The Holy Graal of Programming: Games' Engines

[Prof.Wizard]

Let's talk about game engines, shall we? I heard once from a student in Informatics that there many lucrative and pleasing projects in the programming business but nothing as such creating, launching, and distributing a successful game engine.

Look at Half-Life for example. Its engine sold like crazy for the original game but also made itself respected in the IT world because it spinned off hundreds (including Counter-Strike) of other titles. The engines of Quake and Unreal Tournament are other nice examples. Also simulators with their elaborate graphics and physics are very high-tech projects with lots of work and input, even from pure mathematicians. Nothing can beat IL-2 Sturmovik and Falcon on this!

I had been playing a game almost two years ago, Hitman, and these days beating its sequel, part 2 of the story. I start this thread cause I find its engine Glacier extremely potent and intelligent. It has been developed by a Danish company, IO Interactive, and you can find more information about it here.

Tell me what you think and please give further examples of game engines you like and why.

 

[Groltz]

My favorite is Croteam's Serious engine.

It is the most stable FPS engine out of all the ones I have tried. The amount of configurability available in the preferences is outstanding.

 

[Prof.Wizard]

Serious Engine overview. In case you wanna know more about it...
I also linked Glacier in my first post.

 

[Dozer]

I have to agree with Groltz on this one. I had not tried the game until he recommended it during our visit out in Seattle, but man, it is a slick game interface. You can fine tune just about anything you want in the game. I'd like to see this flexibility incorporated in other games.

 

[Dozer]

BTW, cool avatar Groltz.

 

[honold]

fyi, half-life IS the quake 2 engine...

 

[timwhit]

fyi, half-life IS the quake 2 engine...

It is a very highly modified Quake 2 engine. How much do you have to modify an engine before it becomes a new engine?

 

[honold]

if they're proficient enough to modify it beyond all recognition, one would think they would save the draconian license fees and simply write it themselves

if they're not proficient enough, i doubt they could make technical changes so notable as to merit calling it their own

cosmetics and usability don't count, anybody (including freeware college student mod developers) can do that

 

[Cliptin]

if they're proficient enough to modify it beyond all recognition, one would think they would save the draconian license fees and simply write it themselves.

Presumably assuming the engine met all technical criteria, if the fees were too oppressive then they would have written their own. The engine is not alone in a license. Also included are the world building tools.

One tries to avoid re-inventing the wheel when possible. That said there are really only a few FPS engines out there.

Re the original post: Around the time of Quake and Quake2 when so many amateur mod developers began developing content it was realized that you could make just as much or more money licensing the technology to other game developers as you could selling the game you are developing in stores. And Unreal was born.

 

[honold]

understood cliptin

clearly if it's profitable for one company to make and sell engines, it should be more profitable for the other to make their own (as opposed to licensing one)...

i think all companies, even gaming ones, are most interested in making money.

this circles back around to my original point. assuming that the developers are most interested in profit, they will do what they're doing in a way that would make the most of it. if id can make a killing selling engines, that must be a viable racket, and they also must be charging a decent amount of money.

the only reason i can conceive of (assuming the company actually wants the most money) that a company WOULDN'T write their own engine would be if it was more cost-effective to use someone else's. if it's more cost-effective to use someone else's, it probably means they're not exactly pros at engine design.

if they're not exactly pros at engine design, i doubt they could hack up somebody else's engine enough to call it their own (legal issues aside).

a flow chart would help on this

it's clear why people make random fps games using other engines if they're not uber programmers - they're just artists, writers, mappers, and gamers with what they think is a good idea.

 

[Clocker]

I think you also need to think about time-to-market.

If I had a great idea or some great 'artists' working for me but didn't have my own engine, I may consider just liscencing another already established engine. That way, I'd get my game on the store shelves a lot faster while limiting the possibility for someone else to trumt my idea while I figure out how to make a good engine. Time is money. Also, using an established engine also has the advantage of not having to deal with too many compatability issues.

I'm no expert in programming (I can't program a freaking HP calculator) but there are a lot of advantages to going with proven technology in a fast moving market..

C

 

[honold]

time-to-market was considered - that's why i pointed out the bit about non uber-programmers. for them, it's a necessity unless they want to learn how to program before they begin their project

if i can take a honda crv and rice it out to kingdom come (making it totally unrecognizable in the process), it doesn't mean i made the honda crv (or that i'm remotely capable of designing one).

i don't think taking anyone's work and hacking on it, learning from THEM as you go (via implementing their own api calls, etc) is legally, ethically, or even independent of those issues correctly considered as your own.

if they're so capable of writing a unique engine with compelling features as compared to quake, unreal, serious, lithtech, etc engines i think they would cut the cord and write their own. they haven't, and a 60% modified quake engine is still 40% quake. if the important parts are the ones they left alone, the percentages don't matter.

 

[LiamC]

I work with a couple of hundred software developers. We also get a lot of new graduates every year. Some of them think they are über developers.

Most (nearly all) fall short. Maybe with support, time and good mentoring they might make the grade. I can count on one hand the number of guru's I work with...

Sometimes they have great concepts, but taking a concept into reality where you have to account for every possible scenario, ie make the code bullet proof, it all becomes too hard (or tedious).

0.05

 

[Handruin]

My favorite is Croteam's Serious engine.

It is the most stable FPS engine out of all the ones I have tried. The amount of configurability available in the preferences is outstanding.

Despite the great combat Groltz and I have playing the game as a team, I find the gaming engine to lack response at times.

Out of most FPS games, I've had the best results with the quake II/half-life engine. Serious sam for example may be highly configurable, but the speed in which the engine interperates my mouse movements bothers me during game play. Perhaps it is the speed of my system causing the issue, but why would quake III and even UT2003 have better responce?

 

[Cliptin]

the only reason i can conceive of (assuming the company actually wants the most money) that a company WOULDN'T write their own engine would be if it was more cost-effective to use someone else's. if it's more cost-effective to use someone else's, it probably means they're not exactly pros at engine design.

The last sentence is where the logic breaks down.
I think you can be very good technically at engine design and still not have a good enough design to sell licenses. It is only profitable to design with this intent if you can sell licenses. It is much more tedious and therfore time consuming to make code modular enough to have broad application. Not to meantion that you have to plan development time so far out that you are coding against features that are not even available in hardware yet, and might never be. There is some risk involved.

You assume an infinite demand for game engines. If your engine sells no licenses, you have spent more time on the engine than you can make back in pure sales. As you know time is money. Not to meantion that it is cheaper to develop in some areas of the country than others.

if they're not exactly pros at engine design, i doubt they could hack up somebody else's engine enough to call it their own (legal issues aside).

IIRC, One of the main contributions the HL team made to the engine was massively improved network code.

 

[Groltz]

but the speed in which the engine interperates my mouse movements bothers me during game play.

Did you run it with Vsync turned on or off?

Did you adjust the netplay settings for optimal results?

http://www.techspot.com/tweaks/serioussam/serious-8.shtml

I have noticed controller lag when in SS when playing with Vsync turned off.

I just found this on another forum:

sometime last week on irc people were talking about this issue i think and one thing that we found out is that for some reason the higher your max player setting is the slower everyone seems to go no matter if theres only 2 people on the server. its like sam reduces the available bandwidth based on the max player setting. so servers set to 12 players max for example would lag really bad with this theory and in my experience they do. maybe having servers with like 5 players max would work a bit better. id like to see someone test this theory out. get together with a friend and play through one map only with server set on 16 players max then try again with 2 players max. see what happens.

Yes your theory is correct. I tried it with only 2 people playing with max players 6 = unplayable; max players = 5 = unplayable; max players = 3 = very good; max players = 2 = unstoppable!

I had it set to 4!


And this:

We tried quit a few different MAXbps settings and found that lowering the default settings in the game helped quit a bit. I run Sam at a MAXbps of 4000 on the server (DSL 128up/512down) and I play at 5000 online. This has helped my connection as well. I have suggested lower settings to modem players. The ones that have tried it get back to me and say that it has helped and they don't f9 anywhere as much as they used to and the game is playable for them now. He seemed to think at the time that Servers shouldn't be set any higher then a MAXbps of 6000 and that if everyone was set that low or lower the game play would be better.

 

[honold]

The last sentence is where the logic breaks down.
I think you can be very good technically at engine design and still not have a good enough design to sell licenses. It is only profitable to design with this intent if you can sell licenses. It is much more tedious and therfore time consuming to make code modular enough to have broad application. Not to meantion that you have to plan development time so far out that you are coding against features that are not even available in hardware yet, and might never be. There is some risk involved.

well spoken. unless the company intends to go out of business or stop making fps games immediately after making this one (??) they are making either a lasting commitment to blow money on future licensing or develop their own engine (rendering their initial license dubious).

IIRC, One of the main contributions the HL team made to the engine was massively improved network code.

while we all appreciate good work, netcode is still trivial compared to the graphics. the recent challenge pro mode q3 mod also features signifigant netcode improvement and tunability...

 

[Prof.Wizard]

IIRC, One of the main contributions the HL team made to the engine was massively improved network code.

Correct. And this implies code with adaptability. This was definitely a good thing and made CS so successful in the multiplayer field.