Did anyone tried to create custom shader? https://docs.aws.amazon.com/lumberyard/latest/userguide/mat-shaders-custom-dev-intro.html
I think it’s market standard, if you leave everything hosted and managed by integrated AWS systems and your game hits 10.000+ players that is kind of bills you should expect; this really turn me off when I think of indie multiplayer… It’s just too expensive to deal with dedicated servers and building it all yourself is faaaarrr from easy or cheap too.
Would be if we could make a F2P online game and never bother of monetization, but as you can see then you would bleed money; that’s why F2P multiplayer games are nothing much more than cashgrabs and developers simply can’t avoid designing their games that way while still MANY go out of business after a year or so.
Btw, they give you 1 year free plan of very small datacap so you can beta-test and bank a community of ‘whales’ before they beging sucking your blood-monies from your veins 
After that they charge you, no matter what. The thing is… If you hit small you’ll owe them small bucks, if you hit BIG you still make close to nothing and they earn A LOT of money; your game will basically work to pay the server bills and you may keep a little from what is left. When you think of it is no wonder why EA always shuts down game servers after 1 year.
REally ? From what I have read on stackoverflow from the link you provided, corutines let you just setup task, which can be executed over duration of time with period. Which simulates behavior of concurent tasks (but they are still running on single thread).
For me it looks like setting timer with object and letting it run idefinetly.
I really appreciate what Amazon is doing with Lumberjack, especially the attention to scaling network. But, in my opinion, acquiring unique Content is problem for a majority of the Game Devs. Even Single Player Games not connected to a network need content. Systems to assist Game Developers in the generation, assembly, customization of 2D/3D/Animation/Audio Assets from *within *the Engine/Editor would put the Game Dev System into a unique position. Add the option for Devs to Publish their creations to Marketplace for sell to other Devs, and Presto!
I predicted that future games engines will support Massive Multiplayer Networking out-of-box 1]. Today I have a new prediction based on the speculation that Future Game Development Systems will make Game Dev even easier… as we get closer the ‘Make Game’ Button. High level features such as In-Game Content Management Systems are a step in that direction.
I actually dont find the programming that cumbersome. Have only been playing with it for a couple of days and already got my head around the flow of the engine. Its very different to UE4, but kind of sleek in its own way. Already working on loading custom meshes bypassing the pipeline.
Next on my list, as I have a need for some certain shaders
I don’t know… When I see a challenge hard to master, that is the moment when my heart beats.
Would be smarter to say ‘-no, thanks!’, but at the end I can’t avoid head jumping into it. My own enemy…
I think the only real improvement that UE4 needs as far as “one and done” is to have a better “one-click solution” to multiplayer / mmo stuff. Basically, we need the ability, in engine, to connect to various types of databases and then be able to directly upload to server. Perhaps even a “back-end designer”, similar to Dreamweaver… where you specify file structure / database structure / etc…
(yes, this wouldn’t be a “one click” solution… i know. it’s just a term)
Fun fact.
One of the file names:
system_xbone_pc.cfg
In C++, or in Lua?
C++ of course, don’t really like LUA, but will use it if I have to.
This is why I dropped UE but more so because lack of clustering. Don’t have the time to write the code for it and develop other things. Since no one interested is my project it is a one man endeavor so going backwards was the best option.
However, looking at LY here and seeing how this may be capable of doing what I want it to, I am considering it a valid choice.
Except… It’s still CryEngine and have abusrd problems with precision beyond 2km from origin point.
I don’t see why would you need something like clustered world, when you can’t have map big enough to support it.
Coroutines let you just write simple linear code. Instead of writing state machines with callbacks.
I still don’t see the appeal.
But on the other hand I utterly hate dynamic languagues.
The engine supports streaming of levels so it can be OW and huge at that which will require clustering. In BW engine the client issues are a major stumbling block with their indie release. (no occlusion, no morph targets and no source) but the net code is rock solid and can cluster 100s of servers if required. A single server for a couple people is fine but the world is not alive and persistent. Complex Ai requires considerable amount of processor so distribution of active code needs more than one machine to execute.
No it doesn’t. It support streaming content within single level. And more the point single level can have only single terrain block. So good luck with performance, when you have single 4km x 4km terrain chunk.
I tried out Lumberyard, the flowgragh’s don’t have any nodes for gamepad support like CryEngine did…and i had trouble installing Wwise authoring binaries for audio. They said it will improve over time, but i’m not going to drop Unreal Engine 4 any time soon…
Oh that I must have misunderstood. Well guess it is BW then with its sucky client.
And how have you exposed the properties to the Editor? I am really curious about that, in the ‘original’ CryEngine it was quite ugly 
or it could be like UE4 where you try the tutorials only to find the key commands have been deprecated. When you do figure which commands you should now be using the editor crashes on you anyway. I will be giving it a good try and testing its stability, which comes first in my book.