather than anything else, and then all the testing to ensure that those changes don't break other things. Plus, they'd then need to rebalance everything anyways. So no, not easy, but it's probably not what I'd call a major change to the engine.

  7. #57
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    Modding and actual source-code alterations are generally completely separate experiences in terms of difficulty to do something; something that's almost impossible via modding can be extremely easy if you have access to (and understand/familiarity with) the source code.
    To try and give a simplified response, (for space and anti-derail reasons) to this point for you:

    1. you'll see a lot of fans of a mod asking "why can;t you do X", and even if X is eminently possible and simple overall they seem to regularly underestimate how just how simple it is.

    2. A lot of former modders and mod teams turned full developers tend to go on about just how much more complicated things are than they assumed as modders.

    TDLR i'm playing glass half empty .

  8. Homeworld Senior Member  #58
    I think people are assuming game engines give two shits about physics. A game engine has no concept of physics until you actually write a physics engine, you don't need to observe the laws of physics when designing movement. It's actually harder to make realistic physics engines than Newtonian ones, as you don't need to worry about pesky stuff like gravity. You just need to tweak the acceleration and max speed values (unless you go fully Newtonian) to give it a realistic sense of movement, otherwise you end up with Inertialess motion (which is exactly what the Beast MS was using in Cataclysm, a complete lack of physics).

    HW1's system would probably be nice to have in HW2, as overall it's a more fun way to portray spaceship fights and it probably wouldnt break the game like the HW2 system in HW1 has. HW2 went for more dogfighting style movements, which are all well and good for atmospheric stuff but they don't make a huge amount of sense in space. Just watch Babylon 5

  9. #59
    Member FriendlyFire's Avatar
    Join Date
    Jul 2006
    Location
    On the imaginary axis.
    @Carl: Flip side, I, as a modder, can tell you for a fact that if I had access to the source code, many many many things would go from nightmarishly complicated to trivial. There's a reason why I still dream that one day they'll release it as a bone to the community.

    Your point (2) is also not correct in the current circumstances: you're talking about a mod team going full dev team, but we're talking about actual, experienced devs working on a pre-existing engine. Not starting from scratch. I'm sorry to say but an awful lot of mod teams don't know much about games development because modding is far removed from actual game development (the art side is fairly close, but "code" in modding terms is often laughably simplistic), so of course they'll find it hard, but for a developer the perspective is rather different.

    @Mnementh: As I said, I wouldn't be surprised if HW2 still had a Newtonian engine, because there are many advantages to using one, even if you then go and fudge values. It lets you use SI units, for one thing, which can unify the dynamics of the game under a set of well-understood laws. Even games you'd never think are Newtonian actually do use Newtonian physics, they just add things like drag in space (which makes no sense, but gives the WW2 dogfighting feel).

  10. #60
    GentleM3N OutriderVS9's Avatar
    Join Date
    Feb 2005
    Location
    A Mad Cat
    It would be the easiest and most effective fix though. I am sure given enough time you could balance individual strike craft against squadrons, but the amount of time and effort required would be completely out of proportion of what anyone could reasonably put in.
    But we don't really know this for sure now do we? Squadrons would be one solution, but like FF said above making the HW1 strike craft into squadrons would take away something that in unique for the HW1 races. There are other ways to balance I hope.

  11. Homeworld Senior Member  #61
    Hiding out on LM-27 Norsehound's Avatar
    Join Date
    Jul 2003
    Location
    Space.
    Unique, in dying unnaturally easy to things that shouldn't be able to kill them?

  12. #62
    GentleM3N OutriderVS9's Avatar
    Join Date
    Feb 2005
    Location
    A Mad Cat
    Unique in the sense that those formations were (and still are) mechanics allocated to only the Kushan and the Taiidan. This even goes back to the whole idea of remastering a title, plus the fact that Gearbox mentioned that they did not want to change too much in the transition from classic to remastered. So if the classic HW1 formations are forsaken for the squadrons of HW2, all of a sudden you have 4 races which are becoming more and more like one another. Why take away the flavor when squadrons aren't necessarily the only workable solution to the problem?

  13. #63
    Chappy FooF's Avatar
    Join Date
    Mar 2006
    Location
    On the brink of insanity
    There are other solutions to "the problem" but "the problem" is something Gearbox inherited, rather than created. When HW2 was envisioned, its design goals were fundamentally different than HW1's so when Gearbox decided to use a modified HW2 engine, they were naturally trying to shoehorn HW1 into an engine that was not made for it: it was a hack. One could argue that Gearbox took a path of least resistance by not playing with the RNG numbers or AI behavior enough to more closely emulate HW1 mechanics (and I do believe more could have been done from seeing HW2 mods that seem to get pretty close) but in the end, it was never going to be perfect unless they created a new engine.

    I guess what I'm saying is that I'll cut Gearbox a little slack for inheriting an engine that simply was not designed to do HW1 stuff. Maybe with the MP Beta still in testing and with enough feedback, Gearbox will make some changes but if not (or if not to our satisfaction), mods will likely get pretty close.

    From my perspective, squadrons are the easiest fix for the time being. There's little downside except for nostalgia purposes. I, too, don't want the factions to become too homogenized but unless formations can be improved, RNG tables fixed, and attrition rates fixed for the HW1 factions, their strike craft will always be gimped. The band-aid right now is squadrons.
    Semi-Retired


  14. #64
    GentleM3N OutriderVS9's Avatar
    Join Date
    Feb 2005
    Location
    A Mad Cat
    From my perspective, squadrons are the easiest fix for the time being. There's little downside except for nostalgia purposes. I, too, don't want the factions to become too homogenized but unless formations can be improved, RNG tables fixed, and attrition rates fixed for the HW1 factions, their strike craft will always be gimped. The band-aid right now is squadrons.
    For the time being there is no easy fix without the mod tools!

  15. #65
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    A bunch of useful stuff i dug out of the files for modding and balance consideration:

    Notes:


    Vaygr Assault frigate actually seems unchanged but need to check missile data. Expect it to be ok vs fighters, (1 individual fighter of any faction every 2.5 seconds average), but not great vs corvette's, (37.5 seconds per corvette for HW2 unupgraded. 52.5 to 60 seconds per HW1 corvette, and 67.5 for fully upgraded HW2 vette's). Anti-frigate DPS is 114.


    Kushan Assault frigs are flat out better than Vaygr atm except vs fighters. (4 seconds per HW 2 fighter, 8 for HW1 one's). Corvette killing power is much better however, ( 14 secodns for unupgraded HW2, 22-26 for HW1 corvette's, and 28 for fully upped HW2 vettes). Anti-frigate DPS is 173DPS


    Khushan DD guns have same accuracy values as HW2 DD's, but hit harder per shot so they kill HW1 and fully upgraded HW2 corvette's in 1 instead of 2 hits. Fighters are a 1 hit for both, however the Khushan has 1/8th the number of guns so eh. Anti-Frigate/capital DPS is 859 just as it says in game. 300 from the kinetics and the rest from the ions. Ions are 12.5 second between shots and do just over 3.3K a shot.


    Khushan HC's have similar anti-fighter story to the DD, though they have 6 guns. the guns hit even harder but since no strikecraft can survive the DD's that hardly matters for this. That said this puts it much closer to a higgy BC's 8 guns than the DD comparison. Ion's do 10 per beam, so 20K per turret on a 12.75 second cycle. so DPS is spot on the 4800 and change listed in game. his IS less than a higgy BC or Vaygr BC.


    HW1 Bomber cycle their bomb launchers faster and thus get more shots per pass plus deal more damage per hit, about a 1/3 increase over HW2 bomber's (note individual vs individual, not individual vs squadron).


    HW1 Fighters have more health and 2 guns. A HW2 fighter has marginally better accuracy plus a LOT more damage per shot. to the tune of about 2/3. This means that whilst a HW2 inty will kill a HW2 fighter in 1 hit and a HW1 fighter in 2, the HW1 fighter needs 2 and 3 fortunately the twin gun setup means it will score twice as many hits on average. However there's one disadvantage. a HW2 inty can score a kill on the first shot of each burst more often than a HW1 fighter, and if either do this the 3'rd and final shot will be targeting a new opponent, which can radically increase kill rates in squadron vs squadron engagements. Arc of fire is smaller though.


    As i predicted way earlier HW2 fighters can out turn HW1 fighters, not by as much as i thought but it's not insignificant, ()about a 10% drop off), and they're about 4% slower to boot. That said turn rates seem to be down across the board once the new multipliers in the code are taken into account. the defender dos match H2 ships in turn for the curious.


    Defender guns themselves are very weak, i think they're weaker per shot than scout guns, RoF is slightly lower than inty guns too, accuracy isn't great vs fighter's, (though it's awesome vs corvettes), that said they do have 3 guns their RoF is still good and the arc of fire is huge so combined with their good turning radius and low speed they can maintain fire on anything they're shooting at better than normal inty's of any faction.


    Missile Destroyers do enough damage per shot to in theroy 1 shot any fighter or any unupgraded HW2 corvette and 2 shot any other form of corvette, however as we've all seen the misses couldn't hit strikcraft if their lives depended on it. Anti-frigate DPS is worse than an ion frigate, anti-capital is better but still only about 2/3 that of a normal destroyer. Will investigate in a moment. gonna run trhough a few more ships first.


    Light corvette guns deal a fair amount of damage per shot, enough to OHK any fighter and take out any corvette in 4 to 8 sots, (116 per shot, so divide target health by 100 and round up to a whole number). Acurracy and rate of fire are poor though.


    Multi-gun corvette guns deal just enough damage to 1 shot HW2 fighters, 2 shots vs HW1, and only 23 damage a shot vs corvettes. Accuracy is comparable to the light corvette. Higgy Gunships are similar weapons but fire about 60% the number of rounds per minute, however the way they're setup they'll probably deal less overkill, 9i.e firing 5 shots at the same ship and the first kill it before the others get there), because of how they fire.


