Advanced Open Source Engine Based On Quake 3

An anonymous reader writes "Phoronix is running a news story about the XreaL project, which its lead developer claims is the most advanced open-source game engine. XreaL is based upon the vintage Quake 3 engine, but it has been rewritten over the course of many months such that it no longer resembles the original id Software engine. The XreaL engine has its renderer written entirely in GLSL with compliance toward the OpenGL ES 2.0 specification in mind, but it supports the new OpenGL 3.0/3.1 specification and is able to take advantage of its new features. XreaL has also added an HDR pipeline to its engine and on modern hardware is actually GPU — not CPU — bottlenecked. XreaL can also load game content from Unreal Tournament 3. This engine, which is described to be as powerful as what can be found in Doom 3 or Call of Duty 4, is written entirely with free software. The XreaL project has created plug-ins for Maya to broaden their game development capabilities."

Comments

'bout time

[tannsi]

The pure potential is awesome. If, however they are uptight about letting people develop non-open-source games for this it will fail, hard.

great

[Turiko]

hopefully this will lead to more modern-looking open-source games. That's the reason the regular gamer won't play open-source. Unless there's somethign else i nthe game you can't find anywhere else :P.

Re:great

[Opportunist]

What's funny is that it's all there. You have a ton of people who love to be creative, draw and model. Various pages dedicated to displaying your art are proof enough. You have equally many people who enjoy coding. Linux alone works as a proof for that, together with the lot of FOSS that's readily available and of high quality.

Yet for some odd reason they don't "find" each other. FOSS games are usually mediocre. And I wonder why. FOSS software does not have to hide behind commercial software. Hell, some is better than its commercial counterparts! "Free" art (from music to animations and even movies) is by no means worse than commercial art. Especially true when it comes to music IMO.

Why does the combination not work out?

Re:great

[bit01]

Why does the combination not work out?

C. P. Snow's The Two Cultures. Each needs to make overtures to the other.

Re:great

[ardor]

I know some artists who tried this out, only to find themselves in a failing project.

The way to go is to create some prototype with placeholders as game art. Something that can be demonstrated to artists. This attracts a lot more people than a paper with fancy ideas. However, guess what - most of the time, the paper with fancy ideas is presented. No wonder the experienced artists stay away (unless they get paid).

Also, most projects don't have something resembling a lead designer; instead, one of the programmers is the lead. This is a bad idea, since the designer is the one who cares most about the "big picture", the overall design. The lead designer is the one who takes care of keeping things together and coherent. This is a full-time job, and often underestimated, especially by open-source game projects.

Re:great

[Tenebrousedge]

The reason is that doing a complete set of artwork for a game is hard, and extremely time consuming. Most people, if they have the time, skills, and interest, will join a mod project, rather than develop something completely new from the ground up. Most of those mod projects subsequently amount to nothing due to poor interpersonal communications, inability to meet deadlines, real life getting in the way, etc.

So for a FOSS game artist, you're asking that a person be talented, dedicated, able to meet deadlines, not interested in the mod scene, technically adept (probably), good at working with others, and willing to work for free.

Then you have to find the same thing in a half dozen other people, some on sound, some on levels, environment, character models, etc.

Making a video game is a tremendous undertaking these days. Anyone capable of making a good game for free probably shouldn't sell themselves for that little.

Re:great

[Haeleth]

hopefully this will lead to more modern-looking open-source games. That's the reason the regular gamer won't play open-source.

No, that would be because open-source games tend to have poor artwork, horrible animation, tediously derivative gameplay and level design, and no plot.

For some reason, people with talent in these areas and an interest in giving their work away for free tend to concentrate on making mods for commercial games. Perhaps it's because they want their work to be free-as-in-beer rather than free-as-in-copyleft.

Re:great

[drinkypoo]

The simple truth is that trying to build a FoSS game based on a commercial model is stupid, unnecessary, and in any case bound to fail anyway. The great thing about an Open game is that it can be worked on for fucking ages and still come out to be positive. A good game is always good whether it is the latest and greatest or not.

Re:great

[Peganthyrus]

Artists also act like this because of the way so many people shamelessly swipe their images against their wishes - there are a lot of people who ruin everything they post online with huge ugly watermarks because they'd rather that than find it reposted to some kid's Deviantart account as "theirs".

Also, the culture of artists has been handed down through eons of scarcity: we have been trained to see our art as a scarce resource, and to require money for its creation and/or reproduction. The culture of programmers comes from a world of plenty, where getting the respect of your peers pretty much MEANS sharing everything needed to recreate your final product - in part because it's so damn easy. Sharing a painting is hard, it's made of atoms. And we still have artist culture based on everything we make being atoms - hell, I still see people whinging about how only traditional media can be "real art" because of some mystical connection between the artist's soul and the blobs of paint they laid down on the canvas.

Re:great

[Peganthyrus]

I won't disagree on the code-is-art front; I've done enough programming to know there is a near-infinity of ways to solve a problem, and an art to picking which one to use!

Visual artists stand on the shoulders of other artists too; we constantly steal from each other (or are influenced by each other - same thing, really *grin*). We need each other's critique, we need each others' lessons. We don't rely on each others' work quite as directly as a programmer - I can as easily sit in a cafe and draw in a sketchbook as I can do it here at home with a ton of reference and inspiration in reach. But I wouldn't draw the way I do without having had them in the past.

Young artists also tend to be insanely paranoid about ART THEFT. We tend to see our art as this super-precious excretion of our SOUL and there is no WAY we'd let just ANYONE play with THAT. And again - this is the culture surrounding the core skills. This is what we're taught as we decide to become artists. A scarcity culture.

