I don’t know where to post this, so I’ll post it here.
This is a CryEngine2 realtime scene runs with 12fps on quad-core intel processor.
I really wish someone like Dloince JP to do this kind of eyecandies in BGE. I’m sure that the right person with the right knowledge could do as beautiful scenery as this in BGE.
ha ha! blender will never be able to provide this awsome graphics!
It’s not just a matter of technical problems, though. A huge amount of work goes into graphics of that quality, usually more work than a hobbyist is willing to invest.
A huge amount of work goes into graphics of that quality, usually more work than a hobbyist is willing to invest.
Well, I think this will be the real problem!
Technically Blender is good enough for something like this(especially with the soft shadows patch)…maybe not as big and not as much objects but a fragment of this could be done and I believe it will run in real time, pretty good.
Spectacular. But yes, as you said, personally I would have to be paid to invest that amount of time. I’m doing my best on Ayleth, and some of the individual elements of my game could rival the quality of some of the elements in those screens, but all together, hell no! Shadows add a lot, and for now, the Blender GE just can’t handle them as well as other engines.
I just get jealous when I see things like that 
And I can’t even imagine how much time that took
It would be nice to see an large interactive physically enabled environment with that lighting and explore it.
All I would need is a 1 million dollar supercomputer.:spin:
you know ive seen some pretty amazing work in the BGE one is Iron Steam
i think the BGE is capable of many amazing works it just takes a lot of time and effort, since most people aren’t getting paid to use it they have no incentives other than as a hobby. I will agree it is not on par of the most commercial cutting edge technology, though it is still amazing not being a completely commercial project.
+1
Further graphics enhancements to the BGE are, at this point, utterly useless.
Even if the developers could somehow compete on the underlying technology side, it’s very unlikely that we, as game developers, could ever truly compete in terms of content.
Game-play is the only area where we have a really good chance of surpassing the industry “big-boys”, but, ironically, the logic programming facilities in blender have advanced very little, while the few developers that actually work on the BGE are focusing on graphics more than on anything else.
Most of that would be possible in the GE excluding the shadows. However, for one person, it would take months even years to make that. The GE is fairly capable, but its FPS is terrible. Blender can have the best visuals in the world, but if it doesn’t improve it’s FPS, it’s worthless. :no:
The pretty eye-candy is all we need for the good gaming.
Moguri is pretty much the only dev. working on graphics capability right now.
NexYon has stated he’ll be starting the part of his GSoC project that enhances and improved the python API (already commiting the changes needed to get Yo Frankie! working), Nicks has gotten a good pathfinding solution mostly completed now and is working on automatic navigation mesh generation.
Dfelinto continues work on the RNA based logic brick interface which is supposed to replace the old crufty version, Benoit is currently spending his time as a GSoC mentor for Nick’s project.
In order to solve the architectural problems, as they relate to the scripting system, he will need to rewrite the core interface to the lower level engine functions to be exclusively python.
Is he doing that, or are these “improvements” on top of the old logic bricks mess?
Nicks has gotten a good pathfinding solution mostly completed now and is working on automatic navigation mesh generation.
That’s nice, but if I have to call that functionality from the crummy interfaces that are available now, it’s not really such a big benefit.
Also, path-finding was never the critical deficiency within the BGE - we already have a wide variety of A* python libraries that do the job quite well.
Substandard logic programming facilities are the main reason for why most people never actually find the need for more advanced path-finding solutions -> they never get that far, because all the logic woes limit them to something much, much smaller.
Dfelinto continues work on the RNA based logic brick interface which is supposed to replace the old crufty version
So, the logic bricks will remain as the core interface to lower level engine functionality?
The commits logs hint that you will create a navigation mesh via the modifier stack, also, you define an object as a navigation mesh through the object types UI and not the logic bricks.
NexYon for the most part is improving the python API in the vein on getting the sound functions and other missing parts back and new functions, and on top of what Moguri has done so you don’t need quite as many bricks like being able to do keyboard events completely in python.
Or if you still don’t like the direction the BGE is being taken and just want to bash the developers why are you still using the BGE?
BGE maybe cant handle that many polygons and the shadows for now, but to create HQ games you need to be an good arstist, it is not a matter witch game engine you are using if you cant even texturize a cube! This kids keep complaining about BGE because they dont know how to make a decent work.
And thank god there is the logic bricks, they are life safers for people like me, and i think that blender probably wont have 1/2 of its users if wasnt for the logic bricks, they give people that doesnt know nothing about scripting a chance to create their own games.
People keep complaining about the logick bricks on the system, why??? it hurts on them??? it make some substantial change??? came on god save logic bricks and ***** the people that are against them.
There’s nobody who thinks the logic bricks should go away completely, you’re misunderstanding the problem. Right now, in order to do pretty much anything in python, you have to connect your script to all the relevant logic bricks that actually do those things. We just want a separation of logic bricks and python, so you don’t need five or six sensors plugged into a python script just for basic stuff. If python could access the keyboard directly, for example, instead of having to go through a keyboard logic brick, it would remove a lot of the hassle of using python in Blender.
I understand that logic bricks are vital for the very beginners of the BGE, or those with no interest in making a full game (IE artists who want to showcase a character) but as soon as you want to do anything above the very basics, you’ll need python. Right now, the integration of python in Blender is not very, well, integrated.
Well, if you say so, I think that i prefer mutch more emphasis on graphics than on the logick bricks issue, but, you know mutch more about python than me, if you think that this is a priority i accept that otherwise we should complain about something more important since we all agree that the logick bricks are the blender game engine ’ s charm. I ll not even try to argue about the integration problema because i dont know it very well, I can create just simple programs on python but I dont create a single thing for blender yet but if this step is needed for now, and for the whole BGE existence why complain? its probably hard to separate stuff, waist more time on something that isnt is a real problem. If you guys hate this logick brick stuf so bad it probably cause some problem right? like a problema that cant be solved, like we have on the alpha shadows or soft shadows.
:rolleyes:
It’s called constructive criticism. Stop trying to make it sound like I “just want to bash the developers”, because it’s simply not true, and with all its flaws I still love blender, which is why I use it (along with other libraries/engines).
To recognize the problems that are there, and to question development decisions -> that’s what it means to care about blender as software.
Wise up.
@ Captain Oblivion
By and large I agree, no one wants to see the logic bricks go, because they do bring in a lot of people who would otherwise be intimidated by the rigors of game development.
However, having access to basic user input from some python object is not enough to solve the real problem. Even with that in place, you still have to run an always sensor to actually run the script, and, more importantly, you can’t (easily) use the basic OOP features that really hold the power of python as a programming language.
We need the ability to write a python class for a any given blender object, and then set methods with proper event callbacks, which would be run by the underlying system when they actually need to run.
That class we create should be the one tracked in the global game object list, and all the methods we have on that class should be available when we get a reference to that object, along with the basic game object functions that would be there by default, like setLinearVelocity(), or the worldOrientation setter/getter.
We need to think about the game as a python application, not just a “pitch” into the logic brick system.
Nothing worth doing is easy.
We need python as the core system, because we need the flexibility that arrangement would offer in the logic programming department.
I mean, the only area in which independent game developers can compete with big corporate shops is game-play, not graphics.
Therefore, logic programming facilities are very important, and their inadequacy in the BGE is a real problem.
Or, did you mean to say something else? -> sorry if I misunderstood you; your English is somewhat difficult to read.
No you understand right, i think that you just dont understand my point, about my english if you want i can speak in portuguese or spanish do you like the idea?
why you guys that know python more than us all dont try to implement a moded version by yourselfs? i think that all your geniality togheter can make the diference right? so join a team and implement what you want instead of wait for this “big problem” to be solved! I cant right now change a single line on blender s script im learning , im not a pro like you guys, and my problem need to be solved in c++ lenguage not in python so you “complainers” are mutch more close to your own goal than I am.
Good luck with that, and with the football world cup too!
However, having access to basic user input from some python object is not enough to solve the real problem. Even with that in place, you still have to run an always sensor to actually run the script
I wonder how easy it would be so you would only need the controller and not need the always actuator, a button that will ‘always run the script’?
I propose it’s done that way because you might need bricks and the inter-object connection capability for a few other things, but for systems that can be done completely in python you’d even eliminate the always actuator.
If it ever comes to a point where you don’t need bricks for anything and everything can be done through python only maybe even have a textblock in a tab above the properties tab where you put in the name of a script and it controls all the logic for that object.