    Heavy corvette guns deal about 2/3 the anti-corvette damage and need 2 instead of 1 shot vs hw1 fighters but otherwise match the light corvette gun. Generally better but not doubled up. the burst fire shoots what amounts to a half radius 80% damage flak frigate shot. So you need roughly 5 heavy corvette's to equal 2 full flak frigates, single salvo obviously.


    Drone frigate drones are frightening, only 70% the rate of fire of an interceptor. But they have much better damage output vs non-fighters and are slightly higher damage per shot, they still need 2 per fighter, but now any fighter goes down in 2, not just HW2 fighters. Oh yeah and they launch quite a few of these duh.


    Checked missile destroyer missiles, and yep, their rotation rate is crap, they'll never keep up with a fighter in the real world, also their lifetime value is way too short, they'll despawn before they reach the maximum range they can be fired to. For reference, turning radius of missile is 160m, higgy int is 180m and higgy gunship is 105m. the anti-corvette submunitions of a higgaran cluster torpedo are 100m.


    Repost from the modding thread fo those who don;t check there btw.

  16. General Discussions Senior Member  #66
    Senior Member roflmao's Avatar
    Join Date
    Feb 2007
    Location
    Austin
    I guess I wasn't crazy thinking something was iffy with the fighters corvette flailing wildly not accomplishing much of anything.

    Looking at the HW1 non remastered to remastered comparison on how strike groups operate .. wow.

  17. #67
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    Well really the biggest issue for the Corvette's vs fighters is that the turrets have limited arcs and fighters will attack them from behind where they generally can't shoot.

    Also a couple of other things i've dug up since then, posted it as tips in the multiplayer area, but i'll add em to the balance pile.

    1. Vaygr Lance fighter are better than assault craft vs HW1 fighters.

    2. The Flak frigate is utterly broken vs HW1 fighters, the torp frigate is better. This is down to the lack of squadron's tough the exact amount of extra health the HW1 fighters get doesn't help as it up's the number of salvos required by a third, (4 vs 3), compared to 1 health less.

  18. #68
    Um, just two cents after skimming 2 pages.
    Most all of this is quite moddable, Strike craft behavior is quite easy to modify though tuning it to look and work good is another thing. You can have them break sooner, or evade with more dynamic movements.
    Also there are ways to turn off RNG and just use the miss behavior (ie if it misses it will be tracked and see if it hits something else). Also all of the damage, targeting and hit % can be changed per class group and you can make your own groups so you can make them as big or small as you want. Also turret tracking rate and ship maneuverability really can play havoc and can really mess up a ships usability.
    Also on top of that, you can simulate a more Newtonian movement for all craft by playing with top speed and acceleration rate.
    So there is ALLOT can be done with this engine!
    Also a pure Newtonian and physics braced system while desirable, is a complete pain to tweak and is VERY unforgiving. Also for allot of people it will look wrong (even more so when people want nice WWII naval battles in space. ;P ) and be hard to control. Again I do not mind but most people do.

  19. #69
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    Tareog:

    1. Break distance controls when they swing around for another run, i did notice there seem to be a flag that may? control when they break formation, but it isn't certain and i'm pretty sure mos of the fighters will end up never shooting because the target will never be in firing arc, see point 3.

    2. You can turn projectile damage flag on sure. But:

    A) a miss disables velocity prediction so it will never hit anything that's moving.

    B) the miss dispersion angle has to be a whole number, that means even at minimum dispersion at fighter ranges the shots spread across a circle 50m across, which is a hit rate of 0.5% vs a fighter class craft. It gets worse as the range goes up, you can reliably hit most frigates and all capital ships with such an angle, (cap ship KE guns use it to produce more random firing patterns allready), but you can't use it to tune accuracy the way the old HW1 weaponry could.

    3. You can make fighters take an age to turn slow down, speed up, e.t.c. But it's nothing like Newtonian. they just start acting like racing cars on ice, skidding through space. What people mean when they say newtonian is full newtonian, i.e. new BSG style fighter movement, which is what HW1 had, (weather it was full Newtonian in the engine or not), that's something HW2 cannot do, if your pointing in a specific direction you will not continue to coast in another direction. You CAN set up a slide function, but that fixes the direction of movement relative to the direction of the target, you can't have the ship moving in one specific direction whilst rotating and firing at different targets around it at different angles as it coasts.

  20. #70
    So I decrypted HW1ships.big and took a look at some of the strike craft files, and they are all just like HW2 ships, and you can set squadron sizes via line 12(for kus_interceptor), NewShipType.SquadronSize=#. This is set to 1 by default for the kushan interceptor, but can easily be modified. I did so, packed it up and tested it, sure enough I got a squadron of 3 interceptors for one paid unit.

    i.imgur dot com/k6UqvRF.jpg

    It needs some adjusting so they don't shoot each other but at least it can be done, and only one value changed.

