Disenchanted with 3D Game Development

The whole reason I've been a huge Blitz Research supporter is because of the 3D solution it brought to the table. Blitz3D was revolutionary back in the day. Today though, it is dated and seems clunky in a way. With the release of BlitzMax, there was hope of a cutting edge way to access all the latest 3D features and goodies with Max3D. I know a lot of people who've been waiting for this to materialize over the past couple of years. Sadly, it hasn't been released. I'm not getting any younger, and my dream of making a stable feature packed 3D game may be just that...a dream. I'm not bashing. I'm being realistic. I'm to the point of moving on.

And now for my question. What do you guys recommend for a solution to 3D game making that has all the latest features and is fast?

Cobra3D is virtually here. So hang fire. Lots of people are hoping that Cobra3D is going to be a sort of Blitz3D v2 type of thing.

I just want something easy like Blitz3D but with some advanced features like shadows, shaders, normal maps, and high speed multi-player (threading?). Is that too much to ask in 2007?

Wait until later this week.

Well I've been waiting this long, what's another week? (attempt at humor)

The problem with Cobra is that whatever they shown so far crashes. The demo language on SqueakyDuck Design's website crashes, the shadow demo crashes...

NOTE: I've cleared up this comment I made her to SqueakyDuck Design's, and I am sorry if I might have touched the reputation of GrahamK. He's definately a great man to know, and there are a lot of good things comming to Cobra that I think everyone should look at, if they already know Blitz3D.

Well that sounds interesting doesn't it? I've taken a quick look at Halo's engine and it seems solid - plus I can understand how to use it ;-)

Leadwerks engine seems to be fine, but for me, not welcome yet since I cant see how to put 3DSMAX in to pipeline with that engine..and some of the demos I ran was ridicilously slow /just few bricks is on scene/...I dont know why is that since I didnt change any settings, but its really slow on that particular demo..

I was talking about this with Damien, Bmax will be as old as B3D was when Bmax was released pretty soon, and not much to show for it really :(. I'm glad we looked for our own solution sometime last year.

On the Mac i suggest Unity. On PC it depends no what you need exactly.

I don't normally comment, except to address misinformation.
I've only heard of one person so far, who had a very odd setup where the demo has been crashing. As for the shadow demo, that's written in delphi, but uses the same 3d engine as cobra, only thing that is likely to cause that to crash is outdated drivers, or alternatively an old vid card.

If you know otherwise, I'd be interested to hear your machine config over at the squeakyduck site so I could address the issues.

Yes. Bmax will need something of a miracle over the next year or so to be as successful as B3D. It's a shame. You can understand why its evolving a little slower than blitz- there's much more to deal with.

People have been wanting a 3D engine for the last 3 years. (yes, dec 2004 means in december this yeat Max is 3 years old..) Some have done their own. MiniB3D is pretty awesome at what its intended to do, but it doesn't cut the mustard for high-end features. Halo's engine is coming along. Then there's our team working on Flow. We want Flow to be *complete* when released.

Once youre used to it, Ogre is pretty damn good. I've been shown other engines and havent been impressed by any really.

Wrapping Ogre properly takes time though, but if you know what features you need and what you don't for definite, you could get the basics running in no time.

Depending on which metric you use to measure "success" it already is.

what does Halo got cooking in his kitchen?

BlitzMax is great. It's definitely my language of choice, bar none. That said, Mark's recent anouncement of a new direction with the 3D engine did disappoint me a bit. It feels like we waited all this time, only to find out he's 'started again'. I'm sure it will be great when it is finally released.

My thoughts on what to do for a more near term solution:

1) Check out Halo's engine. It has certainly impressed me and seems very easy to program. It also lets you stick with Max as a language.

2) MiniB3D, although not quite there yet, is probably the best bet for command compatibility with Max3D when it comes.

as successful as B3D.

I'm waiting on FLOW, patiently. :) I'm also waiting for the Leadwerks engine to be updated to see how its progressed before purchasing.

I think Bmax has been very succesfull. Those who don't want to wait for Max3D have a plethora of 3D Engines already available with wrappers that they can use with BlitzMax.

I guess if you haven't done it in Blitz3D yet then the likelyhood of you doing it in Max3D is bleak.

:)

There is also another good reason for some people to wait for mark, and his 3d solution.

We all know that for mark this pays his bills, and for as long as he gets an income from it he will continue to develop the language, well for the people without jobs of their own, and limited money to spend on a 3D engine, its in their own best interest to spend it on Him, rather than some 3rd party. as it helps secure the future of the product line.

I may be wrong tho, just something that came to mind today.

I am looking forward to cobra3D as well.

>I'm waiting on FLOW, patiently. :)

What is FLOW? Got links?

what? who? where?

>what? who? where?

you forgot: why???

Chroma,

I'm gonna be blunt here. Have you got a thorough game design already waiting? That's what you need to be working on if you haven't. Otherwise when the 3D Game Engine of your dreams arrives you will have to start then! Think of ALL that time wasted! I do know exactly what you mean though, I too have been using B3d since day 1 yet I still yearn for a new feature rich engine...but at the end of the day B3d still cuts it for me.

My latest project will be available for viewing in September/October and I have to say to you that if you like Nintendo games, you'll enjoy the visuals on this. I can tell you this much about it, the animations are superb, the game worlds lovely and the characters quite a lot of fun to watch...and not a shader in sight.

Keep your faith dude, get the design sorted and when the engine arrives you'll be all ready to go - in the meantime there is a whole lot of planning to do.

IPete2.

IPete2, I do have a thoroughly designed game already waiting. That's what I did while waiting for Max3D over the past couple years. I guess I could design another 3D game....you get the point...

Then again, waiting for something better may also be an excuse for a break in your development. Realisticly, todays 3D games reached a level, content as tech wise, that none of us bedroom coder inc.s will ever reach, at least not without a team of 10 professionals.

Then again I agree it may sound unprofessional when you tell people you're using DX7. Hey I got an idea, simply print "DirectX9c required" onto the box - welcome back, up to date appearance.

I think there are projects which look professional and still can be done by bedroom coders and there are projects which can't. You have to pick out those which can be done.

Beside of this i find it pretty hard designing a finished game design without trying out several stuff in between. Either you're a genius or i don't see this working out well...

Well I do have a full-time job so I'm not really "in development" like a lot of you guys are. But I still want to be able to make some 3D games with the latest features. Sorry to sidestep both yours and IPete2's insinuations but neither are accurate in this case.

Yep, it's not hard to come up with a design but when you're doing the whole shebang yourself it's nice to be able to experiment with your 3D game here and there. There's so many ways to skin a cat you know. As long as it comes to the finished product then it's all good.

Chroma, I appologise for seeming to insinuate your unreadiness, it certainly wasn't meant like it appears to have come across. Sorry, I wasn't trying to be disparaging.

It's refreshing to hear of someoe having a proper design doc ready to go, good one!

Like you I do feel the urge for the features of 9c and above. What about modelling and texturing, concept art, sound and other things for your project, are they also well under control?

Also, like you I love the power/simplicity ratio which B3d offers, and find it impossible to fnd anything remotely close. I would suggest XNA but it is rather more involved from my brief experience with it, although the more I do get involved with OO languages the more I can see where they crossover in philosophies and I like it.

If you do come across that "Holy Grail" let us know, please!!!! I think that my next investigation will be further inroads into XNA - I have a feeling that will turn out good in the end.

IPete2

Well I didn't mean it the way it sounded IPete2. Sorry.

I like the simplicity of Blitz. But I thought it was obvious that a 3D module for BMax would be a necessity. I'm kind of amazed it's been out 3 years and we haven't seen it yet. I would be happy with the functionality of blitz3d in bmax with the 9c features. It seems so simple to me. Is it just me or am I really asking for too much? Who can do this for us? Isn't 3D a natural path of evolution?

I don't want to turn this into an "OMG I want Max3D" thread. I'm really open to any viable solution.

Do you not understand what the word "metric" means? How do you measure success?

Yeah I know exactly what you mean, Avon and I have been here from the start (Avon being an original Blitz 2d tester) and we don't understand the move away from BMax 3D module either.

It is easy for me to say but in my mind, what Mark should really have done is ignore the Apple and Linux OS's and develop the next generation B3D language that would have probably halved the dev time and had every feature we now so long for. From what I can tell it's probably never going to happen now anyway and some other hotch potch game maker is where Mark seems to be headed.

