WCU 0.3 TODO

Discuss the Wing Commander Series and find the latest information on the Wing Commander Universe privateer mod as well as the standalone mod Wasteland Incident project.
charlieg
Elite Mercenary
Elite Mercenary
Posts: 1329
Joined: Thu Mar 27, 2003 11:51 pm
Location: Manchester, UK
Contact:

WCU 0.3 TODO

Post by charlieg »

What should be in the next WCU release? It should be sooner rather than later, and goals should be realistic (i.e. either small or almost complete or can be made usable quickly) and stability shouldn't be awful but perfection not needed ("dude, it's a development release").

Ideas:
  • New models with (multiple?) textures: Demon, Hornet, Hellcat
  • Other preliminary WC1 content (what is on the radar?)
  • Initial OGRE-based VS engine (what more needs doing, klauss?)
What else? Let's get some focus into this project!
Privateer Ferrius
Venturer
Venturer
Posts: 538
Joined: Tue Mar 21, 2006 12:54 am
Location: Delta Prime
Contact:

Post by Privateer Ferrius »

The Bengal would also be nice :)
Cmdr. Micheal Ferrius (Ret.)
Retired Confederate Pilot

For Hire!
Post jobs to the BBS!

Tengoku de omachi shite imasu
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Post by chuck_starchaser »

Yeah, after I unwrap the Hornet, I'll get back to the Bengal. I agree, there can be no WC1 without it.

Thanks for starting this thread, charlieg; very timely, as Spirit should be back this week. I wish Fireskull was still around; all the WC1 systems need to be scripted, rather than autogenerated.
I don't think Klauss' work on the Ogre integration can be rushed. What I was suggesting to him is a modellers' version of his "Explorer", so that modellers and texturers can start fine-tuning their work for the new engine. I'd be willing to put in some work on that, just don't know where to start. I suppose I should download TortoiseSVN and see if I can compile Explorer...
I'm also hoping to convince Ferrius to move character modelling to the top of his to-do list.
Yeah, I think we should be working specifically towards a WC1 Remake; if we just keep fooling around with Privateer, it becomes just debugging work that's never finished. I think a departure from the norm would inject fresh blood into WCU.
Privateer Ferrius
Venturer
Venturer
Posts: 538
Joined: Tue Mar 21, 2006 12:54 am
Location: Delta Prime
Contact:

Post by Privateer Ferrius »

Only after the Demon and Hornet are done, Chuck. Im experienced with modelling humans, but I cut my teeth doing textures, so those are where it's at for me in terms of the greatest experience :)
Cmdr. Micheal Ferrius (Ret.)
Retired Confederate Pilot

For Hire!
Post jobs to the BBS!

Tengoku de omachi shite imasu
charlieg
Elite Mercenary
Elite Mercenary
Posts: 1329
Joined: Thu Mar 27, 2003 11:51 pm
Location: Manchester, UK
Contact:

Post by charlieg »

I don't want to rush klauss, per se, but a bit of pressure can add a bit of focus. Not that I'm accusing klauss is doing this, but it can be easy to lose sight of the short term goals in pursuit of the long term dream i.e. get done what really needs to be done, whereas in a "when it's ready" atmosphere people tend to be rather cavalier or idealistic in their planning - "Oh, that's not the perfect way to do it, let's do it this way!"

Lofty aspirations are admirable, but they need to be tempered with short term progress-oriented goals in order to be even partially acheivable.

I've been about FOSS a long time now and I've seen hundreds of projects spawn out of dreamy desire and die horribly languid deaths in a pool of red, red realism. And it can happen to even the most popular of projects...

If you have regular, playable releases then new blood has a fresh starting point for involvement. That's probably, in my experience, the single most vital thing for the survival and nourishment of any open source project.
Privateer Ferrius
Venturer
Venturer
Posts: 538
Joined: Tue Mar 21, 2006 12:54 am
Location: Delta Prime
Contact:

Post by Privateer Ferrius »

I could not agree with you any more, Charlie :D
Cmdr. Micheal Ferrius (Ret.)
Retired Confederate Pilot

For Hire!
Post jobs to the BBS!

Tengoku de omachi shite imasu
mkruer
Site Administrator
Site Administrator
Posts: 1089
Joined: Thu Jan 02, 2003 10:07 am
Contact:

Post by mkruer »

@charlieg + everyone else.
Read the Roadmap guys. Its been there for a few months now. You can check the history, most of the edits were me trying to reformat the page. :LOL: But the content is the same.

http://vegastrike.sourceforge.net/wiki/MOD:WCU:Roadmap

@Privateer Ferrius
If it meets or beats the current models sepcs, then it probably will.
I know you believe you understand what you think I said.
But I am not sure you realize that what you heard is not what I meant.

Wing Commander Universe Forum | Wiki
Wing Commander: The Wasteland Incident
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Post by chuck_starchaser »

I think charlieg was trying to inspire us to set some short-term goals, like to put a tentative date on them, and run for it.
klauss
Elite
Elite
Posts: 7243
Joined: Mon Apr 18, 2005 2:40 pm
Location: LS87, Buenos Aires, República Argentina

Post by klauss »

charlieg wrote:I don't want to rush klauss, per se, but a bit of pressure can add a bit of focus.
Totally. Feel free to put pressure on me.
I won't hesitate to ignore it when nothing can be done about it, but when it's not hopeless, it tends to make me work.

Matthew: I think Charlieg here meant an even shorter term than the roadmap. I'd suggest focusing on quality - right now, we have too many features but too loosely attached. We have to start working on those features to make them integrate better.

So... a next release could focus on "playability". Not bugfixes per-se, but it obviously includes them. I mean taking the hackishness out of some features. Balancing looks and performance (flak guns, remember?), stuff like that - stuff that kills gameplay.

When an issue is identified, and a possible workaround found, I would suggest adding it to a tacker in SF's page. Either the bug tracker, or the task tracker, depending on the kind of issue.

How does it sound?
Oíd mortales, el grito sagrado...
Call me "Menes, lord of Cats"
Wing Commander Universe
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Post by chuck_starchaser »

Task tracker says "no subprojects defined", whatever that means. The bugtracker... Not sure... It is definitely the "proper" way to go, but my question is, are people going to use it? I mean, I've used bug trackers before, and it takes a bit of work to fill out the form, and it doesn't provide a means for "venting", like when people report a bug, sometimes they feel angry and want to call everyone in the project names, which is okay, we get the bug report and that's what matters; but if you're full of anger you can't fill out a form.. :) so we miss the bug report :(
And the problem with the sticky in the forum is that people naturally skip stickies. At least I do. If you ask me what are the titles or topics of the stickies in wcu, I ONLY remember the bug reports sticky because I just looked at it today, and the database project, same reason. But I know there's another dozen stickies there, and no idea what they are.
Not sure what the solution is; maybe one of us could take charge of shoving bug report posts into a wiki page. I'd be glad to take on the task if the idea is liked; tho I might miss some, from time to time.
charlieg
Elite Mercenary
Elite Mercenary
Posts: 1329
Joined: Thu Mar 27, 2003 11:51 pm
Location: Manchester, UK
Contact:

Post by charlieg »

klauss wrote:Matthew: I think Charlieg here meant an even shorter term than the roadmap. I'd suggest focusing on quality - right now, we have too many features but too loosely attached. We have to start working on those features to make them integrate better.

So... a next release could focus on "playability". Not bugfixes per-se, but it obviously includes them. I mean taking the hackishness out of some features. Balancing looks and performance (flak guns, remember?), stuff like that - stuff that kills gameplay.
Bingo.
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Post by chuck_starchaser »

Klauss, with respect to the Ogre thing. First of all apollogies due for being opinionated without getting my hands dirty... What I think would pay off is to prioritize the features list. I would NOT advocate a "minimum" list to start with, since I know you like to shoot for the stars in everything you do, and so do I, and I know that doing so often helps the process. My prioritization would be in terms of what's needed for "correctness", and what's an extra; though I'll offer a rather loose definition of correctness.

Correctness:
Basically, those features that will enable a texturer to know what a texturing job will look like in the final engine.

Examples of features needed for correctness:
* Diffuse and specular lighting formulas guaranteed not to change in future releases.
* Ambient lighting that agrees with background (Long ago, I was suggesting that the code could read the lowest mipmap (single pixel) color for all 6 sides of the cube-map and average them together, to obtain a figure for ambient light/color; --rather than read it from an xml file and risk that the figure might be wrong. Same thing for the color of the star, and direct lighting.)
* Environment mapping whose level is optically consistent with ambient lighting and specularity equations, and guaranteed not to change. Also, shouldn't the specularity of most materials change with angle? This should be trivial to add. Like, if I look at this printed circuit board I have here on my desk, it looks not terribly specular, mostly green, at 45 degree reflection; but at a shallow angle I can almost read the text on my screen reflecting on it... Just a thought.
* Shadows; whether stencil or depthmapped; without shadows things look too unnatural to be able to judge with any certainty whether a texturing job looks okay or not, so I'd throw shadows in.
* Bumpmapping/normal mapping: We need them now, so that texturers can see what their bumpmaps look like.
* Detail textures and detail normal maps. I'd say that, to get people to use them, best way would be to have the engine use a generic default, if none is provided/specified.
* Faction decals.
* Basic HDR, as pertains to dynamic range only.

Things I'd deem NOT needed for "correctness":
* Fancy HDR glows and lense flares and all that.
* Self-shadowing, self-occlusion and texture offset bumpmapping. Better NOT to have them for now, I'd say, or modellers might go bananas and start abusing bumpmapping from the start.
* Self-radiosity, other than baked-in; global illumination.
* Shadows produced by one unit being received by another. Only a unit's self-shadowing needed, to judge the correctness of a model.
* Compatibility with lesser hardware: Let would-be modellers and texturers make sure they have the hardware they need to get the job done; ps1.2 compatibility is nice for the final thing for release, but people involved in development should sell their kids (or parents) if that's what it takes to get the proper tools.

Of course, if adding on some feature later would require a redesign, might as well do it up front. Possibly the global illumination thing falls in this category; but it could be like a minimal version of it for now.

As a further simplification, where self-occlusion, self-shadowing and texture offset bumpmaps are considered, you might as well NOT consider many of the other features, such as HDR, global illumination, or environment mapping. Best use of fancy bumpmaps would be for asteroids and planetary terrains, which need basically nothing else. For most ships, standard normal maps should suffice. (The way I'm thinking, normal maps would have higher rez than the diffuse texture; so, if so, texture offset for ships would be unwarranted. So these might as well be different shaders.
klauss
Elite
Elite
Posts: 7243
Joined: Mon Apr 18, 2005 2:40 pm
Location: LS87, Buenos Aires, República Argentina

Post by klauss »

chuck_starchaser wrote:Of course, if adding on some feature later would require a redesign, might as well do it up front. Possibly the global illumination thing falls in this category; but it could be like a minimal version of it for now.
I think one of the most importants features in that category is compatibility with lesser hardware. I'm not entirely sure, but I think you can get in dark alleys if you forget about them (and intend to add support later on).