  21. #71
    So I so not know what you are talking about, I just played a HW1 game and tried very hard to get the interceptors and scouts to do anything like you are saying (fully believing it was me just misremembering)
    I am so glad I did.
    All I got was normal strafing runs, that fly pass the target (big or small) then a flip over and repeat. I did miss remember now this was totally I forgot how painfully basic it was. The ships often bounced off the target. wiggled lamely when flying straight and did allot of 180 flips. almost but not quite flipping on a dime. Not once did they do any side strafing B5 style, or thrust in any vector other then ahead. The ships did skid a little but only barely, and mostly on the flip over. The acceleration was just too high and any "physics" did not have a chance to happen.

    "But it's nothing like Newtonian. they just start acting like racing cars on ice, skidding through space"
    First that IS a good description of how a Newtonian ship would look like if the main thrusters where on the entire time and made a heading change. (I know I have made a Newtonian engine that simulates the movement that more then once over a good 30 years for various projects. I even added gravity but the real deal is almost inconceivable for most and had to flatten the curve - removed the squared and replaced it with just 2x. )

    So ??? I do not know what to say. . . HW1 does not do what you are saying, at least to any greater degree then I got to work in HW2 10 years ago. you may want to try it again yourself, I honestly forgot how clunky it was.
    Also I had the weapons working just fine, I had allot more groups and very selective behavior for each, fire rate was much higher also which fixes some problems (no one parameter will do all of this it requires allot of tweaking). I had frigates swarming a larger ship moving around trying to not get hit. Bombers firing from a long distance breaking and getting out before passing the target. And interceptors doing all sorts of maneuvers including the flip over, loops and sliding 90s, and so on.
    Again, I am not finding a real problem here, and now with some live Devs we may get some nagging problems fixed that was stopping some workarounds that was a pain.




    Also ships

  22. General Discussions Senior Member The Studio Senior Member Boardwars Senior Member  #72
    Player Hater Langy's Avatar
    Join Date
    Sep 2001
    Location
    CA
    I think the main behavior that'll be difficult to add into HW2 will be the formation behavior; currently ships leave their formations as soon as combat begins, which is not at all what HW1 formations were like.

  23. #73
    so the HW1 kushan strike craft are basically copies of the HW2 counterparts, just without the squadron buddies, and a little cheaper. Putting them in squads of 3 definitely seems to improve their effectiveness, but without campaign testing it is hard to say. I'm going to bump up the price a bit and do a playthrough of the campaign and see if they get shredded instantly in combat like my first playthrough.

  24. #74
    Member FriendlyFire's Avatar
    Join Date
    Jul 2006
    Location
    On the imaginary axis.
    @Langy: They do seem to stay in formation if grouped though, so the behavior is there somewhat, it's just not applied everywhere it should be. Now, there are other issues with group formations, but hopefully those can be fixed at the same time.

  25. #75
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    Teraog: Try actually using formations in HW1, you saw it best off top of my head with an X. The fighters out on the ends of the arm, (at least as long as they where considered in rnage which the AI is bad at working out for larger formations, i only got about a qater of a 55 scout formation to fire), would rotate to fire down into the center of the formation as enemy strikecraft passed through. That said given the opportunity, yes HW1 strikcraft do prefer to behave HW2 style, formations are the only thing that makes them behave any different because it forces them to hold position which can mean they have to turn different amounts to each other to shoot. (And yes i just tested it, have screenshots but you can barely see anything in them because of the low res graphics so i'll only add them if you really want them).

    Also yes Newtonian does look like your suggesting if the engines on all the time, but as noted HW1 strikcraft can turn the engine off and just coast. Something HW2 doesn't allow, you can simulate the effect for know firing angle's, but not for random angles.

    Also again i'm not saying i think the HW1 weapons physics can't be properly done in HW2. I'm telling you it's flat out impossible.

    Sure you can do it in frigate vs frigate and cap vs cap an frig vs cap fights, it's allready used as i said to make hit locations more random. But because velocity prediction is turned off on a miss, (forced missing being only way to simulate HW1), if the target is high velocity like a strike craft and isn't moving almost directly towards or away from you it will fire at a point way behind the target. Likewise the bare minimum miss angle, (above zero anyway), is 1 degree. It's too large to create something that looks anything like HW1's system for strikecraft vs strikecraft. You can create a system that is effectively using physics in HW2, but aside from no accounting for target movement you end up with hit rates worse than the minimum accuracy value vs fighters. Corvette's are a bit better because they're larger, but they'r also still nowhere near HW1 in hit rates or apperance.

  26. #76
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    Okay taken a look at the multi-gun corvettes hod files, can't get the models to load but the harpoints show just fine. The Taidanni are screwed basically, the Kushan are flawed but fixable.

    First the taidanni turrets are all restricted to a 20-30 cone of fire centered forward, they can't even fire to their sides, they've got worse fire arcs than defenders. No matter what attack style you use that's going to create serious balance issues for them. It's fixable with a very minor model edit to give the side guns a full fire arr, but you'll never fix the bow turret's. Then again the Tai multi-gun is a crap design as such things go, so no surprises there.

    The kushan have much broader fire arcs, but suffer majorly from the two front turret weapons files being re-used for the side turret's. What this means is that whilst each turret can train to the side it's moounted on and to within 25 degree's of directly aft, it can't point all the way aft, it also can't fire full to the side and can depress enough to shoot through the hull. However with the way the turrets are mounted at an angle to the horizontal it's quite possible to give them 180 fire arcs if you reduce the depression angle. Elevation is a bit limited but fixable. Overall with an altered attack pattern they should do fine.

    Kushan Defender guns are arranged a little oddly, but otherwise share identical fire arcs. Not sure on the way their hod files are setup. Something very odd is happening there. Oddly better arcs above below than to the sides.

    Taidanni Defenders have a more normal and sensible layout of guns, and better firing arcs.

    Edit: Seems giving the Side guns on the multi-gun better arcs is gonna be harder than i thought, they really altered the model. looks better but makes for a harder job.

    Example of max traverse, (elevation is worse though), here.
    Last edited by Carl; 3rd Mar 15 at 2:01 PM.

  27. #77
    Member
    Join Date
    Jun 2007
    Location
    Canada
    Perhaps the Taidanni multi-gun could have higher damage output to compensate for lesser time being on target? Say make them have a shotgun approach, low rate of fire, long down time between targeting opportunities, but a high probability to destroy the target?

  28. #78
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    Wouldn't really help much. It's the defender problem again. Gunships and kushan multi-guns can plow in firing continuously, which means more volleys out targets engaged, and they spread the fire which produces less overkill.

    That's the thing you see, 6 multi-gun corvette guns firing have an average of just shy of a 100% hit rate per volley and you need enough time between volleys to let the first salvo arrive before the next is fired or you'll end up wasting the extra shots. Sure you can do a lot with that, but maneuvering as Taidanni corvette's do drops the time on target percentage to just a 20% rate. You can't push them to 5 times the hit potential.

    Of course since kushan's circle they're not using half their guns 90% of the time so it's not a big deal ATM, the Taidanni will under-perform, but not my as much. But compared to gunships or fixed kushan models they're not even in the same room.


    Also fighters attack by sitting on a corvette's ass. Can't fire behind you and you WILL lose to fighters unless you have a raw numbers advantage.

  29. #79
    While i havent gone through this entire discussion just yet i found an intresting post elsewhere on the topic that I highly recommend being checked out by anyone looking to make this mod or otherwise push gearbox to do it.

    forums.gearboxsoftware.com/t/homeworld-1-remastered-weapons-mod/111157/4

    sorry for lack of link

  30. #80
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    @Koopak: Try reading my post 3 pots up. I allready debunked the work-ability of this method in strikecraft vs strikecraft situations. It works when shooting at anything frigate sized or up, but it does not work for anything smaller. Unless they turned velocity prediction on for misses they're going to fly 300-400 m wide of any strikecraft unless they're coming right at you or moving directly away. In which case their hit rate will still be awful, (0.5%).

  31. General Discussions Senior Member The Studio Senior Member Boardwars Senior Member  #81
    Player Hater Langy's Avatar
    Join Date
    Sep 2001
    Location
    CA
    Quote Originally Posted by Carl View Post
    @Koopak: Try reading my post 3 pots up. I allready debunked the work-ability of this method in strikecraft vs strikecraft situations. It works when shooting at anything frigate sized or up, but it does not work for anything smaller. Unless they turned velocity prediction on for misses they're going to fly 300-400 m wide of any strikecraft unless they're coming right at you or moving directly away. In which case their hit rate will still be awful, (0.5%).
    They could have turned velocity prediction on for misses and made it so the angles can be floating point numbers rather than integers, of course. Neither of those things sounds like it'd be a big job.

  32. #82
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    Yeah they could. but why would they? It's not like they needed them and it's extra re-coding to do.

  33. #83
    Member FriendlyFire's Avatar
    Join Date
    Jul 2006
    Location
    On the imaginary axis.
    You say you can't balance them because of those things, and then say they're not needed? Which is it? Contrary to popular belief, Gearbox does want to balance the game and make it good. Shocker, I know!

  34. #84
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    *headdesk*


    What i said is you cannot copy the HW1 behaviors with those things as they currently are, they aren't fine grained enough to copy the HW1 firing model properly, you end up with awful minimum hit rates vs strikcraft.

    I then pointed out to Langly that since gearbox isn't even attempting to use those things to balance the game there was no point them modifying them to allow a facsimile of HW1 behaviors. It's code they won't use so it's a waste of time and effort.

  35. #85
    Member FriendlyFire's Avatar
    Join Date
    Jul 2006
    Location
    On the imaginary axis.
    You're assuming they won't use them. I have no idea why you're doing this.

  36. #86
    OVERSEER NovaBurn's Avatar
    Join Date
    Mar 2003
    Location
    Fort Worth, Texas
    No one person is an expert on what will or will not work to get the fighters and corvettes fixed. Keep it civil guys. With the big files now decrypted and extractable, dig in and see what the HW1 ships are using and go from there.

  37. #87
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    You're assuming they won't use them. I have no idea why you're doing this.
    I don;t see any reason to expect them to waste time re-coding that section of the engine for it if they haven't done it allready. Not to mention the HW1 system is flat out inferior in every way to the HW2 system. The only thing it allows for that the HW2 system dosen't is for formation to give actual benefits and downsides, and that's a small price to pay for all the headaches it removes.

