Game engines

[SimplySerenity]

So I've got an idea for a game that I would like to work on in my spare time, and originally I was going to create it as a quake mod. But I've come to realize that quake just isn't the engine for me, and I'd rather use something a bit newer. Any suggestions on what engine I should use? Any ideas would be great.

edit: or maybe I should just take the time to learn my with quake modding..?

[mh]

I've come to believe that Quake is an awful tool for modding.

It looks simple on the front, but it's just so restricted and inflexible, and there are limits beyond which you just cannot go without huge architectural overhauls. You may find yourself in a position where you have to pull back or otherwise compromise on planned features (or put a hell of a lot more work in than you originally anticipated) owing to it's limits. As id's first version of this technology it was the one that they got to make all the mistakes on; later engines are much better.

Just look at all the forum posts asking "can I do this in QC?" - many of them great ideas, but so many doomed to failure from the start. So long as you're happy to work within the bounds of what Quake can and cannot do, it's probably OK-ish, but as a general rule, as soon as you try to push things beyond you'll start hitting walls. For sure some fantastic things can be done with it, but just look at the amount of work that's needed before you get there - work that's already been done for you with other options.

The correct answer depends a lot on what type of game you want to make, so these are just suggestions. For staying within the comfortable confines of id technology Q2 or Q3A would be suitable choices. Q3A in particular will give you much more flexibility as well as client-side scripting (all of which would need to be added from scratch to Quake). If you want to stretch outside then the free versions of UDK or Unity might be good choices.

</eeyore>

[toneddu2000]

UDK and Unity are the most common choices but be aware that UDK content creation / scripting process is HUGE and frustrating (of course, compared to quake's one UDK does 10000 other good things ). The problem with UDK I noticed is that I didn't find any good learning resource; on the internet you can watch a billion of crappy video tutorials (except and )but no just a good commercial manual (I bought UDK Game Programming with UnrealScript by Rachel Cordone but I found it not so well written). Unity users said that's simpler and it has that the advantages that you can use Javascript to make game logic. There's another game engine I'd like to describe, . It seems very promising, with amazing tools and visual goodness and, last but not least, it's made by italians (he he a little note of proudness! )!
I'm using Darkplaces Engine and I love it/hate it, but, as MH it's a tough work make games with quake engines; every step you make you don't know if you'll find anyone in the world that had the same problem of yours to help you!

[SimplySerenity]

Thanks for the all the advice guys. I've used UDK before, and I don't really enjoy it. Haven't really messed with unity yet so I guess I'll check it out. You scared me with the link to the s2 engine though tonnedu because I had been thinking that if I ever made a game engine it would be named the same thing. I guess someone had beat me to it.

[qbism]

Try looking at this another way: turn the process upside down and let the engine be the last item considered. Think about the end product first and count the steps (and hours of work) that lead back through various options.

[Spike]

a large part of the problem with modding quake is that people tend to stick to qc. for a mod, this is fine. but for an entire game? baaaad bad bad bad. qc is just far too limited and/or slow, and you will want additional features to distinguish your game from the million other quake mods out there if you did.

if you do decide on using some 3rd party quake engine, be prepared to fork it. just keep your changes as self-contained as possible and you'll be able to upgrade (git is supposedly good for things like that, though I don't personally use it).
Certainly you'll want menu+hud+etc changes.

Quake is good because its Free, but its bad because its Free by force, which can result in some real limitations on how you distribute/sell your game. Same with q2+q3+doom3.

[Seven]

[SimplySerenity]

Once again thanks for all the help everyone. So I guess I'm going to make sure I know entirely what I want from my game before I begin. I've got a pretty solid idea so far since I've been thinking about this for a few months now. But from what's been said, and it does make sense, I should really know what I want before I start so that I can choose my path correctly.

[goldenboy]

Quake, really, is a pain if you want to make a game that looks OK, has basic features you'd expect today (proper rotating doors... curves... mapmodels... pushables... physics... ambient sounds... cutscenes... terrain), and has halfway decent gameplay (AI, puzzles, movement etc).

The lowest acceptable level for me, where idtech is concerned, is a much improved, actively developed engine with features and performance (DP, FTE, QFusion...), a proper map format with proper tools (support for large maps, good looking lightmaps, support for mapmodels etc) such as q3bsp, rbsp or fbsp, and if it's QC it pretty much needs to be heavily modded and extended to be viable.

I think the best looking idtech based games right now are Warsow and Alien Arena, with honorable mention of games such as Steel Storm, Open Arena, Blood Omnicide, and idtech4 mods such as The Dark Mod and Hexen: Edge of Chaos.

You can do good stuff with idtech, but Quake 1 is a VERY harsh mistress, be prepared to suffer if you use Quake 1 map and model formats and horrible stuff such as Worldcraft and QME that should be piled up and burned.

[Jukki]

I am hijacking this thread for a question.

Me and my friends have a plan for rpg fps game. In nut shell we have 2 choices. Quake or source. (source because mapper knows the source maps like his own pockets. ) Currently the plan is to go with darkplaces because we are more familiar with qc and ngine code in it. Is this good choise in your opinion? (we would use every possible feature in darkplaces that we can)

[goldenboy]

If your mapper knows Source and its tools very well, don't underestimate that. He would have to learn Radiant for idtech mapping (at least for q3bsp which is what you'd likely use).

RPG-FPS sounds like both platforms can do it. TBH Source in its current form would probably look better and the tools might be better developed, but GTKRadiant, q3map2 and Blender aren't by any means bad either. It's not just about the engine, really.

The longer I think about it, the more I want to say "Source". You might be more happy with that in the long run, even if you can't use QC. I guess there are a lot of Source mods from which you could probably borrow stuff. And vanilla Quake (QuakeC) doesn't even support e.g. proper rotation, while Source likely has those basic features available already. That can save you a lot of initial pain.

Plus Source is an actively developed platform with good documentation. So I'd be slightly in favour of Source if I were in your shoes.

[mh]

Definitely agree. Pick Quake if you want a technical challenge and/or have fun fiddling under the bonnet (but be sure that's what you want as you'll be doing a lot of it). Pick Source if you just want to make the game.

[leileilol]

[mh]

Well Source was one of the 2 choices Jukki mentioned, so hence the recommendation. For a more traditional id-style content pipeline, and if a GPL toolchain was considered important, Q3A is probably a better bet.

[goldenboy]

I've seen very pretty Source maps at Mapcore and elsewhere. I especially like the lighting. Everything depends on the skill of the mapper, as always.

I'll take your word for the crappiness of the content pipeline.

@ Jukki: Try both. Make a testmap, some models and some gameplay on both platforms. Then make an informed decision.