hi,
i am just curious about this...
what do you think is the reason that there aren't more high profile games done with irrlicht?
what is missing from this engine?
do you think that development goes into the right direction?
what is missing from irrlicht?
like a whole next-gen deferred pipeline which I planned for Blooddrunk?
Honestly... I'd say the fixed function pipeline gets in the way (Horde3d is faster)
Shadows... no other official way than stencils which are crap. Xeffects wont ever do the job because its quite slow and doesnt have the LoD optimizations for outdoor scenes. PSSM was done by me and others but its a tacky implementation each time, I am creating my implementation (to use in Blooddrunk) but it will be closed source because someone is paying me $100 to make it for their game (Blooddrunk is open source but this part will be "protected"). I am hoping to create a full implementation of the PSSM + LiPSM + VSM for ultra detailed soft-shadows in time for "Screenshot of the Month February"
Honestly... I'd say the fixed function pipeline gets in the way (Horde3d is faster)
Shadows... no other official way than stencils which are crap. Xeffects wont ever do the job because its quite slow and doesnt have the LoD optimizations for outdoor scenes. PSSM was done by me and others but its a tacky implementation each time, I am creating my implementation (to use in Blooddrunk) but it will be closed source because someone is paying me $100 to make it for their game (Blooddrunk is open source but this part will be "protected"). I am hoping to create a full implementation of the PSSM + LiPSM + VSM for ultra detailed soft-shadows in time for "Screenshot of the Month February"
-
Radikalizm
- Posts: 1215
- Joined: Tue Jan 09, 2007 7:03 pm
- Location: Leuven, Belgium
It's not about what's 'missing' from the engine, it's about the design decisions the development team are taking
It's not irrlicht's goal to be a next-generation engine which runs on all the latest hardware, irrlicht's goal is to have an easy to use architecture with a high stability on a wide variety of platforms, leaving the interface open for people to expand it with next-generation features
The problem here (if you could call it a problem) is that the majority of the people on the forum do not have the ambition to create engine extensions and provide support for them, they'd rather build games using the engine
There are some very impressive projects done with irrlicht (even with older versions of the engine), but the developers just do not publish the engine extensions they wrote because maybe they don't want to support them, or maybe they want to keep the technologies to themselves
If you want to do basic graphical stuff, vanilla irrlicht is the way to go, if you want next-generation features you'll have to write them yourself (or find someone willing to write them for you)
It's not irrlicht's goal to be a next-generation engine which runs on all the latest hardware, irrlicht's goal is to have an easy to use architecture with a high stability on a wide variety of platforms, leaving the interface open for people to expand it with next-generation features
The problem here (if you could call it a problem) is that the majority of the people on the forum do not have the ambition to create engine extensions and provide support for them, they'd rather build games using the engine
There are some very impressive projects done with irrlicht (even with older versions of the engine), but the developers just do not publish the engine extensions they wrote because maybe they don't want to support them, or maybe they want to keep the technologies to themselves
If you want to do basic graphical stuff, vanilla irrlicht is the way to go, if you want next-generation features you'll have to write them yourself (or find someone willing to write them for you)
-
fmx
^ totall agree with you, sums it up nicely.
But IMO irrlicht will need to evolve on newer hardware so it can be taken more seriously.
I think most beginners feel daunted when they see irrlichts way of doing things for the first time, and decide (rashly and unfairly) its not "good" or "easy" compared to other engines.
On the flipside, advanced developers become frustrated when irrlicht hinders them and they decide to jump to something else, simply out of frustration and not because a solution wasn't available
Adding new features and extensions wont be enough, the process of developing custom extensions needs to simplified and at the same time provide more flexibility with low-level details (eg flexible vertex format control, easier render-order management of nodes, generic post processing frameworks, etc etc)
@devsh
you really dont have a clue how much your skills are worth, do you?
I dont take sides and I have no idea who was kind enough to offer you $100, but I would happily pay you double that just to keep your project opensource so everyone could benefit from it (its not an offer BTW, just a fact)
But IMO irrlicht will need to evolve on newer hardware so it can be taken more seriously.
I think most beginners feel daunted when they see irrlichts way of doing things for the first time, and decide (rashly and unfairly) its not "good" or "easy" compared to other engines.
On the flipside, advanced developers become frustrated when irrlicht hinders them and they decide to jump to something else, simply out of frustration and not because a solution wasn't available
Adding new features and extensions wont be enough, the process of developing custom extensions needs to simplified and at the same time provide more flexibility with low-level details (eg flexible vertex format control, easier render-order management of nodes, generic post processing frameworks, etc etc)
@devsh
you really dont have a clue how much your skills are worth, do you?
I dont take sides and I have no idea who was kind enough to offer you $100, but I would happily pay you double that just to keep your project opensource so everyone could benefit from it (its not an offer BTW, just a fact)
-
Radikalizm
- Posts: 1215
- Joined: Tue Jan 09, 2007 7:03 pm
- Location: Leuven, Belgium
As devsh mentioned, the fixed function pipeline is actually what is holding irrlicht back the most I think, but on the other side, a programmable pipeline would create a very steep learning curve for completely new users (although a programmable pipeline is just awesome once you get familiar with it)
Irrlicht would also need a huge restructure to support a programmable pipeline, as you can see in some of the DX10/11 projects done for irrlicht they really have to come up with some hacked together interfaces to let irrlicht's base classes work together nicely with a programmable pipeline
This is also why I started my own rendering engine project in my world simulation/game engine, just because I felt that writing all the extensions and engine modifications I needed for irrlicht would take me only a little amount of time less than building a rendering engine from the ground up with only the features I need
For example, I'm not interested in support for mobile or embedded devices, nor do I want to support really outdated hardware, although I'm structuring the engine in such a way that support for different platforms could be added transparently later on
But that aside, irrlicht is still a very good and powerful engine with a small learning curve, but as I said before, if you want next-gen results you'll need the knowledge on how to implement these yourself
Irrlicht would also need a huge restructure to support a programmable pipeline, as you can see in some of the DX10/11 projects done for irrlicht they really have to come up with some hacked together interfaces to let irrlicht's base classes work together nicely with a programmable pipeline
This is also why I started my own rendering engine project in my world simulation/game engine, just because I felt that writing all the extensions and engine modifications I needed for irrlicht would take me only a little amount of time less than building a rendering engine from the ground up with only the features I need
For example, I'm not interested in support for mobile or embedded devices, nor do I want to support really outdated hardware, although I'm structuring the engine in such a way that support for different platforms could be added transparently later on
But that aside, irrlicht is still a very good and powerful engine with a small learning curve, but as I said before, if you want next-gen results you'll need the knowledge on how to implement these yourself
I'm a beginner and do not understand much of irrlicht but I will try to tell my
opinion about this...
I think there should be two kinds of versions of irrlicht.. a Beginner IrrLicht and a Pro one. both free and opensource. Then the users would be able to choose between an easy, nice learning curve ( not to steep), with not great features ( less realistic ) and more user-friendly irrlicht ( for beginners) or a more advanced, great features, maybe steep learning curve and difficult irrlicht for experienced programmers
What do you guys think about my opinion?
opinion about this...
I think there should be two kinds of versions of irrlicht.. a Beginner IrrLicht and a Pro one. both free and opensource. Then the users would be able to choose between an easy, nice learning curve ( not to steep), with not great features ( less realistic ) and more user-friendly irrlicht ( for beginners) or a more advanced, great features, maybe steep learning curve and difficult irrlicht for experienced programmers
What do you guys think about my opinion?
“Any code of your own that you haven’t looked at for six or more months might as well have been written by someone else.”
(Eagleson’s Law)
(Eagleson’s Law)
-
Radikalizm
- Posts: 1215
- Joined: Tue Jan 09, 2007 7:03 pm
- Location: Leuven, Belgium
This wouldn't make sense, since the devs would have to do twice the work they're doing now, they'd have to provide support for both engine versions, they'd get feedback and user problems for both the engines on the forums, etc.
Also, this would probably create confusion among newer users about which engine version they should use
One rendering engine project is complex enough to design, implement and manage as it is, having to keep track of 2 engines at a time would be more than a full-time job, and we have to remember that irrlicht is developed without any sponsoring and in the devs' spare time
Also, this would probably create confusion among newer users about which engine version they should use
One rendering engine project is complex enough to design, implement and manage as it is, having to keep track of 2 engines at a time would be more than a full-time job, and we have to remember that irrlicht is developed without any sponsoring and in the devs' spare time
(i think he'd agree)Working with irrlicht is as easy as it can be. I'm not an expert in c++, but as long as you know the basics you're good to go.wahagn wrote:I'm a beginner and do not understand much of irrlicht but I will try to tell my
opinion about this...
I think there should be two kinds of versions of irrlicht.. a Beginner IrrLicht and a Pro one. both free and opensource. Then the users would be able to choose between an easy, nice learning curve ( not to steep), with not great features ( less realistic ) and more user-friendly irrlicht ( for beginners) or a more advanced, great features, maybe steep learning curve and difficult irrlicht for experienced programmers
What do you guys think about my opinion?
And I agree with Radikalizm, your post doesn't make sense. Instead of whining of how hard it is to use irrlicht you should go and learn c++ basics (from start till about templates), after that you wouldn't really think of saying such things.
You can't really start programming without learning the language you're going to use, its like going for a first time to the gym and trying to do 100kg bench press, or trying to speak Chinese without knowing the letters.
Working on game: Marrbles (Currently stopped).
serengeor wrote: Working with irrlicht is as easy as it can be. I'm not an expert in c++, but as long as you know the basics you're good to go.
And I agree with Radikalizm, your post doesn't make sense. Instead of whining of how hard it is to use irrlicht you should go and learn c++ basics (from start till about templates), after that you wouldn't really think of saying such things.
You can't really start programming without learning the language you're going to use, its like going for a first time to the gym and trying to do 100kg bench press, or trying to speak Chinese without knowing the letters.
I didn't say irrlicht is too easy or too difficult
... What I'm saying is that now irrlicht is a bit easy and a bit pro in one package and that you could divide irrlicht in a very easy and very hard package..btw: I also agree with Radikalizm on the fact that it would be a lot of work... but not useless because then the easy version would be more a sort of training where you could get comfortable with irrlicht and the pro one for serious projects... that's my opinion on this question, so: by doing that you could make great features for the pro version so there would come serious poject wich would get noticed and irrlicht would get notice too and still keep irrlicht user-friendly and easy at the same time
“Any code of your own that you haven’t looked at for six or more months might as well have been written by someone else.”
(Eagleson’s Law)
(Eagleson’s Law)
-
Radikalizm
- Posts: 1215
- Joined: Tue Jan 09, 2007 7:03 pm
- Location: Leuven, Belgium
Irrlicht is not an educational library, and it is generally not-done in software design to make an 'easy' version of the software and a 'hard' version
A library is not meant to cater to specific programmer groups, a library is meant to get a certain job done, and it's the programmer's job to master the library so he/she can use it to build an application
If however a concept or technique would be implemented in a library in a non-logical way or if there would be an easier way to implement it with the same results, the programmer could notify the library developers about this and suggest this better implementation, but you cannot make a concept or technique easier than it actually is
And as I mentioned in my previous post, it would be close to impossible to manage 2 libraries, implementing advanced next-generation features is not something you can do in a day, you need to do research, read papers, check out other implementations, watch presentations, etc.
Something I forgot to mention also in my previous posts is that there is no such thing as a 'one implementation fits all'-solution, large outdoor environments for example do not use the same LoD optimizations as indoor scenes, and I could go on like this for a while
Irrlicht does a great job at being a flexible and powerful base platform, but (and I feel like I'm repeating myself here) if you want to get advanced features, it's all up to you
A library is not meant to cater to specific programmer groups, a library is meant to get a certain job done, and it's the programmer's job to master the library so he/she can use it to build an application
If however a concept or technique would be implemented in a library in a non-logical way or if there would be an easier way to implement it with the same results, the programmer could notify the library developers about this and suggest this better implementation, but you cannot make a concept or technique easier than it actually is
And as I mentioned in my previous post, it would be close to impossible to manage 2 libraries, implementing advanced next-generation features is not something you can do in a day, you need to do research, read papers, check out other implementations, watch presentations, etc.
Something I forgot to mention also in my previous posts is that there is no such thing as a 'one implementation fits all'-solution, large outdoor environments for example do not use the same LoD optimizations as indoor scenes, and I could go on like this for a while
Irrlicht does a great job at being a flexible and powerful base platform, but (and I feel like I'm repeating myself here) if you want to get advanced features, it's all up to you
@wahagn Irrlicht is as easy as it gets. I can't think of anyway to make anything easier. To be honest you're the only person on this forum to even mention that irrlicht is hard. Perhaps you lack the skills needed to use any library?
What would you even suggest to be in each package? Also if you divided irrlicht into 2 separate libs, that means you would have to use 2 to do the same thing as one. It would be a huge pain in the ass to everyone but you. Irrlicht has been around for a while and is popular for a reason. It is good the way it is. I personally think the only thing that desperately needs improvement is the animation system.I didn't say irrlicht is too easy or too difficult Confused ... What I'm saying is that now irrlicht is a bit easy and a bit pro in one package and that you could divide irrlicht in a very easy and very hard package..
multum in parvo
This is the main problem, we need solid contributors who can not only spend the time researching and implementing such features, but are seasoned developers and have good OOP design skills and will carefully refactor their code to fit into Irrlicht with minimal API disruption. That's asking a lot.Radikalizm wrote:implementing advanced next-generation features is not something you can do in a day, you need to do research, read papers, check out other implementations, watch presentations, etc.