As for the rest, I agree. But you have a killer there: shadows.
chuck_starchaser wrote:* Shadows produced by one unit being received by another. Only a unit's self-shadowing needed, to judge the correctness of a model.
I imagine you said so because you imagine this would make things easier to get it working before getting it perfect. But it doesn't, actually. In fact, treating shadows anything but globally makes your rendering loop very segmented (like one ship at a time), which hinders performance a great deal. Efficient implementations need a "globalized shadow implementation", as in "providing shadows for the entire scene" (or at least a sizable portion).
I still can't see a good algorithm for this.
I think this is one of the features I'd leave for later, simply because I have no clue as to how to implement them. If it needs a mild redesign (it won't be a deep redesign, that's improbable), so be it.
Oíd mortales, el grito sagrado...
Call me "Menes, lord of Cats"
Wing Commander Universe
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Post by chuck_starchaser »

How about a slight modification of the initial plan we were discussing by email?, namely: Specify that a ship's (only) lesser LOD must be a closed solid, and use that with stencil shadows for inter-ship shadows; and then use individually optimized z range depth-map shadows for intra-ship (self-) shadowing. May not be as elegant as a global solution; but, heck, if a global solution don't exist, cans it not be AND be elegant ? :D
klauss
Elite
Elite
Posts: 7243
Joined: Mon Apr 18, 2005 2:40 pm
Location: LS87, Buenos Aires, República Argentina

Post by klauss »

I've been thinking (and benchmarking) about that, and stencil shadows are simply not well suited for this kind of scenary (space) - overarching shadow volumes that intersect no geometry are all too common, high or low detail, and it produces nasty peaks on fill rates.

Plus, doing self shadows with textures has the drawback of requiring per-ship set ups, which subdivides scene rendering in way too many steps.

No... performance-wise, that's a mess (even if results are, or could be, rather nice).
Oíd mortales, el grito sagrado...
Call me "Menes, lord of Cats"
Wing Commander Universe
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Post by chuck_starchaser »

Doh!
Okay, in that case I'd say let's forget shadows; maybe even wait for the hardware manufactures to solve the problem... Actually, I'm having another idea; --need to think about it some more; I'll email you about it...
charlieg
Elite Mercenary
Elite Mercenary
Posts: 1329
Joined: Thu Mar 27, 2003 11:51 pm
Location: Manchester, UK
Contact:

Post by charlieg »

With the OGRE port, how's about taking an even simpler look at it.

Minimum that needs doing: replacement of the old engine.

Anything else is a bonus and should only be included if it actually works and is stable. Sure, shoot for the stars, but do it step by step. First step is to get a working OGRE-based renderer. Fancy new features can come later!

So, in that vein, what needs doing before the OGRE port of VS can be considered a complete replacement of VS 0.4.x?
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Post by chuck_starchaser »

What I was suggesting to Klauss --forgive if I repeat myself; I just can't remember whether I said it before in this particular thread-- for a short term goal, to come up with a functioning ship visualizer, rather than a full engine integration. I believe Klauss' Explorer is almost there, and that all it needs added to it is a command that reloads the mesh and textures, so one doesn't have to close down and restart the application every time. The advantage of this is that we could start making versions of the ships for the new engine. Now, I may be the one causing delay here: Klauss asked me for a ship with full textures, and I don't have one. I was hoping to give him the Demon, but that's kind of uncertain where it is at. I might just have to come up with something quick.
Anyways, while we modellers and texturers then work on converting ships for the new engine, Klauss could take his time finalizing engine integration. So, this is just my idea for how we could parallelize the pipeline...
Privateer Ferrius
Venturer
Venturer
Posts: 538
Joined: Tue Mar 21, 2006 12:54 am
Location: Delta Prime
Contact:

Post by Privateer Ferrius »

Well, the ship is fully textured; the texture is just not complete. That is to say, all the detail I want to be there, is not. I DO have some significat issues making believable metal plates on the curved sections due to the distortion however.

Perhaps what Klauss could do is say, what he wants to have for the next release, and what needs done before that; the prerequisites.
Cmdr. Micheal Ferrius (Ret.)
Retired Confederate Pilot

For Hire!
Post jobs to the BBS!

Tengoku de omachi shite imasu
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Post by chuck_starchaser »

Well, my understanding is that Klauss needs a fully textured ship like... now, to finetune his shaders with. With radiosity baking and all. Or was it without? I forget now...
Privateer Ferrius
Venturer
Venturer
Posts: 538
Joined: Tue Mar 21, 2006 12:54 am
Location: Delta Prime
Contact:

Post by Privateer Ferrius »

Well worse come to worse I still have the old retextured Demon, as horrible as the mesh is.
Cmdr. Micheal Ferrius (Ret.)
Retired Confederate Pilot

For Hire!
Post jobs to the BBS!

