Share Ideas!

hi , how many times have we seen threads where people mention
what is wrong with gamestudio and in each of these threads
you would also find the people that at the same time say :
"hey i love this engine ", "hey this and that is acceptable"
"could it be graphics ,workflow or ..." (ratchet your being missed already) ...

so i have this idea and perhaps jcl would take it to heart..

why dont we in this thread list all dislikes and likes of the
current versions of gamestudio ,wed ,med...

and ideas for what would be great to see in the comming future for gamestudio ?

there are many things i have noted among the common requests for gamestudio
developement future ..

some of these are things like

a beautifull collection of shaders almost as the athour of shade-c was doing ,
did you notice how many responses there were to the shade-c package? that alone
should show you that it is indeed something we need because not everyone is able
to create beautifull shaders and i have to point out that pre made ready to use
shader packages is a big thing it takes alot of work away and makes things so
much easier for any person hoping to achieve a good look in a short amount of time!!

a terrain editor with the ability to paint onto the terrain , that is a big feature
that all of us would welcome and it would certainly inspire more terrain based games
being made for sure ! ,now editing like that is great but we also need streaming
of these joined terrain's so that we could really develope large beatifull worlds .

some sort of build in path finding system ,as game designer we need a pathfinding
system integrated and ready to use ,there are many possibilities for this as there are
many diffrent way's pathfinding but even a simple system would suffice as long as it
takes the work away from people who dont want and cant and dont see the need to develope
it themselfs ,pathfinding should be a standard among engines ,even it would be some
simple a* system with a floodfill/node edit tool to make things easier.

a gui editor ! this is one of the things i would also want to see in gamestudio ,
a editor where we could design gui windows and layouts ,save it to file and be able
to load it into gamestudio and just go and code our events , lbgui the author of that
sure saw the need and as always it is the users that develope and try to meet the needs
of other users ,i for one must say that any person that contributes tools for gamestudio
is certainly a treasure in the mids of sleeping giants !!..

there are so many people on this forum that saw the need for tools and missing features
and i encourage these people to keep up with it ,only to mention a few of the projects
shade-c ,lbgui ,all the script editors ,all the templates , sivan with his builder

there are many many more to mention but to much for this post...
/////////////////////////////////////////////////////////////////////////////////////

what would you like to see in the future ??????
what do you like about the current ???
what do you hate about it ???

and we may have users here that see a need for tools to develope
if they want to mm?

last night I also thought of time to make a wishlist thread in connection with the new WED , as its 75% state is probably a stage where it has meaning.

I fully agree with the statements above.

to keep up with other engines' toolset, an advanced terrain system (lod, streaming areas) with some vegetation (and other object) placement functionalities is needed today, in basic editions too, to help fast placement with auto alignment etc. and some optimization afterwards e.g. by mesh merging of models assigned as grass (probably easier to utilize than instancing).

I expect a more integrated editor with much more visual editing possibilities, for model material and shader settings, particle effects, gui, navmesh pathfinder, beside keeping the good old properties (fast object prototyping by blocks, templates for beginners in a few genres like rpg and shooter). and an easy editor plugin system.

yes, and an advanced shader system that smoothly works together is also a must. writing your own optimized shader framework is a bit overloading.

"fast object prototyping by blocks"

indeed !! do not throw away the blocks they are very very usefull

another thing i would simply love in wed is the ability to lightmap
any imported mdl model!!! (i have not seen this done with wed)

Something different I'd like to have is direct-x 10 / 11 support.
Yes, there might be more important features but DX9 is more than just outdated (11 years old).

Also, DX10/11 has some great features and is faster:
My brother's application needed 900mb ram and rendered at 180fps. After just switching to DX11 it needs 300mb ram and renders at 800fps.

The Acknex gods are on your side. From the Forecast:

@JustSid
As far as I understood him he meant lightmapping ON models, not FROM.

Kartoffel@ I guess lightmapping ON models without FROM them is going to be a little bit weird

I would like to see some improvements on OBB system, like different shapes (CAPSULE maybe), for all c_ functions.
Plus it could be nice to see some physX3 implementation, instead of the current one, cause it's integrated sloppy and well.. it's already outdated.
All-in-all I have nothing left to add, Wjbender nailed it pretty much effectively. But I'm afraid, that this thread is going to be ignored pretty much..

PATHFINDING,
IOS, ANDROID SUPPORT

What I actually like is the whole absence of an engine tied project management. It's supereasy and straight forward. You have your script and your files and you access them the way you have them stored. Couldn't be better. (No, I'm not ironic or sarcastic. I really like it this "oldschool" way.)

I hope the new WED brings some tools for terrain creating. Everything about that is already said.

For MED: I usually use it as an importer and finetuner for MDL files. And I wouldn't care if the abilities for modelling aren't further developed, if instead the workflow for importing and handling is improved. There will always be better tools and I don't think MED will ever be able to catch up. That work is better placed on engine features.

Well my overall answer to this topic is that I'd like to see 3DGS arrive in this decade. In my opinion there has to be a turn of tides in terms of the general orientation. I mean look at the homepage and read what's there - there are quite some things where you notice that the mentality of 3DGS got stuck many years ago. As an example:

That is said pretty much at the beginning of the engine description on the homepage. This worked back in the days when software like Blender didn't exist or wasn't nowhere as advanced as it is today and where professional software like MAX did cost a lot more than it does today. I mean think about it - here they try to encourage their potential customers to use MED instead of whatever else they are using. These times simply are over. THIS is how the industry works nowadays. You fire up your software, you create a high- and low-poly model, you create a diffuse, specular and normal map plus maybe animations or even LOD stages (if you don't want to use engine internal LOD systems). Then you take your asset and expect the engine to accept all of this with no sweat. This is where 3DGS in my opinion falls flat as it never really adapted to this trend...

I mean this is no AAA magic. Pretty much everyone expects an engine to immediately handle a model with a normal and specular map. That also becomes a problem when looking at the editions: The Extra version (you pay 100 € for) has no shaders - they don't become available until the 199$ Com edition. I think this really is horrible as we in my opinion now really are talking about the most basic tech level an engine should deliver nowadays and I'm not even mentioning that the handling of said shaders also by far isn't that "self-evident" even if you have the Com edition. 3DGS needs a "fire and forget" solution here. You want to use a diffuse, normal and specular map? No sweat - just tell which textures to use and here we go. Works with the lights you add and in general is a no brainer. That especially is needed for those "beginners" the homepage also talks of who are meant to use prefabs in order to click together their games...

This sort of also applies to the resource packer. I still don't get it why you need to buy the Pro edition just to prevent people from seeing your raw files. Not mentioning that this leaves an extremely unprofessional impression...

Maybe a small addendum to me quoting the 3DGS homepage as this probably should be the first and most easy thing to fix. There are quite some sections that should be revised or even removed. I'll name just two examples I won't even comment on:



So with 3DGS getting a new orientation what would I like to see? Well at first I think it's way too late for the mobile market and the Acknex engine also lacks certain qualities to really compete. I know that this actually is in development but I really doubt that this is what 3DGS desperately needs right now. There simply are so many competitors from specialized lightweight solutions to full sized allround engines like Unity all offering very efficient workflows and literally years of experience with those mobile platforms. I don't think mobile deployment for 3DGS ever will go beyond a "nice to have" feature for the majority of users...

I think 3DGS should aim for a niche. It's no use trying to compete with an engine like Unity with its in comparison insane level of workpower nor has it big chances on the mobile market as there don't just exist competitors like Unity but also all those smaller engines already (I can't remember good examples right now but there for example exist engines just made for 2D games which are very efficient at their job both in terms of performance as well as workflow for such games). In the end I can imagine 3DGS going down the "FPS Creator" route where it aims for newbies. What would be needed here are good editors together with good templates. The editors (hopefully) already are changing to give a way better / more modern workflow and the templates always have been there too although they should be a bit less tedious to use and understand. Basically this would be about what Unity sort of offers when thinking of those Kits they sell in their asset store. Things like ORK or the Playmaker are what I have in mind here. Especially the later with easy to understand node based development is nice (and sells well) and sort of ideal for the development for newbies as it offers way more potential / influence than just applying template scripts. 3DGS should opt for such easy non-programming methods. I think THIS LE3 Flowgraph video shows how nice such a node base system works and gives extremely easy access to gameplay logic without a single line of script. I also think 3DGS especially needs this as counterpart to the lite-C scripting which as this proprietary C-like language really isn't all that appealing to programming newcomers...

EDIT:
Maybe as small addendum: Once 3DGS would reach a state where it has those nice editors, nice node based non-programming system and extensive templates together with some tunes to the featureset going for the mobile platforms might make sense as people would be able to create not too complex games in no time. "One step after another" is what counts here though and so the basis for this high attractiveness for creating simple games should exist before the relevant export options. Having them now means that all engine changes already have to be double checked if things still work for the mobile platforms (eating up development time 3DGS doesn't have in abundance) thus slowing the development plus the platforms we get now might not exist in that shape when all those nice engine / editor features are done (which again means precious development time taken away in order to update the exporters to the latest iOS, Android or whatever version)...

So this is not about me hating mobile platforms but I think that there are so many competitors in this market that a newcomer needs a good concept / good reasons in order to somehow stand out and I don't see such a potential for the current 3DGS...

LE3 still uses AABB.. some say it's so outdated. In my opinion it's could be still great to have it.

@Toast
well said, I hope jcl doesn't ignore this whole thread...

Well, I think GS isnt that bad, but its simply a bit old and outdated at all.
The first thing would be DirectX10 or 11 support since I know its much faster, also Kartoffel said this.

Well I don't see why DX11 would help much. Adding DX11 support doesn't flip a switch and you get great performance. Actually it doesn't boost the performance by a lot as far as I know. From the switch itself I wouldn't expect more than a single-digit percent improvement. Maybe if someone skilled in writing shaders would jump into the fray there might be some further improvements but even then I don't think that this is the engine's bottleneck. Way more important than the simple featureset are how LOD, culling and in general the scene management is handled - that's where the real bottlenecks and performance improvements are. That's also why the DX11 switch in Unity did nothing groundbreaking in terms of performance...

So while DX11 support definitely should be on the list for future improvements I don't think this is what holds 3DGS down. I as described think the first engine related problem to address is the general workflow and toolset...

Yeah gs is cool ..it just needs work to bring it up to date

with time determination and work , it could become what it
needs to be but with that said i would not know how
good it was coded , would it be an extremely difficult task to
upgrade or was it coded well enough for extending/upgrading ?

Everyone of us knows it is hard work and takes time
Like has been said "steps" and the only way to move forward
is step by step ..every single feature and improvement
pushes it forward and hopefully into an exciting direction..

There are so many engines each with theire good and bad points
i dont think we should be comparing engines against eachother
because it boils down to what you need and they are theire to choose
from if you need unity then use unity if you think its easier to
go with gamestudio or which ever engine then use that engine
the great thing about engines not being the same is the options
you have to find what you like and want and need..

Yes i agree gamestudio has fallen behind with features and tools
and it really needs to step up..

maybe dx9-dx10/11 performance difference is engine dependent too, in another engine I realized significant fps boost (jumped from 22-24 to 38-41 in average) when the same c++ code was compiled in dx10/11 mode (of course using the same level and settings). I have not tried out unity4 yet.

I would also like to see DX11 support (please no 10!!)
In my opinion a gamestudio IDE would be freakin' awesome. Not like Visual Studio but more for gamestudio. Like combining a coding environment with a level editor, model editor and interactive shader editor (change code -> see results)

I don't want a change of the programming language, i really like lite-c. But also a better support for other languages would be great. Or a good scripting support like Javascript, Lua, ...

I'm not sure if it's engine-dependent but I've already said what I know:
my brother's application (programmed in C#) runs 5x faster (900fps) with DX11 and needs just 1/3rd of the ram usage (mostly geometry-data) comared to dx9.

Well there of course are some areas which got some nice improvements or make a more efficient coding possible. When we're talking about an actual use in a game under "normal" conditions the DX11 performance advantages should shrink down as usually the scene management is what becomes decisive for the performance. So when speaking of a general performance boost through DX11 I wouldn't bet on huge improvements. I mean you can try playing AAA games in DX9 and DX11 mode as that's possible in some titles. Changes in shader code and additional DX11 features (e.g. better shadows) aside you won't be able to skyrocket your fps. Improvements are definitely there but not in a magnitude that I'd say that this is what 3DGS needs to become a beast...

What I would love to see is:
-DirectX 11 support (as I already mentioned some time ago)
*support for geometry and tesselation shaders
-The possibility to render into a depth texture which can be used for shadow sampling for much cheaper PCF and thus a lot faster shadows.
-A new WYSIWYG WED.
-Culling of objects hidden behind terrain.
-more basic, easy to use templates, something like the A5 ones, but cleaner
-better shader library
-some pretty default assets

What is great about gamestudio is that you can actually learn a lot about game development and programming with it as I and probably many others here did, because unless engines like Unity it feels much closer to what is actually happening and still the API is much easier than those of most other engines.
Sure in other engines, you can just use stuff and it works, but most users will never understand what is actually happening there.
Also programming for gamestudio is fun, much more fun than with any other engine I used so far (Unity, Shiva, Irrlicht), not to mention the really streight forward and great to use post processing system.

-an easier way to set up views which render only particular entities like bounding geometry for lights by f.i. filling a view associated list with entities
-texture coordinates which are always in the range 0..1 for pp-effects, which is not the case atm. I don't see much sense in this design choice. It annoys me.
-making it possible to use a custom vertex shader for pp-effects

The first one would actually be sufficient.

-a list for entities which can contain multiple materials and associated views, for render view dependend materials

I really don't like to handle view-dependend rendering over material events.

And of course DX11.

Do you know the following approach?

VIEW* my_view with NOENT and NOWORLD set, then after level_load
my_view.genius = ent_create(NULL,nullvector,NULL); //or any other entity
and finally set every entity's parent pointer to my_view.genius which should be rendered by that view.

No, I actually didn't. ._.°
That's nice! Thanks for the hint.

@Superku: OH, what a gem of a tip

The best thing for me would be a publish option for more systems like MAC or consoles!

That would be nice, indeed but I think at the moment they should concentrate mostly on the PC.
More platforms require A LOT of work but they won't make the engine better and I think only a few people would use this.

There's so much outdated stuff which really needs to be renewed. (the default lighting system didn't change since A5 afaik)

So, the most important thing in my opinion is, that the Acknex Engine will reach a state where it is competitive to other engines.

Please keep in mind that you have to support this platform as well, which means that you need hardware to test your game on and you need to know the platform well enough to give support to your users.
This is coming from someone who is currently working on a project that runs on all major desktop and mobile platforms, and I wouldn't dare to support Linux and Android if I wouldn't work with someone who has experience on these two platforms.

And while an engine can abstract you a lot away from the underlying hardware, you still have to keep certain things in mind (okay, Mac OS X nowadays is x86-64, just like Windows, but on consoles you are suddenly faced with PPC CPUs and if you want mobile you suddenly have to deal with weakly ordered ARM CPUs. Changing platforms is no kindergarten, it's the exact opposite).

@preacherX
...and if you target to release a game which should be available for Mac, ask yourself the question if it's really worth all the work.