Paid programmer work

[toneddu2000]

If you are making a totally new game though, especially new levels, it could be a good idea. It would help to know exactly what your overall goals are.

You reached the point Sajt. I 'm trying to create a brand new project, not a quake style clone. But I used the word "project" and not "game" properly, because I'm not so interested in which monsters or weapons could be inserted in the game or which ambientation to choose. I'd like , to use a parallel, to replicate the Blender Foundation path. YoFrankie (the open game made by BF)is almost unplayable on the 90% of the computers, but it gave to the community a brand new version of Blender, tons of documentation, tutorials, code snippets, and most of all, a new Open Platform for gaming, that could be used also in production studios. THIS is what I'd like to be part of. Darkplaces is a Great engine, and what LordHavoc and Urre demonstrate with Twig (correct me if I'm wrong)is that is very flexible.
So it could be "relative" easy for expert programmers, to insert small "modules" in quake c that expand its power. Anyway, in these last two days, I thought a lot about the "paid work" and I realized that:
1 because the budget, as I said to Swift, is VERY VERY low (about $300) and I don't want to offend anyone's work (because I know that this amount is at maximum a 1/4 of a normal programmer monthly salary), and

2because I know that to implement those changes it would be a very time-consuming work.

I realized that it could be better, if you agree, to start a mini challenge with three prizes (1°,2° and third classified). The three winners, judged by the community, that create a work that follows the guide lines of the requirements, will win different prizes (in money of course). What do you think, is a valid solution or not? Thank you for your attention!

[Sajt]

The biggest problem is that the features you requested are very vague and, even worse, totally open-ended. Each of them could be a year's work, or a weekend's work, we can't really tell until you are more specific about what you want. This is what drives programmers in the game industry crazy, bosses who don't even know what they want, who can't be specific.

For example, you want a Doom3-style in-world "consoles" with dialogues and whatnot. You should state first what you want to do with these, what they should be able to do. Otherwise the programmer will not know where to start, and worse, where to stop!

[toneddu2000]

This is what drives programmers in the game industry crazy, bosses who don't even know what they want, who can't be specific

Yeah, I have the same problem at work!

For example, you want a Doom3-style in-world "consoles" with dialogues and whatnot. You should state first what you want to do with these, what they should be able to do. Otherwise the programmer will not know where to start, and worse, where to stop!

Well, to be honest, my ignorance about programming couldn't permit myself to be clear about the complexity scale of the features I have listed. I'd rather like (and please believe me) that would be the community of inside3d, which has lots of programmers and quake c experts, to tell which feature is possible to implement, which one is too difficult or too time consuming, etc.. and, after all the discussions, to find 3 mini-features or 2 or just 1 (I don't know) that could be the task of the challenge. IMHO a feature that most of others missed a lot in these years is real physics. And, with physics I mean rigid body, ragdolls skeleton system, interaction between skeleton and environment. To explain this last feature I'll make an example: three years ago I used FarCry 1 sandbox for testing purposes. I took a character, I put on the scene (about 1,5 meters above terrain - no contact), and when I pressed "play" button, the character felt down on the ground, and "bent" its spine as it was really felt from a big height!Amazing. So, please be free, anyone to express your opinion about the features we spoke. And, if it's an impossible mission to do.. well, we tried! .

PS if you need a more technical and detailed feature list I'll make it in the next days, copying from other game engines sites

[Sajt]

Well, in the case of physics specifically, somebody or other has apparently been adding it into DarkPlaces. I don't know what the status of it is, or if that person is still working on it, though.

As for other stuff such as Doom3 consoles, it would be cool to program something like it, but it may likely end up that is never used, and that coding it in the first place was a waste of time. Some modder needs to make a mod which uses these consoles. If there were such a person, who was making a total conversion of Quake for which in-world consoles (as seen in SiN or Doom3) are a major feature, then maybe we can get down to business. Otherwise, most of us have more important things to do. When I was younger, I would often code elaborate systems for fun, before I had an idea of what they would be used for, and I'd end up never making good use of it. That was fun then, but now I notice with growing horror the passage of years with no final products really "accomplished".

Then again, the guys on the Nexuiz side of the fence seem to be implementing everything under the sun, even not related to Nexuiz. So, who knows, maybe they are already coding these features.

[toneddu2000]

When I said "Add integration for cameras into Darkplaces" I refer to create an entity "camera" like the doom3 md5.camera object. When in radiant put a camera trigger, once touched, the first person view is switched to an external camera view and it's very useful for "story ingame video".
According to physics in Nexuiz, well, about three months ago alientrap starts inserting ODE in darkplaces engine, but it has been quite soon abandoned.

That was fun then, but now I notice with growing horror the passage of years with no final products really "accomplished"

Completely agree. If I think abot my old projects..

[Sajt]

Okay. People have been doing "cutscene" camera entities for mappers since day one in QuakeC. But you mentioned dialogues and whatnot, so I thought you wanted the camera view to be mapped onto a texture, like a security camera or the TVs in Doom3 or something.

[Urre]

That's also already been added to DP. See examples here and here.

As far as I've understood, ODE works okay in DP. I've been fantasiscing about hooking it up with Twig, to make stuff like boxes with attached cylinders (cars!) or ragdolls. Haven't gotten around to it yet.

[Spike]

While ODE can work in DP, its really not set up for ragdoll.
Additionally, for me at least, it appears to get miscompiled and suddenly throws boxes around the room at infinite velocities.
For me, DP+ODE just doesn't seem robust. Maybe its just msvc.

Regarding cutscenes, yeah, a few mods have those. Zerstorer is the first one I remember that had them. Its open source too.
The thing to remember though is that while making a camera entity trigger, grab the player's view, and follow a track, is easy, what is not easy is providing the functionality to get the monsters to interact in an apropriate way, which is imho more important from a narrative perspective. If you just want to show bridges opening or whatever, showing it through a window like in Q2 (or a duke3d-style security camera) is perhaps more appropriate, and certainly less annoying - it also requires no extra gamecode to do it.
Subtitles could be done with csqc, but then so could the entire camera. :)

[Urre]

While ODE can work in DP, its really not set up for ragdoll.
Additionally, for me at least, it appears to get miscompiled and suddenly throws boxes around the room at infinite velocities.
For me, DP+ODE just doesn't seem robust. Maybe its just msvc.

Regarding ragdolls, shouldn't it be possible to let Twig handle the joints?

[frag.machine]

While ODE can work in DP, its really not set up for ragdoll.
Additionally, for me at least, it appears to get miscompiled and suddenly throws boxes around the room at infinite velocities.
For me, DP+ODE just doesn't seem robust. Maybe its just msvc.

Actually this behavior is observed in other engines using ODE. This and better performance has lead some developers to use Bullet instead.

[Spike]

Regarding ragdolls, shouldn't it be possible to let Twig handle the joints?

Sure, in which case you're not using ODE.
There's supposedly skeletal bone control already, so you'd need to plug twig into that system. Basically each time a skeleton changes, you need to update the skeleton to match, as well as update the joints when the animation changes.
The challenge is updating the animations in such a way that your joints are never broken, and retaining individual control over certain limbs, so the toon pitches up and down and twists at the hip when running perpendicular to his target while still being unable to put his leg through uneven ground surfaces/walls. Its fun. :)

[toneddu2000]

I didn't see either a good stability in Ode + Dp. Infact alientrap practically discontinued development after two months. Maybe it could be possible to do somenthing in Twig, as Spike mentioned, maybe the competition's title could be:"What can you do, Twig BoyZ?":)

By the way, I downloaded and played Urre's gammacar test. It's VEEERY promising: control system is a little disorienting, but the fact you can drive a physics car in DP is wonderful!!

@Spike:thx for the hint about Zerstorer, I'll take a look at the source, I've just downloaded.

Anyway, take a look at this http://svn.icculus.org/twilight/trunk/phystest/ (in windows just execute lhragdoll.exe, in linux cd phystest - make - ./phystest)
If I think that LordHavoc did 2 years ago.. He's a genius! Question: is this useless "combined" to Twig?

[Urre]

I almost forgot, DP also supports Doom3-like interfaces on surfaces using Gecko. It takes it a step further even, you could read this forum from inside a terminal in-game in DP.

lhragdoll is based on the same principles as Twig, except written in C, so it's automaticly superior. I recall LH discontinued that because it wasn't robust enough, particle-based just didn't do the trick by todays standards.

Twig is very limited in capability. It started out mostly as a challenge, to see if it was possible to make a physics-engine in QuakeC.

[toneddu2000]

Well, I've to say thay my idea stops here. I want to thank you all the users gave me precious hints. I'll start a mini game with Unreal Development Kit or a mini game with DP as is. I've to think a little about it!
cheers

[toneddu2000]

I've a little last question for Urre: can you please explain how can you make such a big map for twig, bigtest.bsp? I tried in the last two days to create a very big map like yours(64000x64000x32000 quake units) but q3map2(not Netradiant, which remains opened) crashes when tries to compile it(precisely at the end of FloodEntities) and I've a brand new computer! I also tried to open your .map file but it says:

Code: Select all

Open file D:/games/test1/twig/maps/bigtest.map for read...success
Texture load failed: "textures/"
Texture load failed: "textures/main/sky1"
19:2: parse error at '(': expected '#integer'
brush 0: parse error
entity 0: parse error

How can I do? I also wonder why lighting in quake is different from lighting in nexuiz. Using NetRadiant nexuiz creates a folder with lightmaps inside, instead quake doesn't create anything, and if I don't create a .rtlights file with the origin of the light, the scene is completely unlit! It seems that quake (with Darkplaces engine of course)doesn't calculate bouncing light maps. Or I'm wrong? Thx in advance!!