Tengoku de omachi shite imasu
chuck_starchaser
Elite
Elite
Posts: 8014
Joined: Fri Sep 05, 2003 4:03 am
Location: Montreal
Contact:

Post by chuck_starchaser »

Can you post a link, again? I never downloaded them since I don't even know where I might have saved the mesh of the original demon, if I did at all; and WCU's lighting wasn't working anyways; but I might just download PR and check it out. If it's good, I guess I could generate a radiosity baking to go with it.
Privateer Ferrius
Venturer
Venturer
Posts: 538
Joined: Tue Mar 21, 2006 12:54 am
Location: Delta Prime
Contact:

Post by Privateer Ferrius »

Hold on; I think I have it, but I might have saved the new demon on top of it, let me check..

[edit]: Dammit, I did. I know i have a backup archive somewheres, but you have no idea how many achives I have with demon in the name I kept that were posted in links. Lemme dig for a moment...

[edit2]: Here it is: http://lfs.lfhost.com/wcu/refurbished_demon.rar

I didn't think it was anything special, but people seemed to like it and it was better than the existing texture anyways :P Most of what I got out of thsi myself was what not to do on my new one (the tiny plates, for example)
Cmdr. Micheal Ferrius (Ret.)
Retired Confederate Pilot

For Hire!
Post jobs to the BBS!

Tengoku de omachi shite imasu
mkruer
Site Administrator
Site Administrator
Posts: 1089
Joined: Thu Jan 02, 2003 10:07 am
Contact:

Post by mkruer »

charlieg wrote:
klauss wrote:Matthew: I think Charlieg here meant an even shorter term than the roadmap. I'd suggest focusing on quality - right now, we have too many features but too loosely attached. We have to start working on those features to make them integrate better.

So... a next release could focus on "playability". Not bugfixes per-se, but it obviously includes them. I mean taking the hackishness out of some features. Balancing looks and performance (flak guns, remember?), stuff like that - stuff that kills gameplay.
Bingo.
That’s Fine, but that’s not a 0.3.x release. That’s still in the 0.2.x existing branch. What was originally specified was:
charlieg wrote:What should be in the next WCU release? It should be sooner rather than later, and goals should be realistic (i.e. either small or almost complete or can be made usable quickly) and stability shouldn't be awful but perfection not needed ("dude, it's a development release").
This by its self says it’s a 0.2.x release.
charlieg wrote:Ideas:
  • New models with (multiple?) textures: Demon, Hornet, Hellcat
  • Other preliminary WC1 content (what is on the radar?)
Again these would be 0.2.x branch, and feel free to up this number as frequently as possible. You might want to check with Spirit to see what the last release was at, 0.2.4 I think.
charlieg wrote:
  • Initial OGRE-based VS engine (what more needs doing, klauss?)
This is a major change to the system and gets the 0.3.x treatment or something of equal weight. That’s all I am saying. Hell if you can release a 0.2.x update say once a month, then that, overtime it will warrant a 0.3.x release. That’s all I am saying.

This thread should be labeled 0.2.x > 0.3.x TODO list, which should be added to the wiki instead of this thread anyway.
Last edited by mkruer on Wed Apr 12, 2006 1:38 am, edited 1 time in total.
I know you believe you understand what you think I said.
But I am not sure you realize that what you heard is not what I meant.

Wing Commander Universe Forum | Wiki
Wing Commander: The Wasteland Incident
charlieg
Elite Mercenary
Elite Mercenary
Posts: 1329
Joined: Thu Mar 27, 2003 11:51 pm
Location: Manchester, UK
Contact:

Post by charlieg »

See, there are people who can organise this stuff. *nudges mkruer*
Post Reply