Eh. I use quick fake reflections all the time through texture “reflection” mapping mode. Granted it’s different from Sources cubemaps but that system was really bothersome to use as well with setting up the projection points while mapping and building the cubemaps in-game after every change you made. In the end you see them reflected in the scope of some particular weapon. Not really much of a deal breaker for me.
Bloom . martinsh did a great job on filters, but even though they are slow and not high quality. (because of lack of downsampling)
It runs pretty good for me. I’m not so keen on it though, it’s isn’t the point to blur everything in total white mess but a faint bloom always gives a nice touch.
3.How would you animate those trees without huge performance impacts?
The same way you do it on another engines? LOD, culling, instancing…There’s no magic solution to calculating infinite amount of deformation on BGE or other engines.
4.My bad again, but even with python there is a large cpu overhead, simply because the particles are not batched.
Getting carried away with particles is very easy. Fine tuning something like spawn frequencies, proper merging and layering will make it work. Making particles the wrong way in BGE is easy if not even probable.
5.By workflow I mean a window showing your prefabs, being able to modify one and the rest of them just update. Being able to have them all over you project, not just linking/appending.
But that is instancing? Blender has it? Group instance? Plus possibly a library that you could implement via some python.
6.Indeed, its just your opinion.
Of course it is. I’m not the one talking about who is right and who is wrong. Just bringing my perspective into it.
what causes people to leave bge halfway through a project…
I’ve tried to make a game with Blender… a complete/finished playable game - With proper logic, menus (UI), sound. Lost most of my time for on modeling. End up working with architectural company for real-time projects. Then, I forgot my python logic & need to reprogram my game again.
…and go for another engine?
No. I’m still with Blender. I can make a simple real-time fly-through for architecture, product demo, simulation. I can produce rendered images & animation. I can do video stitching & lots more… oh yeah… I can make games with Blender too… :rolleyes:
Sinan, im not that skilled in programming to use an engine without a lot of examples.
Back to the BGE: the limitation of the engine should tell you how far you can go! I dont like anymore the logic bricks, instead of using them, you should teach how to script your game, since every game have the same type of logic: A character walking, jumping, attacking, collecting items,. etc…
Hey everyone look at my amazing project!! And its just another fps…
I think the learning curve has a few hurdles that some just can’t jump…
Unless you learn to code, model, texture and make music and sounds… you have to rely on others, even if you learn it all, do you have enough time?
I am out of commission as far as testing and coding goes,
But my project lives… I hope to generate a constant revenue stream to the bge development.
Any master coder want to take a crack at heading my team? Right now we are bogged down by a 2.66a version compatibility, and a lack of code making it in the tech demo…
One amazing jaw dropping game could change the face of the bge… so I say
Don’t give up, join a team that has the skills you don’t, and learn by doing!!
blender is an excellent modeling program the best I have seen as of yet. I think that people leave blender game engine because of game engine speed or the lack of knowledge of how to use. I would say that people probably use blender more for modeling then export to another game engine.
On this, if you’re fully dependent on the logic bricks for a game, than yes, you will get a prohibitive mess of wires. This is why one needs to not only use how to make an efficient use of the state system, but also know the ins and outs of the Python API so as to minimize brick use and even eliminate the need to use some types like the property bricks.
In fact, theoretically, the brand new collision API (coded by both Moguri and Agoose77 and committed yesterday), may actually help one avoid the use of collision sensors entirely, this means that you can now avoid the use of collision sensors (or at least reduce them a bit) much like what you can already do with property and keyboard sensors. Basically, this means more ways to optimize logic, the only thing being that it would still be recommended to use actuators for certain advanced functions because the code they run is C++ and thus is faster than Python.
Creating it should be fast, cheep and much better then the original. It will show how good I am!!!
So lets look for some cheap materials. How about that:
This seams to be big enough (I’m not sure how to get it into my car, but I’m sure this is finally not a big deal).
Now I need a tool!
How about that:
It is fast! …ohh… I see it is to expensive … so lets look around …
and I get something cheaper: and
and the best… the original statue was created with the same tools (just the old and outdated version of it).
So you have all I need:
material
tools
myself
some spare time
So lets start… But how?
Ok I need some training first?
No problem I just ask the community …viola! I made this:
Hey this looks cool for the first ten minutes. Lets estimate when the David is done … 2 hours should be fine.
And I got …
after 2 weeks of “hard” work.
Somehow not really exciting :(.
Ok lets ask the community again. Surprisingly there are so many others with exactly the same problem. This can’t be coincidence … what do we have in common? The TOOLS! We all use Hammer and Chisel! That is the proof, we need better tools!
How about that?
Oh dear, this is perfect. It is really fast … but I require a special licence and a lot of money. Why can’t the hammer do the same?! Come-on hammer manufacturer why can’t they create hammers like dynamite … with that I could made not just one stupid David! I could make hundredths of Terracotta Armies … buhaahahaha…
Yeah, exactly! What I feel we need more of is that whenever blender doesn’t provide an industry standard game developing toolset built in like unity or unreal we should give a collective F!@# YOU to anyone who wanders on here expecting that. Thoughts?
Kamiyama; ganging up on people who switch to Unity as if this community prizes itself on mob-mentality would only reflect even worse on the BGE itself, people would then think that there are things that the BGE community is trying to hide and then proceed to list the community as a major weak point of the engine.
A toxic community that adopts a behavior akin to animal instinct would simply do more harm to future BGE use and development than good, one of the major ways to improve the engine and move it forward is to identify weak points in the engine so they can be addressed.
I started to take a significant interest in making games a couple of months ago and turned to the BGE. As a person who teaches programming for living I want to use scripting for game logic and don’t want to be bothered with logic bricks. I have Mike Pan’s book on the BGE but he’s using logic bricks quite a lot. I made a few enquiries here and got a couple of answers after a few days suggesting ways to minimise the use of bricks, but looking for good instruction on scripting without bricks led me to a brick wall. Sure I can attempt to learn the API and work it all out from scratch but I found a better way. I switched to Unity. Fifteen hours of insrtuction on Lynda.com (Unity Essential Training) plus a few other titles, several books from Amazon that address my needs exactly, and I’m even considering the upcoming workshop on CGSociety on Scripting in Unity. I’ve said a few times (on other topics here) that the biggest lack in Blender is training, compared with other platforms. There is good stuff here, but for what I’m specifically looking for now, no. In the past I would have worked it all out myself, been there and done that. But now I just want to get on with it.
I’ve given up halfway through multiple games now, and have regretted it each time, leaving for things like CryEngine 3 or Unreal Engine 4 of late. I’m not exactly terribly familiar with these two other game engines, so I’ve always come scurrying back to Blender, deciding to make a prototype on it.
Anyway, I’m determined now that I will finish my current project - similar to Age of Empires mixed a bit with Total War. My main problem right now is programming an adaptive LOD that will actively change the terrain mesh detail as you move around it.
Ok. I actually took time to learn the syntax behind the python. That’s what programming is and that was what was standing between me and my goal.
There isn’t even an easy to use system on Unity, you’ll have to learn the same thing as with python so SCA is just a bonus on BGE. It’s all documented online one form or another but sure Unity has more handheld guidance as it is a commercial product.
Unity’s workflow stopped me on it. Go back to model and animate on Blender, then set up level lighting and clipping meshes in Unity? Why wouldn’t one do it all in BGE?
and 