    @Nova: We allready have, they're using the HW2 system.

    Also it's not about weather i know better or not what will fix the HW1 strikcraft. When i say that trying to copy the HW1 system will result in a 0.5% hit rate as a bare minimum best case value vs fighters at fighter gun ranges i'm telling you a hard math calculated out fact, (subject to the assumption that HW1 miss angle usage involve true random, (or close enough to be indistinguishable), distribution, it may not, but in the context of converting to HW1 equivalent i'm working on the assumption that it is since that's a requirement for the whole thing to work). I can show you the math to back it up but it really is very basic stuff.

  38. #88
    Member FriendlyFire's Avatar
    Join Date
    Jul 2006
    Location
    On the imaginary axis.
    All I'm saying is that I see nothing that tells me "No, they will never implement decimal angle values or velocity prediction for misses." Both would be useful additions for both modders and for Gearbox themselves as part of the balance process. Neither is likely to require a large amount of work.

  39. #89
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    All I'm saying is that I see nothing that tells me "No, they will never implement decimal angle values or velocity prediction for misses." Both would be useful additions for both modders and for Gearbox themselves as part of the balance process. Neither is likely to require a large amount of work.
    It's the bolded part i see an issue with. As a modder i'd love it as even under the HW2 RNG system it has a few uses for aesthetics. But gearbox gain nothing from a balancing perspective from implementing those features because they've very clearly shown their decision IMO on that aspect. And the HW1 system really is much harder to fine tune, in particular if you've got an issue like the HW1 scout vs inty balance the only way to fix it is to play with health values. You can't just adjust weapons files to make the inty better for example. The big loss as i said is formations don't really affect the outcome of a fight positively and can actually effect it negatively. Hence why the Strikecraft of HW2 don't maintain them, they hinder but don't help.

    Thats not to say HW2's systems don't produce issues on occasion too . For example; balancing squadrons of different sizes against each other is tough as my test results over in the MP section show.

  40. #90
    OVERSEER NovaBurn's Avatar
    Join Date
    Mar 2003
    Location
    Fort Worth, Texas
    ahhh ok I think I see what you are getting at Carl. I'm about a decade or so rusty on my HW2 modding, but what is the math exactly you are using to figure this out. I used to remember but cannot for the life of me recall it.

  41. #91
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    Right let me go through the basic procedure of how a miss works in HW2, just so you understand the background on that since your rusty.

    Each weapon file has a stated accuracy and a stated miss angle. If the shot fails to connect then, (to the best we've been able to determine), the velocity prediction is turned off. thus the shot is aimed at where the target is, not where it will be when the shell has traveled.

    In addition the shot is fired off at a slight angle, that angle being a random number between zero and the miss angle. What this means is that a cone of fire effect is produced. You can see how that loosly simulates part of the HW1 system.

    The lack of velocity prediction means that unless the target is moving straight at or away from you they'll never be in the cone of fire. So that tells us what is visible of the target, Namely the front or rear aspect.

    Thanks to now knowing the max deviation angle and the aspect of the target, (and some measurements taken from the games hod files), we can at any range we choose project the potential cross sectional area the shot will fall in and we can know the cross sectional area of the target, the ratio gives us the hit percentage.

    Fighter guns are 1400 meter's range, you can and will see shots under this, but 400-600 is the break distance and various aspects tend to push the value towards the upper end due to how everything maneuvers, (i can describe this but it's tough so let me know if you want to know the details). Also in case anyone is thinking it, no, changing this too much is a baaaad idea. It stops fighters running into one another and getting in each others way, at least as long as there's not too many. That's why a lot of the HW1 strikecraft are having so many issues, not being in squadrons they're getting in each others way a lot which is causing the pathing code to go haywire. ot the only cause mind, soem of the attack ai files could do with an update but you really don;t want them trying to maneuver too close to each other, it causes pathing problems.

    Tan rule will give us the cross sectional radius, which for angle of 1 is Tan1*1400 = 25. something. PieR^2 is the area of a circle. With a bit of rounding that comes to 1900 square meters. A higgy bomber is roughly 5 wide by 2 high so 10 square meters cross sectional area. Divide that by 1900 and you get the accuracy. Obviously if you get one of the rarer situations where instead of opening fire at max range it gets a shot off much closer in it can go up, but as noted that's not common, and fighter guns are pretty much the shortest ranged weapon in the game so everything else will be worse. As such going with the upper range limit is a much fairer representation of reality.

  42. #92
    Okay taken a look at the multi-gun corvettes hod files, can't get the models to load but the harpoints show just fine. The Taidanni are screwed basically, the Kushan are flawed but fixable.

    First the taidanni turrets are all restricted to a 20-30 cone of fire centered forward, they can't even fire to their sides, they've got worse fire arcs than defenders. No matter what attack style you use that's going to create serious balance issues for them. It's fixable with a very minor model edit to give the side guns a full fire arr, but you'll never fix the bow turret's. Then again the Tai multi-gun is a crap design as such things go, so no surprises there.

    The kushan have much broader fire arcs, but suffer majorly from the two front turret weapons files being re-used for the side turret's. What this means is that whilst each turret can train to the side it's moounted on and to within 25 degree's of directly aft, it can't point all the way aft, it also can't fire full to the side and can depress enough to shoot through the hull. However with the way the turrets are mounted at an angle to the horizontal it's quite possible to give them 180 fire arcs if you reduce the depression angle. Elevation is a bit limited but fixable. Overall with an altered attack pattern they should do fine.

    Kushan Defender guns are arranged a little oddly, but otherwise share identical fire arcs. Not sure on the way their hod files are setup. Something very odd is happening there. Oddly better arcs above below than to the sides.

    Taidanni Defenders have a more normal and sensible layout of guns, and better firing arcs.

    Edit: Seems giving the Side guns on the multi-gun better arcs is gonna be harder than i thought, they really altered the model. looks better but makes for a harder job.

    Example of max traverse, (elevation is worse though), here.
    Looks about the same as HW1 assuming they twist up/down too

  43. #93
    Quote Originally Posted by Carl View Post
    Each weapon file has a stated accuracy and a stated miss angle.
    Would you please point out which specific variable in the *.wepn script controls the miss angle? Because I've searched and searched and have not been able to find it in any documentation.
    Kiith Paktu on the Steam Workshop and

  44. #94
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    Code:
    --===========================================================================--  Purpose : Lua definition file for Homeworld Ship.
    --            Contains loading information and flight dynamics information (among other things?)
    --
    --  Copyright Relic Entertainment, Inc.  All rights reserved.
    --===========================================================================
    StartWeaponConfig(NewWeaponType,"AnimatedTurret","Bullet","Kinetic_Large","Normal",5000,6000,0,0,0,0,1,1,1,5,0,0,1,1,40,25,0.1,"Normal",1,0,0);
    AddWeaponResult(NewWeaponType,"Hit","DamageHealth","Target",650,650,"");
    setPenetration(NewWeaponType,50,1,{PlanetKillerArmour=0},{SubSystemArmour=0.3});
    setAccuracy(NewWeaponType,1,{Fighter=0.1},{Corvette=0.05},{munition=0.05},{Frigate=0.8,damage=1},{SmallCapitalShip=0.6,damage=1},{BigCapitalShip=0.6,damage=1},{ResourceLarge=0.6,damage=1});
    setAngles(NewWeaponType,0,-160,160,0,60);
    setMiscValues(NewWeaponType,4,0.6);
    addAnimTurretSound(NewWeaponType,"Data:Sound/SFX/ETG/SPECIAL/SPECIAL_ABILITIES_TURRET_ON");
    Relevant value is in bold. Check Tanis for info, it's in there somewhere.

    Found it:

    http://forums.relicnews.com/showthre...wepn+file+info

    EDIT: like an idiot i bolded the wrong bit, brain fart moment guys. Fixed now though.


    Looks about the same as HW1 assuming they twist up/down too
    I've done some test and actually their attack pattern means the taidanni are holding their own. They do twist up an down, but by less than the right left shown in the picture.
    Last edited by Carl; 5th Mar 15 at 2:34 AM.

  45. #95
    Quote Originally Posted by Carl View Post
    Code:
    setAngles(NewWeaponType,0,-160,160,0,60);
    I have been messing with that value for quite sometime and I have noticed no real discernible difference; In all my experiments it does not and has not forcefully limited the cone of bloom for missed shots to be within a certain degree.

    From my observations that value only defines the cone within which the ship is given a green light to fire. For:

    • Fixed weapons this means the weapon may fire if an enemy ship is within this cone, regardless of whether or not the shot will land.
    • Gimbals this means the same thing but with the addition of aiming the muzzle of weapon to any point within that cone.
    • Animated turrets this means that the turret can fire randomly within this cone at a target that is flying within or through this cone if the target was deemed a "miss" by the RNG.


    A workaround I came up with for the time being was to create a "missile bullet," or a missile modified to act like the bullet that is normally defined in the weapon script. From my field trials it does work, but the issue then is that ships (Specifically fighters in this case) cannot rotate themselves and aim their forward-facing guns onto the target quick enough to land the shot before the target changes positions.

    The missile-bullet approach then requires two things:
    1. Reducing the angle within which ships are allowed to fire at enemy ships; This prevents them from firing at a target when their aim is still several meters off.
    2. Modifying the ship to be able to position itself to attack more quickly; Either through increasing rotational speed/accel/deccel values (Primarily for forward-facing guns on fighters) or modifying the ship's attack styles and/or attack styles used to give the ship more opportunities to line up shots against its target.

  46. #96
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    • Animated turrets this means that the turret can fire randomly within this cone at a target that is flying within or through this cone if the target was deemed a "miss" by the RNG.
    Erm this is exactly what cone of bloom IS. You just defined it right here. And it does work, but when shooting at strikcraft because it's aiming at where they are when it's fired and because they maneuver so much it won;t give a consistant looking output. Try it in frig vs frig and it works perfectly. I've been seeing it with frigs whilst running my matchup tests.

