Realism, Gameplay, Aesthetics and Canon

The most appropriate place for Questions, Queries, and Quandaries regarding the nature of the Vega Strike universe and its past, present, or future history. Home to the occasional unfortunate RetCon.
Post Reply
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

Vega Strike is a game project that tries to find a balance among Realism, Gameplay and Aesthetics, with its Universe being a major component in bringing these three things together.

But up until now it has had two problems.

One is that it is poorly documented and detailed and jackS and his team does not have the time to fix this anytime soon, so we don't know exactly what VS is or was ultimately supposed to look, feel and play like.

The second is that there seem to be a number of places where VS currently compromises realism, gameplay and aesthetics in ways that do not benefit any of them, while offering little or nothing in return. I personally believe there are some places where this is a really big problem.

And many times it seems like these two problems compound each other, with us content creation folks spreading and building off of badly designed aspects of the game, which are not even canon, because the universe documentation is such a mess and incomplete that there isn't clear enough indication that they aren't canon.

So what I propose is that we brainstorm and discuss whatever we feel are the vague, nonsensical or unworkable parts of the game and universe. Then where the solutions don't seem to conflict with canon we make changes or establish content standards either suggested or enforced. And where the solutions seem to conflict with canon, maybe I can bring it up with jackS and see how much flexibility there is on the issue and the reasons behind it.


P.S. Whenever we go too far off topic on other threads discussing this sort of thing, this is a good place to continue that discussion, just remember to quote what you are responding to so that everyone knows what you're talking about and what was already said in the other thread.
snow_Cat
Confed Special Operative
Confed Special Operative
Posts: 349
Joined: Thu Jan 05, 2006 12:43 am
Location: /stray/
Contact:

Re: Realism, Gameplay, Aesthetics and Canon

Post by snow_Cat »

^ - -^ Seems reasonable. I'll try not to taint the process with a mishmash of sci-fi comedy series.
  • Recorded for future reference regarding the past deliberations about the future. Do try not to use up all the query cycles.
Last edited by snow_Cat on Mon Dec 28, 2009 8:16 am, edited 1 time in total.
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

chuck_starchaser wrote: Politics: This project is NOT abandoned; and we know its project leads are not too keen on realism.
Take the case of shields, for example: According to JackS they work by deforming space. To me, that's THE most
absurd account of shields in any universe (and I never liked any account much, for starters); but it's something
we have to live with.
The explanation behind shields probably follows the explanation behind the FTL methods used in VS which deform space. It might not be a big deal to change the explanation as long as the gameplay mechanic is the same or preferable.

For example, maybe shields could be like modern day reactive armor, only based on small but high power plasma thrusters (which could maybe help explain why some legacy vessels lack lateral thrusters). Or wafers of matter-antimatter explosive reactive armor that the ship slowly replaces for you as they are spent.
Believability: My eternal position about changing accelerations (as well as turning speeds and maneuvering
accelerations) has more to do with belivability than with scientific realism: I cannot accept that the llama is as big
as 40 meters long if I can't "feel" it; --if I can turn it around in half a second--, for instance.
I wholeheartedly agree. This is one of the places where VS hurts its realism, gameplay and aesthetics for no benefit I feel. Realism is probably hurt by the extreme stresses and power requirements that come from the kind of accelerations a lot of the ships in VS currently pull off. Gameplay is hurt because the ultra responsiveness of ships' turning makes combat kind of twitchy, without the skill factor that one expects out of a game that calls itself a sim; as well as gameplay balance by so many of the playable ships being so similarly maneuverable in combat. Aesthetics are hurt because you lose the sense of scale, both within your own too-responsive spacecraft and the things you are flying past, whose artistic details you will likely also miss for the same reason.

Fortunately, I think this issue is really unintentional and not actual canon, possibly caused by there just not being someone with the time and/or motivation to go through and rebalance all of the ships in units.csv with sane flight characteristics. Also, as more modern and detailed meshes enter the game, the benefits of this kind of rebalancing will probably become more rewarding and obvious.
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Re: Realism, Gameplay, Aesthetics and Canon

Post by chuck_starchaser »

When rebalancing, it is important to
  • a) Have written down criteria, models and guidelines
  • b) Make a spreadsheet with old and new values, as well as columns for any data involved in the criteria, such as size and mass in this case.
  • c) Have a brainstorm forum thread during this spreadsheet-based rebalance draft.
  • d) Once the spreadsheet work is done, implement all the changes at once, have some people test it, then merge the branch.
  • e) Add yet another spreadsheet column for any post-rebalance tweaks, and a column next to it for comments and/or forum links.
Otherwise you get the WCU syndrome...
People used to re-balance WCU like every 4 months. Nobody knew it had already been rebalanced, by who, or the reasons why; so they rebalanced it again and again.

At PU, we recently rebalanced ship sizes using the above modus operandi, and now we got sizes of ships and stations we can agree to, and when not, that we can at least prove that we put a lot of thought into it. And if someone in the future says (ship X is too large/small), we can go back to the brainstorm posts about ship X and review the references that were used, and the arguments and/or compromises, and ask the complainer whether he/she would like to put a motion to formally re-open that discussion.
Otherwise it's a "too many cooks" problem.

Rebalancing the accelerations and maneuverabilities in VS would be great; but we also talked about reducing some ship sizes, like the Clydesdale. This would affect maneuverability, so we'd have to do the size rebalancing first.
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

Good points.

I just emailed jackS asking about whether or not the scale of capital and like sized vessels is a fixed part of canon or if it is flexible enough that they may be scaled down for the sake of content quality and gameplay issues. Will see what he says about it.
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Re: Realism, Gameplay, Aesthetics and Canon

Post by chuck_starchaser »

From a modeling standpoint, anything bigger than a kilometer is just not representable. There's nothing you can do as a modeler to convey the size of the thing.
Well, it IS doable; but not as a single model; you'd have to make heavy use of sub-units. This is okay for stations, but it takes huge amounts of work, so it's practically out of the question for ships.
Essentially, conveying a ship's size boils down to two tricks: 1) Meter scale doors and windows; and 2) Inch-scale detail, such as hand-grips, railings, door-handles and buttons. Any human structure, no matter how big, will have meter-scale pass-throughs and inch scale hand-grabbables. At 1 kilometer size, an inch is a quarter of a ten-thousandth of the ship's length. Assuming you wanted a pixel to fit in a one inch thing, and that the ship is unwrapped to the full length of the texture, the texture would have to be 40,000 texels long.
So, it takes all kinds of trickery to unwrap a 1 kilometer ship with even a minimum number of inch-scale details.
More generally, the larger a ship is, the harder it gets to ensure that it looks its size. Presently, the Clydesdale may be 5 km officially, but it looks the size of a truck, in space. Making it 500 meters, but actually taking care of adding detail, will make it look 100 times bigger than it does now.
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

Yeah that's what I tried to imply, albeit in summarized form-
myself wrote:I ask mainly because making objects that big look realistically detailed but
not repetitive or non-canon in layout and still be optimized for realtime
rendering is difficult. At least when so many of the ships are on that
scale. Also there are some gameplay issues that come with having
player-purchasable merchant faction ships like the Clydesdale and Ox be so
large.
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

Meh, detail is almost always better because otherwise you end up with so many misunderstandings. Its just in this case I figure briefness could mean a faster response, because there is simply less to respond to. :)
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Re: Realism, Gameplay, Aesthetics and Canon

Post by chuck_starchaser »

True.
I was just thinking of another way of explaining the scale problem:
Right now there's a split, in Vegastrike, not unlike the split between real and
official value of a currency in some republic where the government says "our peso is
worth one dollar, or else"; and where price controls are in place to try and force the
purchasing value of their currency. So, everybody gets used to there being real and
fictitious values.
In Vegastrike, official sizes of ships often run into kilometers, or tens of kilometers;
but all the evidence points out to their real sizes being orders of magnitue smaller.
So, the problem is not that ships are too large or too small; the problem is this divide;
this split between real and official sizes. It doesn't do well for the game and players
to be told that an Aera ship perceived to be the size of a coffee table is 100 meters in
diameter. It doesn't help to be told that truck-sized cargo ships one sees around are
supposedly many kilometers in length. It just creates gratuitous confusion.
And to close this gap or divide will take pulling both sides of it together:
Detail will have to be added to ships, and art guidelines established for how much
greebling and of what kinds have to be present in order to convey a sense of their
size. But this can only go so far, due to time and technical limitations; so the other
thing that needs to adjust, as well, is the official sizes, so as to make it possible to
have such guidelines adhered to.
I think the current sizes were based on book and movie sci-fi, which don't have to
live within game modeling constraints. For a movie, you can build a model of a
ship or station with whatever amount of detail you want; since you can render
them off-line, frame by frame, using super-computers with thousands of processors
working in parallel.
But for a game, we're constrained to what a single GPU can push; and to keep the
polygon count under a million, we end up with some maximum range in detail size
to overall mesh size, of about 3 orders of magnitude. So, the minimum feature
size for a kilometers long ship would be in the meterS (with capital S) range. But
nothing in the meters range can serve as a size reference. The first, soft anchor
in visual scale perception is the single meter range. Nothing built by humans can
lack meter-sized features, our brains assume. The second, hard anchor is the
centimeter or inch range. Nothing built by humans can lack things that a human
hand can grab.
With these two anchors in place, ship size scale is conveyed successfully.

Image

But, with inch-scale detail, it becomes exponentially difficult to produce models
larger than 250 meters over-all. The above Bengal is 600 meters, and there are
very few areas with inch-sized detail. You have to resort to keeping scale reference
greebling to a few little "islands of detail" around the ship.
That's why I suggest 1 kilometer as upper bound; and a very hard, and exceptionally
reached, upper bound.
JackS may think of this proposal as a loss. Explain to him that while it may be a loss
in official sizes, it will be a double or triple gain:
  • We CAN then begin to convey ship sizes.
  • We CAN have greebling standards and tutorials and enforced rules
  • Eventually there will be no gap --i.e.: You won't have to tell players how big ships
    are; they will feel their sizes.
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

Okay, got a response back from jackS on this subject (the text with ">" in front of it is what I had said):
jackS wrote:> But this has turned up a question which might involve canon. Going by
> various references, it looks like the capital ships and super freighters in
> VS are enormous monsters with whole kilometers between bow and stern. So my
> question is, is it essential to canon that these ships be modeled and scaled
> in game at these enormous sizes?

It is not strictly essential to canon that ships be as enormous as
they currently are, but it is definitely preferred that the larger
vessels be many times larger than the smallest, and that even the
smallest are more boat sized than car sized (a spaceship is not a
Winnebago with wings, although an in-system parasite craft may come
close ;-) ). That said, the enormity of any given vessel should be
taken with the following consideration in mind - in my imagining,
these vessels occupy very large enclosing volumes, but most of that
volume is vacuum. Namely, any particular section of a large ship
should primarily be of the skeletal (think: Eiffel tower) or shell
(think: Eiffel tower inside a space-zeppelin) construction paradigm.
The shell variant, especially for military ships, provides lots of
radiator real-estate on the hull, with vital systems located suitably
far enough apart for insulation and redundancy, while skeletal
variant, for both military and civilians, is a materially cheap way to
enclose lots of space, for cargo or other attachments, or to connect
at a distance potentially hazardous components, such as engines, to
more delicate components, such as crew quarters). As one scales up,
given that volume is scaling as feature size cubed, but surface area
as feature size squared, the scaling of specialized larger vessels
with larger power budgets could easily run quite large. Smaller
vessels, therefore, would have very different aesthetics than the
larger, being comparatively crammed together.

> I ask mainly because making objects that big look realistically detailed but
> not repetitive or non-canon in layout and still be optimized for realtime
> rendering is difficult. At least when so many of the ships are on that
> scale. Also there are some gameplay issues that come with having
> player-purchasable merchant faction ships like the Clydesdale and Ox be so
> large.
>

I think some repetition is certainly ok, and if we build up a library
of common subcomponents, it might even be preferable in some cases,
but I certainly see the difficulty you'd run into - skinning several
square kilometers of ship surface with a single texture is going to
make for excessive pixelation. Short of breaking up the model into
smaller logical pieces I don't see any particularly simple solution -
but as I've been hoping to see the larger models broken up into
smaller logical pieces for some time (for other reasons, as well as
this one), my perspective may be myopic. Certainly, the game engine
has long supported (and there are already examples of, such as the
Clydesdale) models built out of multiple meshes. If done in the
fashion of the Clydesdale, there's still only one "unit" and thus the
physics engine is not overburdened, but the potential for doing
independent damage modeling is only possible when each subcomponent is
not only a separate mesh, but a separately defined "unit" (and the
ship built as a collection of these subunits). However, from a purely
graphical standpoint, the two offer the same thing, and the former is
less work for the content creator.

One thing that may help (although perhaps it won't) is to consider
that the crewed portions of a larger vessel, and the infrastructure
connecting them to the docking apparatus or bay, will be
proportionately much smaller than those on a larger vessel. If, for
instance, a civilian cargo carrier has a very modestly sized command
center at one end that can be reasonably modeled in detail, and then a
set of almost naked engine/reactor mounts that can each be
independently modeled and then duplicated, a docking boom, and then a
body-section consisting of a vertebrae-like sequence of identical
container clamps and a set of rail lines on which cargo cranes move
and another rail line on which a passenger transport moves (perhaps
invisible as it could be in the center, but I digress)... if the
design, like this one, is composed out of replicable parts, then,
clearly, it'll be obvious that some of the parts are being replicated
(in this case even logically so), but so long as the parts that one
focuses on most (vital systems to shoot at, places you dock) are
modeled in higher detail, perhaps the uniqueness of a particular
synthesis of component parts can still come through. For ships with
large stretches of hull area without differentiated function, we used
to have a "detail texture" option to fill in a tiled repeating texture
pattern on top of the overall hull pattern, but I don't know if that's
still functional, or even desirable.

As for gameplay - the player-purchaseability of the Ox and Clydesdale
has always been a bit of an undercooked thing. No ship that large
would be steered like a fighter, nor required to dock like one.
However, until we set things up properly to let the captain of a
larger vessel take their pinnace over to arrange business dealings
whilst shuttles move the actual goods to and from the vessel in
question... I don't know when exactly that will be fixed - we've known
it's not well done for some time, but the nature of the changes
required to do what we really would like to see done with capital
vessels (crewing, etc.) is sufficiently daunting, and half-measures as
temporary fixes sufficiently unappealing that it hasn't seemed like
the most judicious use of time, especially given the fact that the
vast bulk of players will not see themselves into a larger vessel
until they've been otherwise playing for many hours, if ever.

The number of large vessels is also likely to change significantly
when A) better production models are introduced and B) more civilian
craft, both of the private and infrastructure (think: space-bus,
commuter traffic, etc.) craft are developed - the total number of
civilian craft will probably remain in the same order of magnitude,
but many more of them will be smaller vessels. As it is now, if we
want to create 50 civilian vessels, as there are only a few civilian
craft types we have models for, an inappropriately large number of
them end up being Ox class vessels and so forth. Similarly, when we
assign military craft to a system, we do so without any particular
structure or organization - the capital vessels should be arrayed in
fleets, with logistical support vessels and escorts and so on and so
forth - this too will help in terms of how many of these we'll be
seeing at a given time.
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Re: Realism, Gameplay, Aesthetics and Canon

Post by chuck_starchaser »

You can tell JackS that detail textures will be there soon enough. Klauss has alredy coded all the plumbing necessary; and it's just a matter of writing a shader that uses them, and I will do so shortly.
Having said that, there's a question of what kind of detail to use.
My current plan is to have a single detail texture used universally. It's a 256 x 256 tile that has various kinds of "noise": Two channels are dU and dV normapmap encoding to add slight wobbliness.
Another channel has "perlin clouds" to modulate material shininess by.
And the fourth channel has TV type noise for modulating diffuse color.
My plan for the shader is that it will automatically detect the intent of the artist by looking at the diffuse and specular colors and comparing their degree of color saturation: Green diffuse and gray specular, for instance, will be decoded as "green paint" and will use fresnel reflectivity.
But additionally, the application of detail will be suited to the type of material: Matte materials will get TV noise modulation. Paints will get normalmap modulation, giving them slight ripples. Metals will get clouds in the shininess.
All these modulations will be applied VERY moderately, hardly noticeably. Their purpose is to hide the inherent pixelation artifacts of limited size textures looked at in close range; and give the illusion of greater detail, but without calling undue attention to the repetitiveness of the detail texture's tiling and repetition.

What I think JackS is thinking about is something very different: Using detail textures as a way to introduce complexity; --the way Fendorin wants to use them.
The good news is that the detail texture plumbing that Klauss wrote makes room for, by default, two detail textures, rather than one. So, we could have "MY" detail texture AND the other kind; and probably we will.
But there's a few problems with it:
  • A tiling texture cannot be as complete as the the regular texture set. We cannot have a triple set for diffuse, specular and glow, for example. We only have one tiling texture; so we can either use it for diffuse OR for specular OR for glow OR for normalmap; but not all four, or even two. Or we could think of some snazzy philosophy of using the red channel to modulate diffuse intensity, green and blue for normalmap, and alpha for shininess or glow; but the question then is, will such a scheme serve all purposes of all artists all the time? We cannot have a different scheme for every ship, or we'll need a different shader for every ship.
  • A visible tiling texture can get boring, and would probably generate artifacts on small surfaces where it wasn't meant to appear. Furthermore, controlling the scaling of the unwrap is hard to begin with; but with visible, repetitive detail, it must be perfectly controlled, or else we would see the same detail pattern appear larger in one place and smaller in another. All of which means that, at minimum, we need one texture channel in the main texture to control, in yes or no fashion (such as by a 1-bit alpha channel) whether the detail texture is to be applied or not, to a given texel. This way, we can paint on the main texture, where we want detail texture tilings to happen, and where not. The only texture I can think of where we could add a 1 bit alpha, right now, would be the damage texture. That is, we would paint "detail windows" on the main texture, through which the detail texture shows, instead of, or modulating the main. Better than a 1-bit yes or no window, it would be good to have a parametric channel that can control the intensity of the modulation, or perhaps with ranges for different philosophies of application.
  • Even so, given that the detail texture's frequency of tiling is hard-coded at 16 x 16 squares over the main texture, I believe, a given pattern, such as a cargo hatch door that we want to appear in several places around the hull, could only appear at discrete locations on the main texture. What this implies is that the UV islands will have to be placed strategically so that their "detail windows" are aligned with 1/16th subdivision lines on the texture map. This will be VERY difficult. Correction: It would be TOO difficult. UV unwrapping is already VERY difficult, without this added, new constraint.
What all of this boils down to is that detail textures can't be used to produce miracles in terms of conveying ship size. The inherent problems with them dictate subtlety in their use. Creating the illusion that the outer hull of a ship is made of discrete tiles is probably doable; but showing hand-grips and door-knobs and hatches and things like that on them is out of the question: It would simply draw the attention to their boring repetitiveness and their unavoidable scaling and offset artifacts.

The idea of using sub-units to represent the largest ships as assemblies of modules was already on the table, and on nphillips and Fendorin's plans. Repetitive meshes, as opposed to sub-units are less attractive due to lighting bakes. Sub-units would shade correctly regardless of orientation, and they aren't any harder to use; they are easier to use, in fact.
However, the modular solution is one that is rather restricted to cargo ships. I don't see how we'd use subunits for carriers like the Agesipolis. We could make each engine and bay a subunit; but these would be subunits specific to that ship, as opposed to standard container units shared by many cargo ships. IOW, subunits or no subunits, very large ships involve very large amounts of work to represent in ways that convey their sizes well.
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

I agree and as such have never tried to use the detail texture feature myself as a content creator. IMO, the subtle procedural texture effects you plan to use would probably be the best application of this feature. It seemed to me from what jackS said that he didn't necessarily consider it a very effective tool either, in its previous form at least.

I also read this:
jackS wrote:It is not strictly essential to canon that ships be as enormous as
they currently are, but it is definitely preferred that the larger
vessels be many times larger than the smallest, and that even the
smallest are more boat sized than car sized
. . . To mean that we can build the very largest ships to be 1-2 km in length instead of the current 7km. So a reduction of the largest vessel lengths by 1/3 to 1/6, the smallest vessels lengths would not be reduced at all, and ships in the middle could be reduced by some amount in between.



Another thing which jackS mentioned and that I have wanted to bring up for a while, is building ships out of multiple sub-units that can have their own damage modeling.

I wonder if it would make sense to model ships, both big and small, so that they can be damaged and destroyed in pieces, eventually with an engine feature that links up the status of those pieces, or modules, to the ship's ability to function in different ways.

So if you destroy an engine pod, this damage is both visibly obvious and reduces the ships maneuverability. Destroy one of its weapon modules and that weapon no longer fires. Destroy a radiator structure and heat builds up faster. Destroy the crew section and the ship floats derelict. Destroy the main reactor section and it explodes.
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Re: Realism, Gameplay, Aesthetics and Canon

Post by chuck_starchaser »

Deus Siddis wrote:I also read this:
jackS wrote:It is not strictly essential to canon that ships be as enormous as
they currently are, but it is definitely preferred that the larger
vessels be many times larger than the smallest, and that even the
smallest are more boat sized than car sized
. . . To mean that we can build the very largest ships to be 1-2 km in length instead of the current 7km. So a reduction of the largest vessel lengths by 1/3 to 1/6, the smallest vessels lengths would not be reduced at all, and ships in the middle could be reduced by some amount in between.
I read it the same way. Well, largest to smallest ratio WILL suffer; OR smallest size will. Can't have your cake and eat it too. Thing is, the original premise is unrealistic. OR, we could build huge ships out of many subunits; but artists are not an endless resource, and I think it's better to have a higher percentage of ships looking decent than spending all artistic energies into one or two massive ships.
Another thing which jackS mentioned and that I have wanted to bring up for a while, is building ships out of multiple sub-units that can have their own damage modeling.
I believe this is quite doable right now.
I wonder if it would make sense to model ships, both big and small, so that they can be damaged and destroyed in pieces, eventually with an engine feature that links up the status of those pieces, or modules, to the ship's ability to function in different ways.

So if you destroy an engine pod, this damage is both visibly obvious and reduces the ships maneuverability. Destroy one of its weapon modules and that weapon no longer fires. Destroy a radiator structure and heat builds up faster. Destroy the crew section and the ship floats derelict. Destroy the main reactor section and it explodes.
I think this is how it would mostly work right now. In WCU you could attack the individual weapons of large ships. I haven't really tried; but probably in PU as well; I think there was a hotkey for "sub-unit target", maybe Ctrl-T or who knows. But yeah; anything that is added as a subunit is damaged independently. Inter-relating damage effects may take AI work though; like radiator->heat build-up. We'd need to model heat first, for that, anyways.
There's runtime costs with using sub-units, though; both on the CPU and GPU sides; but it should be okay for very large ships, since you wouldn't see a dozen of them together, probably.
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

chuck_starchaser wrote:I read it the same way. Well, largest to smallest ratio WILL suffer; OR smallest size will. Can't have your cake and eat it too. Thing is, the original premise is unrealistic. OR, we could build huge ships out of many subunits; but artists are not an endless resource, and I think it's better to have a higher percentage of ships looking decent than spending all artistic energies into one or two massive ships.
So maybe minimum 10 meters long ships, max 2 kilometers, that should not violate either requirement.

Looking at the wiki, the capital ships go as follows from largest to smallest:

Code: Select all

Dreadnaught: 1
Battleship: 1
Battle Cruiser: 3
Cruiser: 8
Destroyer: 4
Frigate: 4

Corvette: 7

Strike Craft: Lots
So we don't need to worry about the Dreadnaught or Battleship sizes that much, there's only two of them anyway. Battle Cruisers we should be concerned with having a reasonable level of difficulty to model. And Cruisers we definitely need to think about, because there are lots of those. Perhaps something like:

Code: Select all

Dreadnaught: 1200-2000 meters
Battleship: 800-1200 meters
Battle Cruiser: 600-800 meters
Cruiser: 400-600 meters
Destroyer: 300-400 meters
Frigate: 200-300 meters

Corvette: 100-200 meters

Strike Craft: 10-100 meters
I think this is how it would mostly work right now. In WCU you could attack the individual weapons of large ships. I haven't really tried; but probably in PU as well; I think there was a hotkey for "sub-unit target", maybe Ctrl-T or who knows.
Yeah, VS has a similar hotkey to target turrets which can be destroyed independently. The Dodo also has a cargo module that can be destroyed separately, though I don't know if this actually reduces its cargo capacity or destroys any on-board cargo, or how it is repaired.
But yeah; anything that is added as a subunit is damaged independently. Inter-relating damage effects may take AI work though; like radiator->heat build-up. We'd need to model heat first, for that, anyways.
yeah, it would probably require a lot of changes to units.csv as well.
There's runtime costs with using sub-units, though; both on the CPU and GPU sides; but it should be okay for very large ships, since you wouldn't see a dozen of them together, probably.
Plus I think Klauss brought up that it bogs the GPU down to have to deal >30,000 triangle models that have not been broken down into sub-units.
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Re: Realism, Gameplay, Aesthetics and Canon

Post by chuck_starchaser »

Deus Siddis wrote:

Code: Select all

Dreadnaught: 1200-2000 meters
Battleship: 800-1200 meters
Battle Cruiser: 600-800 meters
Cruiser: 400-600 meters
Destroyer: 300-400 meters
Frigate: 200-300 meters

Corvette: 100-200 meters

Strike Craft: 10-100 meters
I'll update my sizes list in the UberShader thread.
Plus I think Klauss brought up that it bogs the GPU down to have to deal >30,000 triangle models that have not been broken down into sub-units.
No; that's a limit in triangles per-mesh; not per unit/sub-unit. And it's an ad-hoc rule of thumb. The actual limit is 64k vertex indexes, which is not really a function of the number of triangles. A more precise calculation would be to multiply the number of vertices that fall on edges marked sharp or on texture seams by two, then add to that the sum of all other vertices. A simpler way is to just multiply the number of triangles by three, on the worst-case assumption of them all being set to flat shading (really, we should NEVER use flat-shading); and such a worst case would result in a limit of 21 k triangles. But that's per-mesh.
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

So now that we have some idea of ship size classes lets move on to weapon size classes. I believe VS currently has six of these- Light, Medium, Heavy, Light Capital, Medium Capital, Heavy Capital. We need to decide if these are enough and we like this naming convention. Then we need to decide on some standards or guidelines for the physical dimensions each class of weapon is in actual meters, for artistic consistency and intuitiveness to players sizing up the firepower of another ship.

I think that one thing we should really keep in mind though as we work this out, is the size of the ships these sizes of weapons will be mounted on as well as how they will be mounted on them. So if a particular size of hardpoint is always fixed instead of turreted on ships of a certain size range (like 'Corvette' for example), it should be because the weapon size is so big relative to the rest of the ship, that there would be practical drawbacks to turreting it.
Neskiairti
Confed Special Operative
Confed Special Operative
Posts: 334
Joined: Thu Jan 11, 2007 4:10 am

Re: Realism, Gameplay, Aesthetics and Canon

Post by Neskiairti »

I'm not sure weapon size is really important.. id say simply class it by what type of ship can mount it. Capital/non capital.. maybe what faction creates it.. an alien weapon isnt going to interface with terran tech very easily IMO.

now, what would be more interesting... classifying weapons by the damage and projectile.. this is what I'm playing with for my own game.

Projectile Class:
Dumbfire Projectile
Seeking Projectile
Energy Pulse
Energy Stream
Energy Wave
Plume
Proximity
Skipping

Damage Class:

Concussive
Piercing
Heat
Electrical
Magnetic
Gravitic
Organic
Upper Photonic (Penetrating frequencies)
Lower Photonic (Non Penetrating frequencies)
Spacial
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

Neskiairti wrote:I'm not sure weapon size is really important.. id say simply class it by what type of ship can mount it. Capital/non capital..
But 'Capital' basically refers to the size of a ship, so that's still classing weapons by size.

I'm not really sure how you can say size isn't important either, when you have such diverse scale across VS' ships. It seems to me that if you have much larger craft you'll want to put on them some larger weapons that might not be able to mount on smaller craft. Because there are things a few large weapons can do that numerous small ones cannot, or at least not as well.
maybe what faction creates it.. an alien weapon isnt going to interface with terran tech very easily IMO.
Sounds plenty reasonable to me.
now, what would be more interesting... classifying weapons by the damage and projectile.. this is what I'm playing with for my own game.
Remember that we're talking about hardpoints that you house weapons in or on, not actually weapons themselves. This is why you can choose what weapons you want to put on a particular ship, rather that it coming with a fixed arsenal.

But what you are describing does exist in VS already, with gun, special gun, missile and special missile hardpoints (in addition to the size classes). The question then is should this be expanded with more categories. Expanding it means more narrowly limiting what weapons you can outfit any particular ship with. That reduces player customization options, but also makes ships more unique and diverse.
Neskiairti
Confed Special Operative
Confed Special Operative
Posts: 334
Joined: Thu Jan 11, 2007 4:10 am

Re: Realism, Gameplay, Aesthetics and Canon

Post by Neskiairti »

true about it being hardpoints, i got a bit carried away.. but anyway.
when I say capital weapons, I mean deathstar lasers, or the beam cannons from the freespace series..
These are where the ship is designed around the weapon. It isnt about size, its about function.

anyway. You can theoretically attach any weapon to any ship, the only question is if you have the energy resources to power it.. and if it unbalances your ship. Having a massive 80ton slug thrower on a light fighter is certainly possible... however trying to move that fighter around would not be so possible, and firing it would probably give you more thrust than your primary engines. thrust in the wrong direction and likely at an angle from your center of balance.

plus you cant likely reload it, or if you can, your store of ammo might ammount to 5 shots :p

now a high energy beam weapon on the other hand, sure you can stick it on, but that little figher's power plant is barely enough to illuminate a flashlight let alone a laser or particle cannon of that scale.

whats the point of limiting by arbitrary factors when there are prfectly good reasons why X weapon just wont work.

the weapon should be mountable as long as you have the infrastructure for it.. a gauss/rail cannon would require electricity and a munitions storage.
some weird sticky organic space borne acid spitting weapon might need some alien storage tanks and glands to create the compound.

soo... you could make that as simple or as complex as you want.. requiring individual components.. or just toggle a bitflag for that type of weaponry.. and maybe add in purchasable components that expand a ship to support other race's technology.
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

Neskiairti wrote:You can theoretically attach any weapon to any ship, the only question is if you have the energy resources to power it.. and if it unbalances your ship.
Those are a couple of the reasons why you have some kind of size limitations. Larger versions of things generally eat more power and have more mass.
Having a massive 80ton slug thrower on a light fighter is certainly possible... however trying to move that fighter around would not be so possible, and firing it would probably give you more thrust than your primary engines. thrust in the wrong direction and likely at an angle from your center of balance.
You're overlooking major engineering issues. Your assuming that stations would have the equipment to easily frankenstein together ludicrous combinations like that so that the ship doesn't just fly apart, instead of simply attaching weapons to hardpoints that are physically compatible with each other. Or that all manufacturers build all craft of all sizes with huge overbuilt hardpoints that can mount any weapon, regardless of how practically useless that extra support is. And your assuming the stress of the weapon's recoil won't destroy the small fighter outright.

You should keep in mind that Vega Strike is meant to be a kind of spaceflight simulator, so things are supposed to make some kind of realistic, believable physical sense. Removing hardpoint size limitations to make any extreme combination possible works against that design philosophy.
whats the point of limiting by arbitrary factors when there are prfectly good reasons why X weapon just wont work.
Because these factors aren't arbitrary at all, see above.

And again these ships and their weapons have to be represented graphically. Which means us modelers still have to know the sizes of these things we are supposed to model.
Neskiairti
Confed Special Operative
Confed Special Operative
Posts: 334
Joined: Thu Jan 11, 2007 4:10 am

Re: Realism, Gameplay, Aesthetics and Canon

Post by Neskiairti »

if were talking leggos vs nuts and bolts, id agree. you cant fit those child leggo blocks (the huge ones) on the normal ones.
but to me, 'Frankenstein' attachments involves making sure the power systems match.. then simply 'gluing' mounting plates together for the two different sizes and attaching.. then running the wiring through it and if needed, the munitions loading mechanism. I know enough welding to do that :P

but i think your overlooking another factor, why the hell would weapons (unless they were some special capital weapon) be so much larger than others. take naval warfare for instance. at the top end you have 16 inch guns which aren't really even used anymore.. this goes all the way down to tiny (ish) machine guns. most guns these days, stay within a size range to be able to be fitted on just about any ship if the need arises.. this is not entirely true, 16 inch guns as an example.. just wouldn't work.. take the Mexican patrol boats.. they are putting up to 127mm guns on the damn things to be effective.

as for artists.. I still think the mounts should be defined on the mesh, but the guns should be modeled separately. none of the ships should come with pre-built weapons unless the ship is designed around the weapon.

also, size does not necessarily equate energy usage nor even impact power. lasers are a good example of power, the larger you get, generally the weaker the beam. and if you are using a gas propelled kinetic, that likely will use less energy than any other gun, mostly just aiming and firing mechanism

if you want size, just make sure the mass of the gun is say less than 10% the mass of the ship :P
but thats all I've got to say, i got my own work to do!
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

Neskiairti wrote:if were talking leggos vs nuts and bolts, id agree. you cant fit those child leggo blocks (the huge ones) on the normal ones.
but to me, 'Frankenstein' attachments involves making sure the power systems match.. then simply 'gluing' mounting plates together for the two different sizes and attaching.. then running the wiring through it and if needed, the munitions loading mechanism. I know enough welding to do that :P
These are still major modifications, to achieve impractical results. If every fighter barracks would do this kind of work for free and quickly, to put on one massive weapon your ship can't handle, why not also be able to put on 50 smaller weapons that your ship can't practically handle altogether? So then we should eliminate both size classes and the number/type limitations of hardpoints on ships, if following this explanation.

The manufacturer-unintended hardpoint modifications you're talking about might be something you could purchase at a pirate base or such, but to assume they are done everywhere so that you can mount any weapon to any ship doesn't seem very believable IMO, nor does it create more gameplay benefits than it demolishes.
but i think your overlooking another factor, why the hell would weapons (unless they were some special capital weapon) be so much larger than others. take naval warfare for instance. at the top end you have 16 inch guns which aren't really even used anymore.. this goes all the way down to tiny (ish) machine guns. most guns these days, stay within a size range to be able to be fitted on just about any ship if the need arises.. this is not entirely true, 16 inch guns as an example.. just wouldn't work.. take the Mexican patrol boats.. they are putting up to 127mm guns on the damn things to be effective.
Because most or all modern naval warships don't even have real guns anymore, just an anti-aircraft gun taking up the one forward turret. A patrol boat larger than a tank, using a tank's gun also isn't much evidence that size makes so little difference. And then in VS you have even bigger scale differences than you see in modern real life 'wet navies'.
as for artists.. I still think the mounts should be defined on the mesh, but the guns should be modeled separately. none of the ships should come with pre-built weapons unless the ship is designed around the weapon.
That can be done, but we still need to know the approximate sizes of these weapons.
Neskiairti
Confed Special Operative
Confed Special Operative
Posts: 334
Joined: Thu Jan 11, 2007 4:10 am

Re: Realism, Gameplay, Aesthetics and Canon

Post by Neskiairti »

well I said I was done.. I lied :p

i never said do away with hard points.. but im not sure you really grasp what a hard point is.. its a point on the hull that is structurally sound, a place where you can put exterior equipment without worry about thrust wrenching it and a chunk of your hull off.. its not some specialty equipment, its simply a point of structure.
and whether the station crew will do it or not...? they just sold a really expensive weapon to a chump.. :p

Also, just because the size of a ship increases, doesnt mean the size of the gun increases. that would be foolish, really. The bigger the gun the easier the target. And you only need a gun large enough to project your payload, whatever that is. A laser can be a single inch wide dot on a hull the size of a planet.. if your lasing equipment is all internal. and that laser, if backed up by enough power and technology, could punch a hole through the crust of a moon.

Big guns have gone the way of the 16 inch, impractical. just because your in space and your vehicles are even larger, doesn't change this. what you DO need to mount and worry about, is heatsinks. as chuck says, heat is the biggest problem in space if you want to restrict weapons so much, do so with heat.. keep guns small and just base them on how much heat they produce.
Deus Siddis
Elite
Elite
Posts: 1363
Joined: Sat Aug 04, 2007 3:42 pm

Re: Realism, Gameplay, Aesthetics and Canon

Post by Deus Siddis »

Neskiairti wrote: i never said do away with hard points.. but im not sure you really grasp what a hard point is.. its a point on the hull that is structurally sound, a place where you can put exterior equipment without worry about thrust wrenching it and a chunk of your hull off..
Infinite thrust you mean? Because that's kind of what you're talking about if there's no limitations on what size, mass, recoil that hardpoint can take. And then shipbuilders like the Aera can only uses weapons small enough for their fighters, since they mount all their (at least fixed) weapons inside their hulls; making size a major limiting factor.
its not some specialty equipment, its simply a point of structure.
It currently is in VS- you can't load missiles where a tractor beam goes or a tractor beam where a standard gun type goes. Removing this would also be a big change that would need to be weighed carefully.
and whether the station crew will do it or not...? they just sold a really expensive weapon to a chump.. :p
They didn't just sell you something, they practically rebuilt your ship into a new design. That's very different from just swapping out standard parts that are designed to work together.
Also, just because the size of a ship increases, doesnt mean the size of the gun increases. that would be foolish, really.
It isn't automatic, but there is a correlation. Larger weapons can generally have more range, more penetration power, etc. There's natural advantages as well as disadvantages to using larger weapons, where possible. So it seems very artificial to say that 600 meter long ships can't use any individual weapon system larger than what can be effectively utilized by a 10 meter long ship.
The bigger the gun the easier the target.
And more easily armored, higher velocity, higher armor penetration, possibly longer effective range. . .
And you only need a gun large enough to project your payload, whatever that is. A laser can be a single inch wide dot on a hull the size of a planet.. if your lasing equipment is all internal. and that laser, if backed up by enough power and technology, could punch a hole through the crust of a moon.
And your tiny laser doesn't just melt from that kind of load? I don't know that much about weaponized lasers, but I thought the idea was they focused onto a small point on the target, not that they themselves had to be very small to create a concentrated beam.

Either way lasers represent a minority of the weapons used in VS.
Big guns have gone the way of the 16 inch, impractical. just because your in space and your vehicles are even larger, doesn't change this. what you DO need to mount and worry about, is heatsinks. as chuck says, heat is the biggest problem in space if you want to restrict weapons so much, do so with heat.. keep guns small and just base them on how much heat they produce.
Actually big guns were replaced by big missiles, a trend which someday soon might start to reverse with the usage of railguns. Heat could work as a major limiting factor, but only for certain weapons- missiles being the prime exception and also one that is physically large.
Post Reply