People will download huge torrents of comics while being worried someone will steal their precious, precious (derivative) creation. One foot in the post-scarcity world where stuff is infinitely duplicatable once it's digital, one foot in the scarcity world where every item is as valuable as the time you spent, because you can only sell it once... but programmers have a culture that came out of the science world, which is all about sharing information, and was shaped by the incredible ease of sharing bits.

Humility would be a virtue

[kripkenstein]

YAFPS

Not only YAFPS, but also the screenshots are poorer than other FOSS FPSes. For example, they lack basics such as dynamic shadows.

It seems the article authors got excited from the claim that the engine is written in GLSL and is OpenGL 3.0-focused. That, and the engine developer is not exactly humble, with claims like "definitely the most advanced open-source game engine".

Instead of dissing other engines - which offer greatly superior visuals, to boot, just look at screenshots - he should let his achievements speak for themselves. They don't, thus far.

Re:Humility would be a virtue

[Opportunist]

"basics such as dynamic shadows"

That's basics? Wow, times change fast these days, even I can remember the time when "dynamic shadows" weren't basics but rather bleeding edge, something you'd need a killer machine for to pull off at all. Maybe 'cause it ain't been a year ago or so.

You know what "basics" too many games lack these days? An interesting concept, some originality, replay value and generally something that makes me want to play them. If you can put that in a game, you can keep your eye candy. Eye candy is like new car smell. Yeah, it's nice, but it wears off too quickly and after it's gone, you only get to see that you have, essentially, the same crappy game that you didn't want to play a year earlier already.

If this means we have now a good FOSS engine at our hands that allows the development of games without first forking over six digit sums and thus being pressured to deliver something "mainstream digestable" (read: uninspired copy of something that sold well), we might get to see a few daring new ideas.

Re:Humility would be a virtue

[zwei2stein]

Bad art direction makes good engine look unimpressive.

And good art will make feature lacking engines look awesome.

Re:Humility would be a virtue

[ardor]

The topic is the XreaL engine. Not games.

Re:Humility would be a virtue

[TheRaven64]

This will probably be modded down, but it's worth highlighting. Diablo II looked great for its era, and was basically a 2D engine. Duke Nukem 3D was really 2.5D, but unless the monsters got really close it often looked much better than Quake (for one thing, duke3D could happily run at 800x600 on machines that struggled to do more than 400x300 with Quake).

Re:Humility would be a virtue

[Anonymous Coward]

The screenshots on the Phornix page are all of modified Quake 3 levels, which still use static lightmaps. So yeah, of course it still looks like Quake 3 - it's using Quake 3 levels, and full-blown dynamic lighting is mostly incompatible with lightmaps.

Have a look at the screenshots from the Xreal website:

http://xreal.sourceforge.net/xrealwiki/ScreenShots

The top screenshots are from a Quake 4 level. Quake 4 levels don't have lightmaps - all of the lighting in that level is being done in real time. It's using more modern techniques than the actual Quake 4 engine, so it actually looks quite a bit better than Quake 4.

Also, pretty much every feature in Xreal can be turned on and off. You can run without any kind of dynamic lighting, with static lighting but dynamic shadows, with full dynamic lighting, and in any case with self shadowing turned on or off, and half a dozen different techniques for actually drawing the shadows.

Re:Humility would be a virtue

[blahplusplus]

Quake always had horrible art direction, bland palletized textures, etc. Doom 2 had more artistic merit then Quake 1 and 2 ever had in terms of range of art and style.

I love quake 3 in terms of gameplay but even the art in quake 3 feels weird, random, and artificial . IMHO iD software has always had a problem finding good artists for quake. Quake 4 (single player) was a total joke, it was basically doom 3 redux except not anywhere near as good.

Re:Cool but...

[Lonewolf666]

The question is, is the engine good enough to be used by commercial industry and would they want to given the fact that companies are a bunch of copyright nazi's?

If you look at the FAQ, you will find that they use the GPL. Not the LGPL. Which means the commercial game companies would have to hand out the source code for the entire game. Not gonna happen.

Game Modding

[i_ate_god]

One of the things these people have to do is take a page from Valve and Epic and look closely how the two created modding tools for their engine.

Well, Unreal Editor doesn't really allow you to make an entirely new game out of the Unreal engine, but it's an incredible mapping tool, much better than Hammer for the Source Engine.

But, Valve has other tools as well, such as Faceposer to help in lip syncing your models. As well, the event based choreography of NPCs and physics seems to me to be unparalleled. NPC see's enemy, fire an event, which triggers the NPC to freeze, since the enemy was MEDUSA ALL ALONG! It's very intuitive programming.

So this engine needs to have an infrastructure in place to make modding as intuitive, as well as tools that make use of that infrastructure.

Re:Game Modding

[ToasterMonkey]

Uh, hello, give a little credit to ID and their modding support. Don't forget Team Fortress and how Robin Walker got hired at Valve.

Re:Cool but...

[Opportunist]

That's sad and true. But there's actually two kinds of artists, or rather, two different reasons.

First, the "it's MINE and you're not allowed to touch it kind". They basically operate under something like a CSS mindset. Look at it at their page, but don't you dare copying it.

And then there's the "my art is PERFECT, don't you dare to RUIN it" kind. They don't mind if you take their art and spread it, but they view it as a personal insult if you "improve" it. In their mind, it's perfect as it is and you should worship them for being awesome.

Neither group is well suited for open development. The first because they don't want to. The second because it leads to the invariable headache when you try to tell them that the polycount of their model is simply out of whack.

You need the artist that like to create art and that likes to see his art "come alive", when others take it and develop on it. It's a different mindset. And yes, those artists do exist.