This almost makes me as sad as the day I realised I needed a PC and no matter what I thought, the Amiga was dead and buried.

In the meantime I'm looking for a replacement, XNA is very promising but too early to tell for me personally. Leadwerks has an interesting approach and I may take another look there soon too, Josh is a very clever guy, and has been around a long time - and being a developer knows what we want in a game engine/editor suite - roll on CShop 6.

Whatever you do - do not get tangled into the luring web of Torque, well not this generation anyways, it is so contrived as to slow the defveloper down not speed him up.

A6 is weak (A5 was where I stopped), in fact anything which is short of the really McCoy (i.e. C# or C++) is never really gonna cut it, so check out Alan Wake or Crisis and dream on - along with the rest of us.

[Edit - I should add that I still absolutely love B3d and all it allows me to do - so I'm going no where else for my current projects].

IPete2.

I was saying that from day one that we heard it was going to be multiplatform I think. :)

Lol, there goes me being a smartass and I misread what you wrote yourself.

True, we don't have any accurate figures to see how it's actually fairing.

For my needs they could have left out a) the linux version and b) the gui module but definately not the osx version! :O)

linux might be nice in the long run but i just don't need it now. linux first has to grow up.

Indeed - I think the Linux version should have been abandoned.

It'll be interesting to see what DBPro X10 will be like (and especially if there will be a SDK version of it). If Lee can get things right this time (efficient machine code and better testing), it could good.

Hey, don't knock the Linux version!

If you ignore the IDE, setting it up to the point that it'll compile modules is way easier - particularly on a Debian based distro - than on Windows :)

For many years I have dreamed of a Game Engine with the latest DX10 features with a Blitz3D BASIC for Scripting (all at an affordable pricetag). In fact, I have suggested this to a few popular Game Engine Developers such as Garage Games.

Over the years I have come to realize that I'm strong coder and extremely weak 2D/3D Graphic Artist. I've accepted that even if such a Game Engine existed, I would not be able to develop the games I imagine alone.

Since this awakening, I have redirected my focus on the development of Games heavily dependent imagination, NOT Graphics.

I still don't see an economically successful future for Linux ON THE DESKTOP and/or as a gaming platform. Linux has fostered an "it's all for free" culture in its userbase, and that is the one reason why e.g. Adobe is not porting their products to Linux. And according to Carmack, even Doom 3 has barely made enough sales on Linux to cover the costs for the port.

Personally, I only care for the Mac version of BlitzMax. id Software and EA are now supporting the Macintosh platform (again), and others will follow. I'd say that Mark Sibly was ahead of his time when he made the decision to release BlitzMax for the Mac -first- and later for the two other platforms. Macs experience a lot of hype these days AND they have a paying customer base.

It would just be foolish for all of you guys to focus on Windows only. Unless, of course, you like to see your products flowing around on eMule and other filesharing networks. You know, the most successful "portal" for Windows software. (I just have to be cynical here.)

Leads me to another point: Why is it that I don't see almost no BlitzMax creations on the software portal of www.apple.com? It's the first place where Apple users look for software (links). Guess where I found most of my indie games for the Mac.

So what is BlitzMax really missing these days, at least in my book? Full Cocoa, multi-threading and built-in database support. Or let's put it that way: MaxGUI should have more gadgets for regular application development. In the database department, it should have an abstraction layer similar to ADO.NET and support at least SQLite and MySQL. (Yes, I do know Brucey's great modules. But the point is that they should become a regular BlitzMax feature.) And sooner or later you need multi-threading when you want to write responsive (GUI or server) software.

The "problem" is that BlitzMax then would leave its domain and become a serious competition for REALbasic or Visual Basic. And probably that is something that the Blitz guys don't want to be. I can understand why they prefer to stick with the gaming domain... At least that's not boring. :-)

Not true, theres lots of commercial software for linux, and if something is good enough to be better than the freebies then it sells.
Even people that make 'free' software like ubuntu, blender & open office make a fortune in other ways, merchandise, manuals, web advertising etc. If software is written from the start with cross platform in mind then porting it would cost much much less.

The reason adobe probably said that, is because their software was originally coded to be too reliant on windows and mac platforms, with no thought to any other markets.
Programs like this would practically have to be massively re-engineered. Corel on the other hand are making linux versions of its products (possibly involving a custom version of wine)

I reckon the Linux and Mac games markets (both account for about 15%) are prime new areas for exploitation as the windows market is saturated.

Ahm, you do know that there are millions of dollars pumped into Ubuntu?

From what i know i can say that the linux userbase for certain cases is surprisingly larger than the mac userbase but the conversion rate is much worse. I would say less trouble and better income on osx so far.

Exactly my point. Which makes it a commercial software enterprise, not a free one.

For some businesses yes but i really don't think for games. The last time i looked at Ubuntu it still wasn't able to detect the hardware of two of my systems properly, so nah not really an option. Let them invest some more dollars and just wait some further years and then let's see where the others are at this point.

which great known commercial software is there on linux, application wise, beside those that just were forced to have Unix versions for the simple reason that rendering/calculation farms are Unix based?

Just name me one and its "successfull" numbers.

There is no casual gamer market as there are to many tuxies doing their GPL stuff.
Sure, you always can say "if you have something better then ..." but you should not forget that many of the linux users are only linux users because its OpenSource and not because its Linux at all. They don't waste money on stuff not opensource.

As well, the opensource stuff, if good, ends up beeing in the distributions repository ... so you are finally out as a precompiled application will never end in there (you see how much they struggle including acceptable drivers in there to which they do not get source)


In the end, worlds tooo much problems to overcome for a the 20-30$ segment ... OSX on the other end has a growth rate like Wii when it comes to casual games and proper drivers so less problems for your user support ...
skyrocketing.

Problem with DX10 is that about 2% of gamers have the hardware to use it. SO its not really a priority on anyones list. DX9c has barely been pushed to its limits yet.

The real problem with next gen graphics is that Artists need to be able to code to have creative freedom with materials.

Last few weeks has been spent starting on new Aerial Antics demo with shaders, and it has been pretty painfull. In that time we have gotten two sophisticated shader model 2 shaders we are using through most of the game. Every other day we find a graphical problem that makes us have to rethink our shader and add an extra atream of data or extra tweakable parameter.

It's so different from fixed function, and every shader has to be specialized that its not at all easy for an artist to get to grips with. Not without an experienced coder to assist them.

I'm using a node based shader editor, which is very impressive. but without a coder like Jeremy I can't get my shader into the engine. Mostly because there is not a definitive unified standard way of creating your shaders, and not all engines load them in exactly the same way.

The tool does 90% of the work, but someone who can code and knows shaders has to do the last 10% that makes it compatible with your particular engine.

Theres loads though obviously not as many as are for windows, but here is just one list for a start:
http://www.yolinux.com/TUTORIALS/LinuxCommercialApplications.html
Lots of people (including me when i knew only windows) seem to think that it is some non-supported, mickey mouse free thing that no-one understands or cares about. Its not.
Eveyone seems to forget that linux is basically a rewrite of unix (which already had/has tonnes of industry standard commercial apps) so porting them is not difficult.

To be honest, if i was into a game, was enjoying it but it stopped after the first chapter saying the story continues... pay 10 quid and get the next nine chapters...
I wouldnt care if it was linux,osx,windows or a mobile phone. Id pay it.

Point being, what ever you think of linux / OSX and their users, they are still a large pair of undernourished markets for games.
Heres a few though: http://icculus.org/lgfaq/gamelist.php?license=commercial
Oh and btw, and ive been using it for 2 years, and i have never even used open source. I always try and get the binaries. Most people i know who use ubuntu use it because they prefer it to windows and some probably dont even know what GPL means.
In fact the only non BMAX compiling ive ever done on it was for a guitar FX program, and a driver.


Back on topic though,

I think the future (certainly for me) is to make games/apps that can be built on all platforms with the simple task of re-compiling the code.
ATM im looking at minib3d as im quite new to BMAX but once ive mastered that i will move on to something like OGRE or whatevers new at the time.

I'm at a loss here. I hope something comes down the pipeline soon.

Well, I had stopped searching for the ultimate and best 3D engine, since every engine seemed to lack something, either it was entity based objects, multi-platform compatibility, speed, quality, ease of use, affordable price, etc...

But then I tested the Leadwerks Engine for BlitzMax (and other languages), it had everything I wanted, plus much more. The fact that it's not DirectX 10 based, but OpenGL, allows it make better and faster graphics on todays PC's than with DirectX 9/10. I don't think there will be a better 3D/game engine in near future than Leadwerks, and if there is, I would like to know about it!

EDIT: I just saw this on the Blitz main page gallery: http://www.blitzmax.com/gallery/view_pic.php?id=1634
I think it shows quite good what speed and quality combo is possible with it.

Of course, ID Software and EA are doing their conversions "for free" in some senses. They have the cost of porting to the Mac, which will be cheaper for ID than EA, but beyond that, they don't have additional marketing costs because their marketing budgets are such that advertising for two platforms ( or three, for that matter ) is scarcely more expensive than doing so for one. That equation is going to slant heavily againt the small independent developer who is going to be spenting ( relatively ) more on porting and supporting, but is also going to be spending much more ( again, relatively ) on marketing in order to actually make the port pay off.

So while I'm not saying there aren't advantages for smaller developers, the advantages are much smaller unless you happen to get very lucky and just stumble into a boatload of free publicity on a platform with less competition ( for now. )

And very rarely has Blitz3D with DirectX7, come to think of it, if Blitz3D did have DirectX9, how many developers here would actually utilize it to its potential? My guess... not many!

Still, thats not the point really, everyone wants to play with the newest toy in the toyshop... I rarely do 3D, though, I'm currently toying with the idea in my head about doing my current project in 3D (When the 2D one is done).

My guess is, realistically, I dont need to expand past Mini3D, yes, it would be great for adding all the fancy pants stuff of DirectX9, but realistically, to myself... Mini3D should do.

One thing is sure, I'll not worry about it!

And thats the whole point, realistically, what can you do against what you want. I know I could handle mini3D as well as the transition from 2D to 3D, and I know I want to convert my game to 3D!

So, I'm sorted! :)

Most importantly, when it comes to the design stage, like my current project, I know what I want, I know what BlitzMax doesnt have... I'll either get over it, or under it.

Not related to 3D, but still a comparison:-

I had major beef with joypads in my current project, and Blitz's joypad setup wasnt ideal in its current carnation, dont get me wrong, the tools are there, but I must have been naive when I wanted joypad support and just banged code in to suit. Little did I know when I plugged a different joypad in it would go all to cack. So I had to write this to compensate:-

http://www.syntaxbomb.com/forum/index.php/topic,625.0.html

The design of the game had to change slightly, mostly the control section, but still, I encountered a problem, designed something to work around that problem... And away I went (Though, I was secretly gutted about it)! :)

And thats exactly what I'll do when it comes to 3D in Max!

I'll pick an engine, use it and if there's something not quite right... I'll get over it!

Either that, or I'll end up doing sod all!

Dabz

Unreal Engine 3 looks pretty sophisticated, but a HUGE pricetag.

It's not that huge. They offer a much smaller up-front license fee in exchange for a higher royalty down the road. The question is whether anyone here is capable of putting together a team capable of using the U3 Engine at all, let alone coming up with the quantity and quality of art assets you would need in order to justify using the engine.

having confidence is a must for indies because of all the obstacles they face. but one of the most important things that they must know are their limits. what can they and what can't they do? it doesn't mean that an indie can't develop a top notch game, it means that they must know their limits and not get over their heads. their world might come crashing down on it. i'm sure that all of you guys know this. one important lesson i learned is that u can't wait for people, u have to make what u have do, or just move on. i have been programming for 4+ years. i have'nt made any completed programs, just demos. i learned a whole from my days of using DarkBasic, Blitz3d, Pascal, C/C++. i remember the days when i wanted to badly program 3d games. when i finally got the chance, over the years the feeling just died. now a days i just focus on understanding different programming concepts, like OOP vs relational, like using tables instead of stacks,queus and the likes. i also wonder what the heck does windows do with all that memory, who the heck does apple think they are ("their operating system is not all that"), and im tired of all those freaking linux distros ("like 200+") and every month or so u are hearing of a new distro. with all the madness going on in the world of computers, a day is coming when everything will crash. but carry on folks and pay me no attention.

How can you even imagine making a game in Unreal Engine 3 when you can hardly make a game using Blitz3D? That is my question. It's beyond price tag. It's usability, understanding, involvment, efforts, persistence, perseverence, determination... Ever so often I see a new complete Blitz product off the dev lines, and that to me is a determining factor of language and tool usability. One of the latest I have seen is Tank Universal, which I have participated in Beta Testing. One man and his project, the ideal indie scenario.

What more do you want? It is paradoxal that people that usually scream for Direct X 9 features, expand the language here and there, actually won't even make the effort to learn C language, Direct X, Open GL, complex math functions, and dirty their hands in the endeavor. I am not saying it is fine to be satisfied with the little that we have, but instead, to move on and produce what we have in our fresh and jumpy minds, and to learn and develop, pursue our goals. I don't thin kit is that hard to do in a basic language that permits 2D, 3D and a host of functions and people around to help you out.

Note: This is not directly aimed at Chroma

Red Orchestra was made by an online team using Unreal Engine 2.

The point I'm making is that being deceived will not help a bit. It's not because you are a millionnaire that you are a happy man. It's not because your wife is a supermodel that you have conquered her heart. It's not specially because Unreal Engine 3 is around that you'll make a game.

lol

As for the "if you can't do it with DX7, you don't need DX9/10" debatte (why does no one like DX8 nowadays :O):

There are things you get easily and things you don't. I want AA and there are zero dev costs involved and it improves the quality a lot. Same with normal maps, easily done and common tech which works on almost everything. Not to mention how much better per pixel lightning looks. Do i have to take care fo the rest of DX9/10? Nope, still i do need DX9/10 for this. Just two examples why it does make sense...well, at least to me. :O)

As far as i know iD and Epic do native ports to the Mac and EA won't (Cedega?). iD Tech 5 is said to turn fine on Win, Mac, Xbox360 and PS3, U3 isn't.

There's no doubt Blitz3d is long in the tooth and that it would be of doubtful wisdom starting a big project with the language at this stage.

Having said that, I like the simplicity of the language and frankly, if it hadn't been as 'basic' as it was, I wouldn't have finished/released Tank Universal.

When it comes to TU v2 I guess I'll be looking at Blitzmax + rendering lib, possibly Halo's engine or something similar.

That's an easy one. Fast shadows, shaders, normal mapping (dot3 or whatever).

You're right. We don't want to take all the time to learn C/C++/C#. That's why Mark is doing this and that's why a lot of us are here.

This is exactly my whole argument. When we have a good 3D lib then we have the whole enchilada. It shouldn't be that hard...we're so close...

Hehe, I didn't even see this line until I was almost done replying above. So this wasn't an attack back at you at all _33. But I hope you can see my frustration.

I totally agree with Chroma.

I don't want to use the full set of dx features whatever the version. But I do want to be able to knock out a game or demo that looks nice without having to learn really complex low level stuff, hence working with a BASIC language.

I'd like to add a feature of a special particle sprite that automatically rolls all nearby ones with the same texture into into a single mesh and stuff. And those convex hull collisions too

It's a where's BMax3D thread, but in disguise.

Well, actually, it was indirectly aimed at you Chroma :P J/K ;) .... But, tell you what, I know a really talkable indie developper (SqueakyDuck Design) that is aching to release Cobra3D. His fingers are just so itchy... And it does have all sorts of shadows, inline assembler (say MMX, SSE/SSE2), GLSL... So, whatever happens, it'll be pretty good!

But, that won't hold me back, I have developpments to do, and I won't stop breathing for this promised miracle that we are all waiting for. Meanwhile, AShadow does exist, which has post process effects, water, and various neat little tricks. There is still good stuff to do in the current iteration of Blitz3D.

To each his own. If you're still happy with B3D then knock yourself out. I want to get a little more advanced (modern).

BMax + 3D + Physics lib. That's what I'm after. Any ideas? PLEASE FOR THE LOVE OF GOD. SOMEONE HELP ME.

Leadwerks Engine.

Well, depending on the size of your project, you can get all those things right now then move to newer solutions as they become available.

For example, you could use MiniB3D and one of the Newton wrappers. If you keep everything tidy, it shouldn't be too difficult to swap to another graphics engine if you want.

Ok thanks. For now I'll revisit MiniB3D and try out Josh's Newton wrapper.