  47. #97
    Quote Originally Posted by Carl View Post
    Erm this is exactly what cone of bloom IS. You just defined it right here. And it does work, but when shooting at strikcraft because it's aiming at where they are when it's fired and because they maneuver so much it won;t give a consistant looking output. Try it in frig vs frig and it works perfectly. I've been seeing it with frigs whilst running my matchup tests.
    Adjusting the cone of bloom does nothing. Consider the following:

    Kus_AssaultGun.wepn:
    Code:
    StartWeaponConfig(NewWeaponType,"AnimatedTurret","Bullet","Kinetic_Small","Normal",2000,3000,0,0,0,0,1,1,1,0.001,0,0,1,0,55,55,0.1,"Normal",1,0,0);
    AddWeaponResult(NewWeaponType,"Hit","DamageHealth","Target",72*0.9,72*0.9,"");
    --AddWeaponResult(NewWeaponType,"Miss","DamageHealth","Target",72*0.9,72*0.9,"");
    setPenetration(NewWeaponType,5,1,
    {Unarmoured=0.71},
    {LightArmour=0.89},
    {PlanetKillerArmour=0});
    setAccuracy(NewWeaponType,1,
    --{Fighter=0.137},
    {Fighter=0.0000000000001},
    {Corvette=0.135},
    {munition=0.1},
    {Frigate=0.8,damage=1},
    {SmallCapitalShip=0.6,damage=1},
    {BigCapitalShip=0.6,damage=1},
    {ResourceLarge=0.6,damage=1});
    setAngles(NewWeaponType,1,-150,150,-10,60);
    --setMiscValues(NewWeaponType,1.5,0);
    addAnimTurretSound(NewWeaponType,"Data:Sound/SFX/ETG/SPECIAL/SPECIAL_ABILITIES_TURRET_ON");
    In a field trial of Kushan Assault Frigate vs. Turanic Fighter:
    • With AddWeaponResult(NewWeaponType,"Miss","DamageHealth","Target",72*0.9,72*0.9,"")

    • Without AddWeaponResult(NewWeaponType,"Miss","DamageHealth","Target",72*0.9,72*0.9,"")



    There has been absolutely no change at all concerning the rounds (excluding the immediate round expiration with the additional "miss" weapon result) against the target ship.

    Cone of Bloom - 900m



    Even with a constraint of 1° the cone of bloom for the bullets remain unchanged and still miss pretty much all their shots. The only time they really land a hit is if the target is really, really close (Less than approx. 100m) or if the target stops moving and even then they miss the majority of their shots.

    This is not a viable solution.
    Last edited by MechWarrior001; 5th Mar 15 at 6:00 PM.

  48. #98
    Member Carl's Avatar
    Join Date
    Aug 2006
    Location
    The Crossroads of Inertia
    Image links are broken. Til i can see those i really can't tell whats going on.

    Also whats with the on miss weapons code?

  49. #99
    Quote Originally Posted by Carl View Post
    Image links are broken. Til i can see those i really can't tell whats going on.
    Should be fixed.

    Quote Originally Posted by Carl View Post
    Also whats with the on miss weapons code?
    It was part of a test to see if RNG missed shots could be made to deal damage; Which turned out false. Missed shots didn't actually deal any damage despite the weapon specifying a DamageHealth effect for a "miss."

  50. #100
    The short and skinny of what MechWarrior is trying to say is that setAngles has ZERO impact on angle at which a miss misses. That is to say that if oyu have a fighter flying in a strait line, and have another attack it directly from behind with the accuracy set to 0.001 it will only hit once in 1000 shots because none of the rng misses will hit, even if the cone of ATTACK is set to 0.01.

    Lets define some terms.

    A Cone of Fire is the cone in which a round may land when fired in reference to its static barrel.
    A Cone of Attack is the cone in which the game engine checks to see if a target is inside of before attempting an attack.
    A Miss Cone is the cone is which the round may land if the RNG rolls a miss.

    Homeworld does NOT have cone of fire as a operable variable. Instead it uses a Cone of Attack to decide when to attack, rolles an RNG value between 0 and 1, compares it to the accuracy on the weapon table, if its a hit, it draws a round chasing the target and applies damage on hit, if its a miss it fires the round off at a random angle defined by the Miss Cone.

    We do not have access to the variable controlling said Miss Cone, which is the problem, since we cant change it it stays the same generally speaking which is why when a fighter attacks another fighter it has zero chance of hitting UNLESS the enemy fighter moves into the round while it is in motion. However against say a kushan corvet from above or below it will have a small chance of hitting with the RNG misses, and agains a frigate on up its almost guaranteed.

    I recommend using Cone of Attack and Miss Cone instead of Cone of Fire in all future discussion to limit confusion.

    If we gain the ability to access the variable controlling miss cones, and then make it be able to "miss" at 0 degrees, then we have a cone of fire

Page 2 of 3 FirstFirst 123 LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

     

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •