Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

Scene Editor rewrite allowing many new features like shadow maps, theora previews etc.

39 (63.9%)

HLSL (pixel shader) support

20 (32.8%)

odnorf font (TM) - (it's here just to check sanity of forum members anyway)

2 (3.3%)

Total Members Voted: 58

[metamorphium]

So here's the deal. What would you prefer?

[Jyujinkai]

I am going to have to go HLSL sharder support. The fact is that while scene edit could have a host of improvments... it dose work. HLSL dosn't. Things like theora privies in the editor would rule but ... well peoepl can already use thorea in the games as it is.

Unless there is a host of other features you are talking about that are not there at all.

[Derrek3D]

Could you (or someone else) please write in a bit more detail how could each of the options help developers and their games? For example, what are shadow maps and how do they look in the final game? Or what can be achieved using HLSL?

Thanks.

[FogGobbler]

I agree with Jyujinkai and also vote for HLSL things, because the scene editor works fine for me as it is. Just had a look at the new Perry Rhodan adventure and in my opinion everything can be done with WME (even the scrolling for big scenes).

But things like reflecting materials, and (sorry for repeating myself again) perpixel lighting with real shadows are vital for a modern 2.5 engine. Normal mapping for characters would also be great.

Please don´t get me wrong, Mnemonic, I think your engine is awesome, and I certainly don´t want WME to become a 3D engine, but with these features everything would be just a tick better.   

Happy Easter,
Oli

[Mnemonic]

Ok, let me elaborate a little. Both the new SceneEdit and support for DirectX effect files (aka shaders support) are features that need to be done, and I'd like to see both in WME 1.9 (perhaps ending the whole WME 1.x series). But both of these are potentially quite time consuming and may take a long time to finish.

Shaders support would be nice for some additional eye candy (and perhaps some more stuff, depending on how deep would the implementation go). SceneEdit rewrite would bring better user-friendliness for scene design (i.e.  the area where developers probably spend most of their time), but mainly it would open door for new features, such as the mentioned direct Theora integration, direct placement of external entities/actors/3D actors, more comprehensive camera tweaking, the "shadow maps" (simply an image which tells the color of an actor in a given point in scene, think more advanced decoration regions), predefined scene states, and more.

So to recap, I want to do both of these, but some has to take priority for now, because my time&resources are not limitless.


While I'm at it, and as I mentioned the end of 1.x series above, some people were asking for a WME development roadmap. Well, I don't have a roadmap, and I tend to change my mind frequently (I know it's unprofessional, but please allow me this luxury at least for my pet hobby project, I have enough roadmaps and deadlines in my daily job). I can give you a rough outline of things to come:

WME 2.0 - completely replaced renderer (based on some 3rd party middleware); completely replaced script compiler/interpreter (I already have a lot of this done, but I can't merge it with WME 1.x because it will not be 100% compatible)

WME 3.0 - complet redesign of the engine/tools; I have many cool ideas, but I'm not going to talk about them just yet

We were even discussing the possibility of skipping WME 2.0 completely and going directly for "the new thing"(tm). Nothing is set in stone yet.


And contrary to what certain AGS users think, I AM listening to the userbase, and I AM interested in your opinions. So if you have something to say, say it now. Whew.

[sychron]

I voted for the Scene Editor. Reason: HLSL support is a cool addon for 2.5D-Games, while the new Scene Editor would allow better development for ALL games made with WME. Second reason: You would have to wait for an SceneEdit-Update anyways if you want some efficiency in placing and editing the new Effects ;-)

[Daniel]

Well, the way I see it, it all comes down to WME's intended target audience. I can see three major groups: a) Developers creating free games just for fun, b) Developers creating relatively small commercial self-funded games, and c) Developers creating full-scale top-notch publisher-funded games. When a developer belongs to group c, technical features like bump and normal mapping, dynamic shadows and lighting, and other 3D effects in a 2.5D game become an issue because some of the larger publishers in the field demand those technical features even in adventure games. The reason is the competition that has sprang in the last two or three years in the newly revived adventure market. I know this first-hand from dealings with publishers and I'm already pushing the limits with WME.

Anyway, what I'm trying to say here is that if one of WME's target audiences is indeed group c, then the way to go would be to add more support for eye-candy features. I agree that better development tools like the new scene editor is indeed very important but if I have to abandon WME because it doesn't support the eye-candy effects that the publishers demand, having a new set of tools won't help me much as I won't be able to use WME at all. And believe me, I love WME and the last thing I want is switch engines.

That's the reason I'm voting for HLSL.

[metamorphium]

Still I see the chronology and this is what Mnemonic wrote here:

Ok, let me elaborate a little. Both the new SceneEdit and support for DirectX effect files (aka shaders support) are features that need to be done, and I'd like to see both in WME 1.9 (perhaps ending the whole WME 1.x series). But both of these are potentially quite time consuming and may take a long time to finish.

My reason why I've voted for Scene Editor is, that Shaders can be implemented whenever in our game development cycle whilst better Scene Editor would speed up the development process a whole lot. And adding eye candy is usually not done at the very beginning of the process. Knowing that there WILL be shaders available in WME is enough so I prefer better tools.

[FogGobbler]

I agree that better development tools like the new scene editor is indeed very important but if I have to abandon WME because it doesn't support the eye-candy effects that the publishers demand, having a new set of tools won't help me much as I won't be able to use WME at all. And believe me, I love WME and the last thing I want is switch engines.

That is exactly what I think.

I can understand that you don´t have time to bother about everything in your spare time, so perhaps the way is to separate the different target groups Daniel mentioned.

So what do you of keeping the WME as it is for hobby game designers, because it has everything you really need to get started and offer regular updates, but in bigger intervals.

On the other hand a affordable commercial "professional" version which we have to pay for (monthly subscription payment or just one amout) with these advanced lighting effects.
I´m quite sure that people would pay for it..I certainly would :-)

[odnorf]

I'm jumping in a vehicle that isn't mine here but let me clear 1-2 things based on what I know and what I have learned since first wme public version (2003). WME was once freeware for free projects and had a separate paid license for commercial games. This experience taught as one thing. It didn't work because 1) Not many were willing to pay, 2) Some people were bitching that they had to pay to create a commercial game (and even making noise in other forums with not so polite comments). From time to time there is some brainstorming going on for which road wme should take and there is some kind of draft in Mnemonics' mind but everything stops at one thing, which is the biggest problem. That is free time.

The new freeware license seems to also be problematic because almost no one donates.

EDIT: Some mistakes

[Daniel]

My reason why I've voted for Scene Editor is, that Shaders can be implemented whenever in our game development cycle whilst better Scene Editor would speed up the development process a whole lot. And adding eye candy is usually not done at the very beginning of the process. Knowing that there WILL be shaders available in WME is enough so I prefer better tools.

What you're saying is true if you belong to the above-mentioned group b. It is very different when you belong to group c, however. As a group c developer, you have to *commit* to what your game will feature *in advance*. I can't tell a potential publisher that the game will *probably* have said feature as they will never put one cent on something like that. When you belong to group b, however, you put in the game whatever is available during the development process. If a new WME version is released with new eye candy features during your development process, then you can incorporate them in your game. If, however, such features don't become available during your development process, your game will simply go out without them and it's perfectly ok because you never had to commit to such features in the first place.

WME is Mnemonic's pet project and he works on it on his spare time. In all the years that I've been using WME, I've never known him to promise any feature whatsoever (and I understand him completely) and I don't think he is going to start now. When he says he *would like* to see both the new scene editor and HLSL in WME 1.9, that's all he means. It's not a commitment on his part and I can't promise anything to any publisher based on that.

What I'm saying is that unlike in your case, I need the eye candy features as soon as possible so that I can commit having them in my game. If a new set of development tools comes out during my development process, I'll gladly take advantage of them but even if they don't, I will still be able to use WME without them.

[metamorphium]

And what I am saying is, that to have shaders in your game doesn't mean only WME support but also someone in your team who is able to program them and succesfully utilize the shaders logic. So this poll also shows how many of the potential users will be able to make use of this feature. If you need them as soon as possible and have someone skilled with shaders, he would be probably able to hijack the d3d device and implement shaders on his/her own. Having shaders (and trust me I've already worked with them in TGA) requires someone who will do the dirty shaders job.

Unlike you I sort of know that the shaders will make it inside which makes me more calm. I've worked on different projects in the past commercial or self-funded and there're not much 2 1/2D games using pixel shaders. What I am missing are mirrors and reflections (although they can be faked to some extent), shadow maps, scriptable lights, stuff like this, which is quite easy to implement and adds much more to the overall impression. And when HLSL makes it in, I know I'll have to pause the development for month or so to learn how to put those beasts in.

 

[FogGobbler]

What I'm saying is that unlike in your case, I need the eye candy features as soon as possible so that I can commit having them in my game.

I´m in the same situation as Daniel. We´ve got everything planned, graphics are good enough for a demo, but I´m a bit worried that publishers say that the feature list is not good enough.

2) Some people were bitching that they had to pay to create a commercial game (and even making noise in other forums with not so polite comments).

Can´t understand these people. If you want to make some money with a commercial game, then why not pay for the engine?

Please correct my, if I´m wrong, but as far as I know Mnemonic does everything alone. Wouldn´t it be an idea to make WME a team-based project? Everything would go along a lot faster, wouldn´t it? Sadly I´m not a very talented programmer, but I´m sure there are enough of good guys here that could help you,  Mnemonic.

[metamorphium]

Well, with dealing with publishers I safely say that there will be shaders automatically. I know that they will be there much sooner than it takes to develop the game. And if they won't be there natively, there's always a way how to dll them in. So as I'm saying it's a nice feature to have natively but nothing which helps you with making the game. It only is an extra feature which can be added via various ways.

Edit: anyway it's just my opinion. I'm not stating any objective truths (just to be clear about that).

[FogGobbler]

...there're not much 2 1/2D games using pixel shaders

What about the new Perry Rhodan adventure? Looks like they use pixel shader here, or am I wrong?

I also think that real shadows and a better lighting system are the must important (I feel like a parrot, because I´ve said this so many times  )

What exactly do you mean with "shadow maps"? In real time 3D graphics I know what they are, but how would you use them in a 2.5D engine?

[metamorphium]

Perry Rhodan is using WME, isn't it? Edit: probably not. I don't know where I've heard it.

the "shadow maps" (simply an image which tells the color of an actor in a given point in scene, think more advanced decoration regions)

(which affects 3d actors and make him much better blend into the environment)

But anyway - for you and Daniel: can you program shaders? It would be cool if we have someone on board who would be able to share some experience with tools and shader assembly language
programming! I know that I've barely scratched HLSL surface back then in torque and it was very complex matter so I understand why games have special shader teams who do only shaders or even specific type of shaders (think Bioshock and their water-artists, fire-artists etc.)

[Daniel]

WME was once freeware for free projects and had a separate paid license for commercial games. This experience taught as one thing. It didn't work because 1) Not many were willing to pay, 2) Some people were bitching that they had to pay to create a commercial game (and even making noise in other forums with not so polite comments).

The new freeware license seems to also be problematic because almost no one donates.

It's a bit out of topic but I must state here that the only reason I haven't paid or donated yet is because unfortunately I haven't reached that point in production yet. I said it before and I'll say it again that I'll be glad to pay for using WME for commercial projects and I thought so even before it switched to donationware.

People that are not willing to pay for using WME in commercial projects are IMHO people that no one should care what they think on this matter, or whether they go bitching in other forums or not. The independent adventure game scene is changing rapidly and what was true two, or even one year ago is not necessarily true now. The WME user-base and game-base are steadily growing and if WME can compete with the other adventure-specialized engines in the market feature-wise, there's no reason why people won't pay for using it.

More on topic though, I do agree that having such effects as reflections, shadow maps, and scriptable lights do cover current-day demands from publishers - at least according to my own experience. HLSL is still not a demand for now but feeling the pulse of the market, it may as well be in some 6 month or a year. If the above poll had a third option for reflections, shadow maps, scriptable lights, and other modern effects that don't require HLSL, I would have chosen it for the moment and leave HLSL for later on in the roadmap, but the poll had only two choices. Maybe a new poll is needed here then.

But anyway - for you and Daniel: can you program shaders?

Unfortunately, I can't, and I do see your point.

[Jyujinkai]

Well I think that no one is saying that we do not appreciate Mnems workload and such. As donateware WME is a very awesome package and all. Anyone that is using it for commercial use should be paying for it. If they havn't then there is something wrong with them morally imo. It is not like mnem is even asking for much.. just do not go to the pub for 2 days and bang you can afford a decent donation cost. 50 bucks or something.... more if you like.

I tink the reason people do not donate more is that there isn;'t anything to give them a guide to donate. You would prob get a lot more donations if you had something listing what donations "could" be.. Like a minum donation thing, or recommended donation.

anyway

But anyway - for you and Daniel: can you program shaders?

HLSL sharders even though they require programming can be made by artists visually in 3DsMax and other major applications and exported directly from app. You need ZERO programming experience if you have access to these visual HLSL shader apps. My experiments with the C4 engine and Irrlicht engine has shown me this.

I just think it is hard to understand the appeal of the editor myself over the HLSL shader thing.. when there is an editor already and MAJOR things like Bump mapping, Normal mapping, Reflective Surfaces, Procedural Skys, Glowing objects etc are not supported at all. I mean if when you are talking about scene edit you are speaking of only a "viewer", not new features that the user can see in game... well all i keep coming back to is... well you can do that already?

Still either way.. as a community.. we will probably be happy witch ever way you go... I think it is important to do wat you need to do and support what projects tehre is currently on that are driving the development of WME.

Mabey we can satisfy both camps? Supprt some major eye candy... then move to the editor or visa versa... rather than full HSLS shader.. get normal, bump and reflective surfaces working... that alone would make the games look 1000 times better.

[metamorphium]

Jyujinkai: hey, that's new for me. Maybe you can write some little thing on wiki how to export shaders from Max? This would be really useful. I was making my own shaders from scratch in some nvidia tool and I remember it required programming of asm language. If there's shortcut like exporting hlsl files from max, it's something completely different. Thanks for this info.

[Jyujinkai]

I guess I'll add it into the tutorials i am doing for wme...

You do need some programming skill... but nothing a gfx artist can not handel...

It all depends on how the shaders are done by mnem. Being compatably with Ogre would be cool as there is a 3ds max bridge http://www.ofusiontechnologies.com/ about... or http://www.virtools.com/ or something.....

Like i said it is up to Mnem... i think we will be happy either way....

[FogGobbler]

@meta: There is y very cool little plugin (commercial) for max called shaderFX. You can visually create your shaders in a shader-tree layout and see the result in max. Then you can export the shader to .fx if you want. You might have to change some semantics, but it works well for me.

[Jyujinkai]

yea shaderFX rules!!!

[MMR]

It would be great to allow language localization in the ProjectMan way!

[loken]

Hmm. Though HLSL is cool, it will bump the requirements up for the game engine.

Correct me if I'm wrong, but does the current WME require a card with Pixel Shaders? Granted, you have to go back before the Geforce3 days to find one, but there are some cards (or rather, integrated chipsets) out there that still do a terrible job with them, namely, Intel Integrated Graphics chipsets.

Then theres the question of which version of PS to support. If it's 3 then you have to have a Geforce7x series or ATI/AMD equivalent. The Pixel Shader version might affect the customization ability of the HLSL shader as well.

I'm not opposed to the idea of HLSL shaders, but there's some things to consider. Also keep in mind the target audience. If this were to be implemented, there might need to be an additional hardware accelerated mode for compatibility, along the lines of the now deprecated non-accelerated mode.

Just some thoughts. =)

[Jerrot]

Hmm. Though HLSL is cool, it will bump the requirements up for the game engine.

1) If you don't want your target audience to require a card with Pixel Shaders - just don't use them.
2) In general, shaders can have fallbacks, e.g. if the card doesn't support PS20, you can offer an alternative for PS11. Of course I don't know about the implementation idea of it in WME, but that's the usual way to go.

[sychron]

Remember there is some hidden geometry lying beneath the 2D scene image.

[atm_deev]

Very need shaders!

[Myxini]

Hi! I'm kinda new here but have used a variety of game engines, mostly as an artist but sometimes as a scripter.

Whilst adding shader support is all well and good, I find that hobbyist game developers can work better with what (I suppose...) could be called "Effects".

For example, the shadowmap idea, defining regions where a 3D actor is in light or shade is brilliant and anyone could use it.

Full scene tonemapping with exposure control and bloom to allow the lighting of a pre-rendered scene to have an element of nature.

Even, dare I say it, Normal and Specular mapping on 3D actors.

These both allow better integration between the 2D and 3D elements of a scene, and honestly I think that should be paramount. This is a 2.5D engine, and I reckon it should be the best one it can be.

So, In conclusion, those who desperately want shaders will probably have the wherewithal and manpower to code them in themselves, access to shaders does not mean they will be easy to use or implement (I know that from experience, thank you Quest3D!), the rest of us would probably benefit from some solid "effects".

But anyway, thats just my opinion... 

[FogGobbler]

Hi!

For example, the shadowmap idea, defining regions where a 3D actor is in light or shade...

I´ve created regions on the floor with an "OnEntry" and "OnLeave" and changed the ambient lighting (faded) to simulate dark areas, but the result isn´t that good, because the whole actor is lit up and that somehow looks unnatural. The problem with using this is also that all actors get lit up 

Would these shadow maps be more accurate?

Could anybody please post a link of an adventure game which uses this kind of technique?

Bye,
Oli

[Mnemonic]

Would these shadow maps be more accurate?

No, but you'll get smooth color transitions if your scene shadow map will contain color gradients.
This technique is pretty old actually, since the old 2D days.

[Myxini]

I'm curious, how would we use a shadowmap? Would it be aligned to the 3d walkplane or just treated a normal image? (sorry if that makes no sense...)