A few considerations about UDK's news and bge

I’m usually out of this kind of discussions, but after reading this a few thought come up.

This is not another engine war, so avoid “i like blender” or “udk is the best” comments( but “we need new features NOW” is more than welcomed :wink: ).

After the change of the royalties of UDK, we should think why we use an engine…

For an hobbyist use, both engines are free.
For a commercial use, you have to pay 99$ to Epic, while blender is still free. Btw, with blender you must release the changes you do to the source code, and there are always doubt about the licences for the executables files(many topics have been opened around this subject). In addition, right now there isn’t an official easy to use way to protect your content.

Epic starts to ask for a 25% for each dollar earned by the developers, for gains over 50000$ (under this threshold you just pay the 99$ una tantum fee)

So this kind of royalties gives a lot of possibilities to small teams, and i really think the advantage of the “freeness” of blender is not that strong any more against UDK.

So, why is the industry using this kind of engine?

Let’s not start talking of the features like “faster rasterizer”, “better physics”, etc.
The engines are different, but even bge can archive very good results.

I start to think that the “problem” of bge is that it is “just” an engine. Ok, it is well plugged into blender, but blender is not a game editor, is a 3D suite that is developed for animations, modelling and all the other cool things blender can do, but not with the bge in mind.
So other engines start to be more interesting to indie companies(and there are a lot of them “making money” with udk) because there is not just an engine, but maps editors, presets and other game specific tools that are not strictly a part of the engine are bundled with it, making the production more easy and fast.

Do you think blender can have all of this kind of features(accepted into trunk*)? Or maybe if bge would be an external engine, more of this kind of tools can be bundled with it?
New engines are starting to be very competitive even for small developers…

On my side, i appreciate the open-sourceness of blender, since i can learn a lot about coding from it, but i don’t think this thoughts can be extended even to artists…

*having a patch that is not updated, distributed, maintained, etc doesn’t make sense…

I heard before that you actually need to do quite a bit of scripting and modding to get various genres that are not shooters up and running in the UDK as opposed to getting any type of game working rather quickly in the BGE, that is likely because the UDK was apparently not initially developed to be a totally generic game creation tool like the BGE, Unity, and other game engines.

Sure you can create eye-popping graphics, but I’ve heard before of you allegedly needing high end hardware to work with the UDK and play games made with it, though I’m not sure how sound that claim is.

The Advantage of the Unreal Engine is its Technology – one can say that you would need some High End Hardware to use it, but what I have seen unto now has also played rather smooth on my Laptop, and my Laptop is getting pretty old already. ( I had pretty much a hard Time playing Portal 2 and Dead Space 2 with it crashing all the Time. x3 )
But if you build one Environment in Blender, import it to the UDK and test on both, the UDK might be faster in Rendering it. So, that is the Advantage of the UDK, its Technology…

But UnrealED (the Level Editor) is practically still the same as back in 1996, just flooded with Dozens of Functions that make it way more complicated to look at (I feel reminded of the »old« Blender Layout) and with non-optimized Controls (moving the Viewport in UnrealED is nothing but a Pain, especially for a Laptop User like me! ˘o˘ ).

As already mentioned: The UDK is made for Shooters. You can do quite different Stuff with it, of Course, but as it is simply MADE for Shooters, it is way to easy to make such one (you get the Weapons, get the AI, get all this Stuff) while makin’ something else might even be harder than for the BGE just because rather few People would even want to do something else than a Shooter plus upload Tutorials therefore…

I personally dislike that you already get so many Things with the UDK. Too easily one simply gets lazy and uses Textures, Prefabs and Stuff that was already sent with the UDK, and in the End it feels more like a Sandbox.

Short said: I like the Unreal Technology as I dislike the UDK Tools.

Post Skriptum: When Updating something about Blender Game Engine, then technological Things, that will be fine, but as an Editor it is good enough already.

I used to be a warm and devout supporter of BGE.

But now all I see is an abandoned game engine that’s left way behind compared to the competitors. I think there aren’t any upgrades after 2.49b. Correct me if I’m, wrong.

I’m about to move to another game engine (probably Unity) and it saddens me to leave BGE, but the impression I get is that noone is interested in adding more features.

New sound system, initial support for real text (not the bitmap text that we’ve had to use all these years), RNA’ified logic bricks that bring enhancements like using a search to find a scene, object, or text datablock without worrying about spelling, numerous enhancements to the BGE API, anisotropic texture filtering, and others I can’t think of on the top of my head right now.

Also, the end of the 2.5x feature freeze means the BGE is getting regular development again
http://lists.blender.org/pipermail/bf-blender-cvs/2011-August/038887.html
http://lists.blender.org/pipermail/bf-blender-cvs/2011-August/038933.html

On top of that, a ton of BGE animation enhancements in the pepper branch will also be coming soon as it is currently being reviewed for merging into Trunk, and there’s various threads started recently in this forum about people working on new features and enhancements.

Sorry for underestimating the work that has been done. It’s that when I tried to migrate to the early versions of BGE 2.5+ it had many bugs and it crashed a lot. On top of that I was making a game and it was in a development stage where I couldn’t transfer it to BGE 2.5, so I stuck with 2.49b.

Still these changes you mention sound more like changes that should have been added years ago (Still,better late than never). I believe that BGE will find its way eventually and keep pace with the rest of game engines.

The problem is that I don’t know if I can wait until BGE wins my admiration again.

I’ll try the new features you mentioned in a small project.

I hope I’ll change my mind.

Moerdn’s been doing a lot of work with display optimizations, and Moguri’s done some things to optimize animations and make them playable via Python. Kupoman’s also been working with bugfixing and working with the BGE. I don’t really see what else is necessary to make high-quality, top-notch games.

You’ve got:
a map editor,
the ability to play animations,
the ability to code complex game logic and AI,
GLSL lighting,
and an engine that is getting faster and faster. A triple-A game can already be made with the BGE - it’s just that most people don’t want to sit down and work with it. Or, maybe people feel like ‘it’s not a real engine’ - I know I did before I really looked at it.

Don’t forget his work to make object groups much more usable in a way that would allow many people to easily use them as part of general game development (do to his work allowing rigidbody joints to spawn with the group when created in the game and being able to move all your group assets to a hidden layer).

Maybe I’ll actually start using groups in my games if this gets in trunk as a result :slight_smile:

Recast and Detour is hoped to be merged into trunk in the first batch, as are the other features. Components will not be until some bugs are ironed out.

Allways when I see this kind of thread I say the same, the biggest problem about this engine is that nobody take it seriously(on the user point of view), you indeed can t do a lot of things but you can make games with high quality with it right now, with what you have, try out yourself but try hard, and you ll make a amazing game with this oltdated engine.
For tell you the truth I m very happy with the soft shadow( that I dont know why in the hell is not on the official built yet) and the huge increase of speed that it had, now if we have better armature processment it ll be awesome.

I think we need more tools like those, that are integrated in UDK, Unity and so on. A simple way to create project structures, an asset and script browser, easy to make && use logic templates, a way for automatic baking of static lighting, a cleaned up GUI for BGE render engine…

Every little feature that makes it more easy for creating games is highly appreciated.

It is easily possible right now to bring a lot of small (or even big) features to bge with addons. But in this section, I didn’t see any development. Why?

greetings, moerdn

As moguri told me, biggest thing that is slowing down BGE is its single threaded (if i remember correctly). someone needs to fix that soon. we can expect atleast 3x faster BGE after that would happen. :slight_smile:

Regarding UDK, ive been eyeing the development since past 1 yr. and they come up with something jaw dropping every now and then.
Yes i’ll agree there are very less people who exploit BGE to the fullest but games like http://www.udk.com/showcase-hawken makes me feel were are lacking quite a bit of things.

btw just wondering can anything be done about license?:confused::spin:

Regarding the license, idon’t know we can override GPL if we just rewrite the code from scratch( maybe it would only require to rewrite variable names function names etc.).

These features would be welcome.

Don’t misunderstand me. I’m not looking for an engine that does everything for me. I just want to have options and tools that can make things easier along with the full control coding offers.

I’m just tinkering with Unity and there are things that can be done very easily e.g. the player controller. Why not have the basic controls for the player ready and then for adding advanced control use whatever method I want.

Or if someone wants to make basic controls by himself, he could opt to. E.g. I 've never used the mouselook script most people use. I have created my own and I prefer it.

Change the GPL takes ether consent from all contributors or a complete rewrite. That’s the good thing with GPL - It protects the freedom of the community and every contribution - No one can lock it up.

But the GPL is only for changing blender and the BGE itself - not for games and other art made with it. That’s usage and usage of GPL programs is completely free, for commercial or any use (porn, genetic research, you name it).

Add-ons are underexploited. But I seen two attempt to make automatic baking a while back - possible there is not to much work bringing it up to usable state. Now - add-ons just like games need not be GPL so it might be some licensing issues but may still be worth to look up for anyone interested.

I saw a lot of “if you do hard work you’ll get a AAA game”

I never said that this is not true, i said that in the professional industry productivity is not a secondary part, and if you need to write the whole AI without any “help”, or to model each tree instead of having to connect a few nodes to generating the tree you want, already weighted for physics simulations, you are less productive.
In fact i didn’t talked about the engines themselves(and their features), but about all the tools around them. While UDK has a lot of tools helping the users to make up games (ok, it is made for shooters. This means that you can make a shooter in no time, not that doing a different kind of game will need be more hard that making it without any tool it gives).

Maybe its tools are old, but something is better that nothing at all.

In the hobbyist and open source world, i often heard “it’s not right to have everything already done, you have to get it on your own” or “let’s get your hand dirty and work it out” or “to really understand [fill here something], you need to spend a lot of time on it”.
Well, maybe i don’t know how it works in the artistic field, but on the computer science field this are a b*****it. You need to archive something, and if the work of someone else is good for you, than you’ll use it. You don’t reinvent the wheel each time, you can enhance it, or her usage. So, i don’t think it is the industry standard to work on something 5 years if you can make it in 1 year.

Final notes:

  1. it is really unfair to the bge developers saying that it is not developed, because this is simply not true.
    2)i’m not criticizing bge, and btw critics should be seen as input for improvements, not as accusations. I’m just underlining that bge needs a bit of love even in making easy to develop games. Right now even only logic is a full pain to use.
    3)making the bge multithreaded would be nice(and i don’t think it would be too much difficult), but it is also a lot of work.
    4)A really nice step into the direction i’m talking about is moguri’s bgui.

i exactly had this game in mind when talking of indie developers making great games

@LaH:
With eight uv-layouts from Kupomans branch, it could be much easier to implement this feature, maybe fully automatic. A big problem would be baking for linked duplicates and group instances. If this problem could be solved, lightmaps would be as powerful as in UDK and also usefull for non bge workflows.

@rarebit:
Tried to PM you, but your mailbox is full. Opening a new thread for bge addon stuff would be great. I wanted to code this already, but atm there is not time left.

Btw, totally agree with Makers_F. Reinventing the wheel every time is unacceptable. BGE makes it very easy to build up complex logic in a short amount of time, but makes it horrible hard to make this ‘buildingblocks’ reusable.

There is another thing to consider: BGE is opensource, and maybe its up to the community to push it further. There is so much awesome work done from members here already but its cluttered all over this forum, hard to find and maybe hard to use for newbies. Why not providing an unofficial blender build that is synchronized with trunk, but contains some improvements and scripts like mouselook, filters, experimental stuff and features…

We have to look at Unity and UDK for getting an feeling for nice workflow and useful features. I’m sure a lot of these could be implemented in bge.

greetings,
moerdn

For me it had some improvement on multithreading compared to the blender 2.49b. about the 5 years when you can make it in one, whell, I really dont agree with that! but I can tell you that for years we complain, me included, really for years, almost nothing changed, and now people had some nice ideas here about the tutorials for bge and the suport of projects too was great, but I hope to see progress on those ideads.
There are the developers but there are few of them unfortunably. But I think what is the reason to complain? or to point it up that bge is outdated? makers f i think that you is an advanced user of blender or something why you dont point out a solution, like agoose or guillaume did I m not desagreeing about blender s outdated thing, Im just saying that blender ll not get competitive with UDK if we dont do nothing, I really dont know what to do, I realy dont know if anything that we do ll make any difference, for tell you the truth I m not caring too mutch about it.
I ll keep working on my games and I ll keep find a way to profit with them like I did for pic a pix 3d. I hope you find a way to get some money with yours too.

I would like to open by disagreeing with “the BGE is just an engine”. I spoke with Campbell and he said, quite truthfully, that a lot of work has been put into integrating the Game Engine with the Blender core. Just think about how easy it is to start and run the game - no compiling, just plug’n’play! The main issue that we face with the BGE is not so much a lack of features, but a lack of direction!

The BGE is renowned for its ability to create logic setup’s incredibly fast and easily. However, having a simplistic approach to game development will eventually slow you down later on, when specific complex functions are required. This said, nodal logic would certainly provide a middle ground between Python and logic bricks, (if we can hope to see Sven’s university funding granted!). I believe that there is a future for the BGE. UDK is an unfair comparison to our Game Engine, because one is funded by income, a capitalist engine if you like, whilst the other is community driven.

You mentioned an unofficial Blender build. It sounds like a great idea, but why not integrate common scripts like MouseLook into trunk anyway? Whether this would be a logic brick, or a component, i’m not sure. For sure, we are lucky to be open source. But, i believe that in order to criticize, you must be willing to tackle some issues. Therefore, i’m embarking on a project to:
Find,Target,Solve.
(features or problems of the BGE).
I’ll post back here with a link, if you like.
Just a note: i’m not comparing UDK to BGE in any respect :wink:

Totally agreed. i think we need a parallel svn just for BGE Devolopment.
Also a separate patch tracker if possible. A good place to host the code will be code.google.com imo. it would be hard to sync with the main svn though. but it would allow us to push some further experimenting with BGE source code.:smiley: