Engine into parts?

[daemonicky]

It is possible to exchange whole Quake engine. Each of these engines suports differend effects like shadows, fog, colored lighting, shaders ... I was thinking, that it might be interesting to separate even more.

Image being able to load level which is not bsp tree but is a heightmap or made of voxels, everything else would be working just fine because all functionality of level would be encanpsulated.

And imagine having voxel based entities instead of polygonal ones. I can see only benefit of being able to do holes inside of these models, so there might be something like in Red Faction. But voxels are not accelerated by videocards AFAIK and I am not sure whether it is possible to combine them with graphics (maybe trough some shaders ...).

Physics in Quake is based on AABB and BSP models. It uses level data format and information about boxes from models (or map), but it might be interesting to add ovals or rotated bounding boxes. Maybe use Newton, ODE or whatever is available. I know there are some Quake C physics.

Basically split Quake into modules which access themselves only through some interface. Interface between Quake C and Quake (think, touch, prething, onconnect, traceline, pointcontent ...) is great example of such interface, not that it would have to be same .

What do you think about it?
I mean it might get more people to Quake community. I imagine Quake support some 2D platforming levels or topdown RPG levels or Descend freemove 3D levels or some simple grid levels and all of them have entities and are scripted by Quake C. So someone can have platformer in Quake and someone else ditch darkbasic, gamemaker and similar tools ...

[Baker]

Won't comment on breaking the engine into pieces, but to make Quake development more popular in any sort of way that matters would require:

1. Premier Projects. Foremost, a successful example mod or 2 that attracts interest. I would firmly put that in the FAIL category. Kurok did a great job, Nexuiz did a great job. Quoth attracted mappers for a while. Without a premier project that showcases capabilities, any idea of re-popularizing Quake modding will not happen.

As Kurok and Quoth were closed source and Nexuiz and DarkPlaces are way too complicated from a code standpoint, Kurok was only successful in showcasing what a Quake engine could do on a PSP, any use of Quoth was solely restricted to mapping (got new monster idea or need a slight QuakeC modification --- forget it).

Nexuiz is complicated enough a guy like Div was the only one able to maintain the QuakeC. And DarkPlaces, while well written and organized, is poorly documented with a feature set that is both deep and well -- undocumented. That isn't LordHavoc's fault --- he certainly isn't 20 people. Thinks like that, MH's blog, some of the discussion at Inside3D and things like that I think are exposing people to more and more information.

2. Documentation A well-organized Wiki with some newbie like examples and some videos. Doesn't seem likely to happen, for various reasons.

3. Decent and Easy To Use Tools. I'm not really sure what map editor you'd recommend to someone that has the right feature set, ease of use and is modern. A lot of people hate GTK Radiant 1.5 (versus 1.4). Worldcraft 1.6 is ancient. Worldcraft 3.3 is ancient. Anything except a Radiant fork will eventually die an ugly closed source death of non-updates.

And what modelling program do you use? And what model format?

Now, I am actually an optimist but I think the above problems are an accurate description of severe problems.

The reason I am optimist is kind of weird. I used to think the "driving force" was widespread knowledge accessibility. But it is probably more about the depth of knowledge in the most active "people that do stuff". One example would be the interesting effect that Seven and Ooppee had last year with DarkPlaces modelling and texturing.

[daemonicky]

I am new to this scene, so forgive my small understanding of who is who and why is something the way it is :

MH's blog ...

Who? Which blog?

One example would be the interesting effect that Seven and Ooppee had last year with DarkPlaces modelling and texturing.

What was this effect?

Anything except a Radiant fork will eventually die an ugly closed source death of non-updates.

What is the reason others die?

2. Documentation A well-organized Wiki with some newbie like examples and some videos. Doesn't seem likely to happen, for various reasons.

The state I see now as a beginner is quite decent. I needed explanations and got reference manual and quake specs and some tutorials or even better there were some state diagrams http://ai-depot.com/FiniteStateMachines ... tical.html . Well, I had to dig into quake source and get some with a pencil.

But there are tools to do erd and state diagrams from code. http://stackoverflow.com/questions/6603 ... rails-code . They woul help me I guess, but digging into code is a nice excercise in language ... It depends on person i guess ...

3. Decent and Easy To Use Tools.

Agreed.

... what map editor you'd recommend to someone ... And what modelling program do you use? And what model format?

For maps Qoole. Easiest editor I found. It does not have tons of features, but has source code (and I believe it is open source of some kind), so it is also extensible.

For models none yet, dont need it, but I am thinking about free 3D Studio Max for games http://www.turbosquid.com/gmax , Blender or Milkshape 3D (actually I want to get some of its plugins functionality and have something similar to model creation in werkkzeug1).

[Baker]

I am new to this scene, so forgive my small understanding of who is who and why is something the way it is :

No problem. Everyone was new once.

Who? Which blog?

MH's blog at http://mhquake.blogspot.com . MH has been delving into some incredibly tedious and painful issues, dissecting them and getting answers. And talking about it on the way.

Things like timers. MH has been a one-man army plowing through complicated issues for 2 years. He made a whole new engine from the ground up in like 2 months.

One example would be the interesting effect that Seven and Ooppee had last year with DarkPlaces modelling and texturing.

http://quakeone.com Seven and Ooppee -- who are both registered here --- have done incredible works for DarkPlaces and Seven has done a crapton of documentation. One shiny object:

Anything except a Radiant fork will eventually die an ugly closed source death of non-updates.

What is the reason others die?

Closed source works eventually become worthless junk. The reason is that they cannot be maintained. Imagine trying to use Internet Explorer 3.0 to browse the internet.

Some operating system change (like 64-bit), technology change or feature expectation will either make the application not run, run terribly or some new expectation will render it obsolete.

Quake is largely an open source place, but some of the tools are aging.

2. Documentation A well-organized Wiki with some newbie like examples and some videos. Doesn't seem likely to happen, for various reasons.

The state I see now as a beginner is quite decent. I needed explanations and got reference manual and quake specs and some tutorials or even better there were some state diagrams http://ai-depot.com/FiniteStateMachines ... tical.html . Well, I had to dig into quake source and get some with a pencil.

The state of things is ok to a very serious individual. Still, critical information doesn't find you, you either have to ask to body of participants in many cases (this forum) or crash into it yourself.

I'm just pointing out some of the tougher questions out there. Quake is absolutely a great game and a great toolset. It just requires a very aggressive attitude to get things done.

[goldenboy]

Anything except a Radiant fork will eventually die an ugly closed source death of non-updates.

What is the reason others die?

Closed source works eventually become worthless junk. The reason is that they cannot be maintained. Imagine trying to use Internet Explorer 3.0 to browse the internet.

There is another factor in this equation. Because Radiant isn't really maintained either. A couple folks adding stuff to Netradiant when they feel like it is not "maintaining".

Still another factor is that some mappers don't want to switch away from worldcraft because they're not willing to leave their comfort zone. Ironically, you (Baker) strengthened Worldcraft (and its stupid Valve 220 format) with the Quake adapter. And as long as Radiant isn't the most used editor in Quake, people will just not see the need.

[Baker]

I have the feeling that eventually I will spend some quality time with the source code for QER Editor. Necros likes it, therefore I'm sold on it. Necros doesn't seem to like 1.5, although it appears you do and what seems to be a minority of Q3 mappers.

I'm rather confident that I could make it load models and such (like within 4 hours of browsing the source code).

But I'd like to take QER and finesse and remaster it in an easy-to-use package. I like super newbie friendly. You know ... click and it installs and then no config, no hassles development.

I did what I did for Worldcraft because: I was getting pissed off at Worldcraft 1.6's slowness in rendering. Worldcraft 3.3 has OpenGL accelerated rendering and Worldcraft 1.6 screwed up textures after rotation horribly. Plus I wanted to "give back" to the community and it was Quake Expo 2006 and I came up with a 2 day speed modding project.

But any of that stuff will be after Quake Expo. I plan to release my engine on Quake Expo. It has an interesting feature set and a very different approach to transparency. It's kind of funny.

[goldenboy]

I'm thinking about a Radiant fork (either 1.5 or Net or a mixture) myself.

[Baker]

I'm thinking about a Radiant fork (either 1.5 or Net or a mixture) myself.

Your stubbornness and dogmatic point-of-view indicates you'd do a great job at that. I still don't like 1.5 and I'd make QER more Worldcrafty, but in a way that wouldn't piss people off

[goldenboy]

I'm thinking about a Radiant fork (either 1.5 or Net or a mixture) myself.

Your stubbornness and dogmatic point-of-view indicates you'd do a great job at that. I still don't like 1.5 and I'd make QER more Worldcrafty, but in a way that wouldn't piss people off

Good luck with not pissing people off.

[mh]

...I think one of the reasons people like it is that it's not Worldcraft-ey.

[Baker]

Any interface changes would be optional. I'm a believer in user preferences. You can have it both ways, you know.

[Kinda how I like maintain a conservative engine and a non-conservative one. There is a never "one-way" that presentation or interface has to be done.]

[daemonicky]

Qoole has cool interface.

QER editor = Quake Radiant ?

[Baker]

Qoole has cool interface.

QER editor = Quake Radiant ?

I called it the wrong name.

The editor name is "QuakeEd" but is a fork of QER. It is a more Quake 1 tuned version of QER by sikkpin.

http://www.celephais.net/board/view_thread.php?id=60225

[I don't claim to know much about Qoole. Although based on what I read, it isn't in mainstream use by any of the heavy duty Q1 mappers, which seem to be using Quark, Worldcraft or Radiant. I'm not sure what this means about Qoole.]

[qbism]

Would be nice if BSP Editor source were opened.

Tread3D is open source, but nobody has picked up development where it left off AFAIK.

I'm not Mr. Mapper, but I prefer Quark, which is open-source and actively developed. Maybe because Quark is similar to the CAD programs I use.

Quake could use a Total Newbie Dev Kit with everything in one place. Doesn't have to be perfect, just lower the barrier to entry. Maybe based off of OQ++ in the interest of being self-contained.

Quake is in the same boat as the Star Trek franchise, plenty of potential fans but in need of a premier project, I agree.