Sorry your browser is not supported!

Hello there community!

You might wonder why I didn't post this in the official thread on the topic. To be honest, this thread is not just about what I think should be implemented in FPSC² / Reloaded, it is also about my personal 7 years of experience with FPSC and several Mods and it is also to educate some of the less experienced users that don't know that much about the subject but still shout out suggestions to TGC.
I'll also point some fingers in this very long thread so...buckle up!
Note that I try to write this in the best english I can this early in the morning

Direct X9 Limitations

A reoccuring argument that already manifested itself in the FPSC X9 / X10 Discussion is/are the limitations of DirectX9. Many substandart engines I've seen and tried blame a variety of just badly written code and clunky designchoices to DirectX9 limitations.(thats just lazy)

DirectX11 isn't that important.

Before you start to furiously punch an angry response to me in your keyboard, please look up what DirectX actually is... check out some of its keyfeatures and you will see that a lot of DirectX11 / 10 effects are way out of a reasonable goal for indie developers. If you are still in the state of mind that you try to achieve graphicsquality as you see in recent games post 2005 by drag and dropping content somebody made for you in FPSC this thread isn't for you. Everyone past that state knows the difference between graphics and aesthetics and can keep on reading

I would like to see TGC adding access to X10/11 technologies to FPSCreator but this is a secondary objective. I dont want to see the TGC team sitting there for hours to give us Direct X features most of us wont even use to a point where it would matter and therefore have less time to focus on more important aspects of the software.

If you still don't agree with me on this...thats okay. Allow me to point out a variety of things you might not yet know:

I know when people yell DirectX they mostly talk about graphics so:

Every game you will ever play on the XBox 360 will run Direct X9 only. I challenge you to create games that look better than recent XBox titles.

I even played a lot of games with recent visual quality like crysis, bioshock etc. on X9 only as I didnt switch to an X10 compatible OS back when these came out. The difference was only noticable if you looked for it in these titles.

From a realistic point of view: FPSCreator can not have the power of Unity or UDK...its better to accept this right away instead of throwing fantasysuggestions at TGC.

I am completely aware of recent DX performance enhancment but then again...if we look back at FPSC. There isnt that much performance enhancment. Those of us who have been there since V1.0 have seen it all but even though it advanced drastically I would much rather like to see a well designed DX9 engine which TGC is used to work with than another X10 that will fall apart. A second pullout as big as the X10 one would shine a bad light on TGC.

What is FPSCreator to TGC?:

(you can scroll down to the last segment of this point if you want to keep this short.)

People around here who have red my numerous remarks about what makes FPSCreator great (creative community and approachability to outline this once again) might have red my thoughts on this but I would like to elaborate on it a bit:

It might seem to be offensive to the TGC Crowd so yeah, I think you guys work wonders and are an overall nice bunch but FPSCreator was never ment to be a serious game development tool. It has been handled, marketed, developed as a game (it has also been received as such by the general public hence all the "I can't believe its FPSC" comments around the web issued to more advanced games made in FPSC.)
I always felt like FPSC hasn't been taken as serious as TGC's other projects. (Dark Basic Pro, AppGameKit etc.)
It has been marketed and developed to reach a teenie audience... (I know this for sure as they described it as "Game Making for Teens" on their new webpagedesign a while ago...the community felt this to be misplaced so they changed it to "explosive gamemaking".)
...too old for the 3D game maker.

It was the community who drove FPSC in the direction it took almost 1 year after its release to an approachable development environment. However, all the updates tagged on a rather unstable matrix didn't solve all the main-issues and led to the software reaching its zenith recently with the memory cap.

Bare with me here, I have a point, I promise! I'll get to that later:

FPSCreator is unique in its simplicity. I can say wholeheartedly, or however you spell that, that it is the tool for newbies interested in gamedevelopment.

I have seen many people picking up FPSC and growing in talented gamemakers. Most have started without any prior knowledge and have grown through FPSC and discovered their place in the medium. Be that designer, coder, model artist, musician etc.

I even know people who just picked up FPSC to try and for some fun that ended up in the games academy and later in the business...and thats awesome!

So how can this be? Simple: FPSC allows everyone, unlike any other engine, to get some results right away. This can horribly backfire and produce a lot of crap that will shine a bad light on FPSC in the general gamer/developer conciousness but then again this is less important. Maybe have an IQ Test on your first start of FPSC2 and if its below ...say 120 you can't use the software? Just kidding!

I can do games in FPSC that get noticed even though I couldn't ever make a game otherwise. I discovered my place in the world of gamecreation as artist and general designer who can't code hello world without looking up the commands. (okay, thats an exageration but you get the picture) However, using FPSC's premade core engine and incredibly ingenious scripting system even I can make complex ingame puzzles happen.

Which brings us to the next focal point: FPSC has given birth to many really cool games that may have never seen the light of day if it wasnt for this software. It is crucial that it reminds as approachable as it is but it would be wise to take a look at some of the more advanced FPSC projects out there:

A lot of them either are or have first person adventure elements (finding clues, solving puzzles.) I would take this in consideration if I would be to rewrite the software.

Essentially, TGC should design FPSC around these core values and maybe expand on a lot of elements. Browsind the community made games by the advanced users and comparing the ideas with the results. the ambitions with the possibilities.
I would much rather like to see a more efficient tool for lone developers/artists/beginners than a 3D Game Maker on steroids being marketed as a game for teenagers.

My biggest Problem:

Using the software for years now, I can handle most of the issues newcomers point out and I believe I could make some legit indiegames that might sell a bit if there wouldn't be the horrid optimisation that we all experience as the gutwrenching terror we call the memory cap.

Recently I tried to create a level which consisted of 2 rooms without any noteworsy content. (only a few barrels and pipes and 2 fuseboxes) and a larger storage area that led to a small control room with some screens. The level is kept as minimalistic as possible using almost only cubic detail props and I don't need to tell you that this is hardly a level.

Its just a kiss of death for a game that tries to tell a story or get the player hooked to gameplay and tone if it beats you over the head with a loading screen every 4 minutes.

thats right! If I test my really large levels I only get 4 minutes of maximum gameplay length. This is inexcusable. Of course, players who are new to the level might need longer as I give them some puzzles and backtracking to do but the point is still valid. 4 simple rooms is no legit level for a genre which often lives of large and complex environemnts. (I mean, just think of the classic fps games prior to the boom of setpiece shooters... Games like Doom and quake had great architecture and large levels at the time)

I can not underline enough how deeply this issue wounds fpscreator to a point where I had to discontinue a lot of projects and even contemplate leaving FPSCreator as a whole.

I will now show you a list of projects I had to cancel as they would not be realisable within the boundaries of FPSC's memory cap. Please take a look at the simple leveldesign which is constrained to mostly small indoor areas as I was always aware of this issue and tried numerous approaches to defeat it with unstatisfying results.
[href=http://forum.thegamecreators.com/?m=forum_view&t=185350&b=25]
Shavra: Dead Frequency[/href]

The levels did simply not build after I added weapons and enemies even though I put a lot of effort in keeping them under a certain amount of polygons, effects and textures.

Psishock

Psishock looked great on the drawing board and it was designed to be a first person adventure with a lot of documents to read and puzzles to solve. I wanted this to be my first legit game since the popularity of euthanasia...kind of to deliver an actually good horrorthemed game. To make it work flawlessly in FPSC I designed the setting to be very claustrophobic and dimly lit.

Now that I have designed a game that relies on small rooms and corridors I can make these shine in, for indie standarts, good visuals right? wrong!

Relict

My best received title in this community. The game had a daringly large scope for FPSC and I managed to get everything working! I had an inventory system that was none but you couldnt tell the difference, I had multiple choice dialogues with animations and voiceacting, I had magic, I had puzzles, I had the visuals.

I couldn't make levels that flew well and lasted longer than 4 minutes.

You see! I pointed out earlier that implementing new DX would not be that necessary and I stand by that point! I would not be opposed to FPSC being a 32 bit application at all... From my point of view: The primary focus of a rewrite/overhaul would be optimising the engine so we could make legit FPS levels.

I dont expect it to handle big outdoor scenes or overly detailed rooms. You can certainly see that I am not asking for much...just please find a solution to this problem.

If I only could pick Psishock and Relikt back up in the near future I would be more than pleased

Cross Plattform:

Technologie is exploding around us and we get major engines like Unity and UDK to support awesome 3D on tablets and phones... and of course some of you want this for FPSC.

I know you do, I even saw someone who I'm not particulary fond of to ask for android support. I beg of you, before you clutter TGC with requests...do some research and once you have a certain understanding of the technologie other than none at all, bring your suggestions to Rick in the stickied thread. We are still having lagspikes and optimisation problems with FPSC on computers. I see no way to have it run on a mobile phone or tablet. I know that some of you want this to make some cash as the shooter community for tablets is not yet Jaded with AAA titles.

Xbox and Playstation ports are a timeconsuming and difficult part of gamedevelopment and can go horribly wrong. Did you know this?

Read this paragraph again.

Now again.

Okay! Did you get the part with this being a large undertaking for professional studios? Good! I see it being possible to get FPSC games to port to Xbox or whatever but I see this as a major undertaking for the TGC crowd to achieve properly.

Mac and Linux? Sure! Once you get that memorycap issue fixed I'd like to be able to tap into these operating systems as there aren't that many games for these to begin with.

I am no coder and might be wrong on this. Its just my point of view.

Bad AI?

The AI in FPSC Games is simple...you know...like in

Return to castle wolfenstein
Chaser
Doom 1 to 3
Call of Duty (yes)

And you can get enemies easily to get in cover behind a barrel and react to the player. You can even make a stealth system using dynamic lights as invisible trigger zones to interact with some of the static ligting (its almost possible to make a legit splinter cell game...if you could make a level large enough for proper alternate routes)

What? You can't do all these things? thats properly because you didn't look enough into the scripting language. A bit more than not at all would have already given you an idea of what you could do.

Okay! You can't have hyper intelligent enemies as in FEAR but you certainly can have simple responsive enemies like in Call of Duty.

EAI's task force characters have outlined some pretty advanced stuff in terms of indiegames... but all you really need for a challenging shooters are enemies that spawn once you enter a certain area and then execute a script that either makes them interact with waypoints or to freely move through your room (depending on the design). I recently discussed this with Bugsy who designed a bunch of responsive enemies that consume very little memory and performance by still being effective.

My point in this segment is that FPSC's script system is so great and simple that you can do a ton of things (if you dont load too many pictures and sounds otherwise you will encounter the memorycap monster) without being a mastercoder. Its perfect for people like me who concider large paragraphs of code disturbing.

Now a lot of people complain about the AI ...and you can call me biased... I just know that they didn't even write or modify one script beforehand.
This seems rather annoying to me.

If you really ... really... expect to just drag and drop enemies in your editor without any work on them whatsoever and you expect them to react the way enemies do in shooters from studios who spent millions on their project ... well... that is a little bit obnoxious wouldn't you say? Honestly?

EAI did a great job on making his chars execute good scripts, including a simple system for allies, for the drag and drop crowd that are not yet ready for custom script and ... ah! who am I kidding here? The lazy folks and the teenies who only click some content together in the editor without any effort from their part.

Bugsy argued that a lot of these people don't complain about the AI but the animations which are pretty awkward on stockchars.

If you, TGC, are going to work on the Scripting and AI commands... please keep them as simple and efficient as they already are...some of us really appreciate and count on it!

Physics:

It would be nice to have a working physics system. I mean, in FPSC x9 we where almost there... If the system we already had would have worked properly we could do Amnesia/Penumbra'esque games.

Problems:
*Lighting. Dynamic objects didn't blend in with the static lighting.

*Objects tend to fall out of the map which would make a game that would rely on a physics puzzle unsolvable.

*Objects tend to fall and jump around. They glitch quite often.

*Maybe a more reliable ragdoll system.

Again: I don't really expect you to get us anything ground breaking...just make what is already there work.


Save / Load feature

The Save and Load game feature needs to work reliably and flawlessly. This is another issue based on FPSC's core design as it originally didn't have a save/load system and the tagged on one never worked without flaws.

It is important. Really important.

Multiplayer

If you have to get into multiplayer...please make it a stable, yet simple addon... I really don't see a point in wasting too much time and energy on it but this is personal. I don't like multiplayer games that go beyond capture the flag and team deathmatch.

For this next segment, I will quote some things I noticed.

Code it with C. Make it modern with all the standard features of this era, while still maintaining the simplicity and innovativeness of the editor and you've got yourselves a cash cow

And a workload for another 5 or even more years
I highly doubt that TGC will use any other coding language than the lates DBPro build. It is their own licensed coding language.

It would be nice to see though but ... I don't need to have it.



Yeah. I remain skeptical about this idea... sorry guys.



True! It would be absolutely necessary to be able to easily migrate our extensive media collection/ our hard work into the new fpsc. The folderstructure can change though, its better than in UDK but could be improved as it gets rather megalithic if you just own all modelpacks.

*****************************

Well, I haven't gotten as far into details as I wanted to, but I feel that if I keep going, nobody will read this and it would be a waste of effort.

So yes! TGC, I believe in you guys, you are great! Please deliver on this. Also if you read this, could you give me a word wether or not you get onto that memory cap issue? Its important to me, if you feel like you won't be able to do significant changes give me a shout so I can prototype media for 2013 projects to be more like early PS1 games (Silent Hill, Metal Gear)

Please, If you disagree or agree with me, feel free to explain and I ... don't know if opening a new thread for this was completely justified.

Have a great monday guys!



-Wolf

100 thumbs up for this, wolf. i totally agree with every point.

thanks for writing this.


gz

agree myself. I didn't realize how long this was gonna be until I had already read it

You can have a third on that one!

Yeah Wolf I have to admit that you've got a point.

Guess I was too excited for a moment when TGC announced the Kickstarter thingy. I was expecting something truly new, a new large scale project to take FPSC to the next level. But to be honest: If we get the current FPSC that is fully working, bug free and performance enhanced, it should be more than enough.

Hello, i was going to make a thread about this but i see it was already made. So i want to give you some of my own suggestions about this matter. First of all i don't believe in Reloaded version. In my opinion code must be rewritten and i shall explain later on. Also i agree with Wolf about DirectX. I want you to look at UDK. UDK uses DirectX 9.0c render and it looks very good. Old DIrectX 9.0c will still be here for a while. And now i would like to write things i would like to see implemented in FPS Creator 2:

Curved segments - right now FPSC interface is made by squares and we can only make square roads or segments (unless we make our own models) would be nice in my opinion to have curved segments to make round corridors and roads.

Fragmentation tool - this tool will allow users to make own destructible walls and stuff.

Possibility to make group objects - What i mean by that is imagine i want to make a level with lots of campfires. For that i have to drag wood models, fire effect/particle, sound zone(in case fire has sound) and hurt zone(in case fire will hurt player). Do all this for 5 times. Wouldn't be easy to make one group object and place it 5 times?

Scene organization - Have any of you worked with maya? Maya has one tool called Outliner. It is made for scene organization purposes. Basically it tells all of the objects that are in the scene AND you can orginize them,in FPSC case it means Enemie Zombies can be place in the Zombies Tab, objects can be separated by dynamic ones and static ones and so on...

Support for animated menu - Would be nice to have support for flash animated menus and buttons.

Possibility to scale ovelays - Imagine i want to put lots of blood on the floor to show where big slaughter was. I have to place like 3 or 4 overlays which consume memory. If i could scale one overlay i saved some resources.

Ok, now moving to most important topic: Optimization.

Ok so i have Dual Core CPU, 2GB of RAM and Nvidia 9400GT 1GB. The thing is i can't make levels larger then quarter ot FPSC max map size. Unless i make just rooms without any objects in it. If i try to makemore than that and put object in scene game will lag. And not just one time in while, FPS will drop to 10-15. I wanted to make PD level, but now i see i will have to divide it in 3 parts.

And look at this image:


This level has only 1 prefab and player start point no objects what so ever. It "eats" 500mb of RAM? hmmmm


In my opinion all that optimization can be accomplished with new code so i vote for FPS Creator 2.0


------------------------------------------------------------------
Some less important stuff:

It seems that every game engine nowdays has Terrain editor so think about it. As a suggestion there is a free program called CityScape 1.8 Promo. It comes with 3ds max 2011 in Support -> partners folder.It's free. It allows you to make cities. You could intagrate that program into FPS Creator or make a connection. Company that made that promo version no longer exists and was bought so i guess there will be no problem to use it.


Anyways i hope that some FPSC developer will read this thread and think about it

here for you (just kidding)



first i agree with you, but cod are not enemies are not so stupid(sometimes i meet in multiplayer worse people). Cod must comes all year again and it must look good, play fluent, designt... and after all this there is not so much time for ai! So they create for each character class one ai script and put a lots of enemies in their game to mask it! Crysis has a long more time so they can set and tweak each character like they want!
So if you create for each character one individual script, the npc can react different on you in touch with the enviroment. But this is a lot of work and so everybody search for a cod-like script, that is great without a lot of tweaks and a lot of work. So of course you can have a script which is like FEAR, but it needs a lot of work and maybe some modding skills

to all the other things: AMEN

Corno

- nevermind

I agree with you wolf.

100% Agree with you Wolf.

The thing that cripples FPSC is the Memory Usage. Back when Half Life 1 came out I was able to build bigger levels and have more detail in levels than what I can currently do in FPSC, without having such random Frame Rate drops.

The kind of debates that I have seen in other threads are "I want it to support 64-bit systems so we can have a 4gb cap" That's the biggest load of *beep*. The problem is the core Engine of FPSC.

You know that every Call of Duty up until Black Ops is still using the Id Tech 3 Engine (Quake 3 Engine). Obviously the Engine has been really updated and chunks of source code have been overhauled. My proof, Check out Black Ops on Wikipedia

Perhaps a complete rewrite of FPSC isn't necessary but key points need to be looked into which consume the most memory:

1. Optimized Static Light Calculation System
2. Enhanced Pipeline Rendering
3. Smooth Mixing of Lighting on Dynamic and Static objects
4. Complete rewrite of the Particle System
5. Implementation of a basic LOD system.

There is a lot more that can be improved on. As it stands, Realtime shadows would be a NO GO for FPSC. Heck imagine realtime shadows with the current particle system running at the same time, Say cheers to your frame rate.

This is just off hand so I'll add more once I'm done with work.

My opinion is, it's not Direct X that is limiting FPSC but how it is written in DB.

Landman

Wolf, you could not have put this down as words better. I totaly agree with you on every word on the above post. I can not even remember how many Projects I have scratched due to either the Mem cap or some stupid problem presenting its self like a bad start of the week.

So Wolf, Thank you for taking the time to point the fingure properly at TGC team and their MAIN problems that should be solved with FPSc before moving on and adding more broken Items.

I would like to see what is broken Fixed before the venture onwards by adding the NEW X10 or X11 or fancy this or Fancy that.

Just get FPSC working properly and that will eliminate 50% of people yelling at their computer screens on the Weekends. I would say thats their best bet for a "RELOAD" project.

But Im just a guy that blabs words on forums and make everything seem painfull, heck my name is PAIN, but I think I just think of pain everytime I end up using the FPSc software with a Project that I design to work with FPSC and ends up BOMBING out or causing some type of new issue along with each update that supposedly helps fix these issues.. which it never has. Honestly.


So Bravo Wolf, have them come back and please their paying Customers who are loyal enough to be sitting her waiting to pay more money for a reloaded project that we already hacve that MAY or MAY not work better. The battle field is on their side. I would like to read a response from one of them to see what kind of Bright Idea they have to come back with.

Will it please us, Will it Make us happy, Will it be Stable enough to make a game with levels longer then (like wolf said) 4 Mins.

Those are the Questions we need awnsered by TGC team. also please make it honest, not a line of crap to make use read it Five times so we arnt running around like a chicken without a head.

Wolf, you are my Supper Man.

PAIN!

Someone give this man a medal. I agree 100% and could not have said it any better. I feel accomplished myself as I think you have vented all my anger and frustration with FPSC for me. I hope TGC sees this because the memory issue is and always will be my number 1 issue.

Read the whole thing. I do agree optimization is the most important thing we need, especially since, as you said, you can't really make a serious level that lasts more than a few minutes.
I more or less agree on the DX argument. I don't think requiring higher than x9 is really necessary, but I think it should be decently high on TGC's priority list to include DX11 as an optional mode, even post-launch.

I only disagree with this otherwise I got a lot of ideas from the whole thing...

Thanks!

Actually, the scripting system is horribly inefficient. For each object each line needs to be looped through and then each command within each line is looped through ( though once a condition fails the line is skipped )

A script like this ( realistically you would never have a script this spread out, but you get the idea)




Is 12 loops per entity that is using it not to mention MOST scripting commands do loops of their own, lets say that each of those variable commands does 3 loops to search through the variable list so now that's 36 loops. So if 10 entities were using that script, it would be 360 loops which isn't anything hard for the engine to handle, but this example is on a very small scale, realistically you'd have scripts more complex than the one in this example and most of the time any commands that have to perform loops of their own would have more than 3 loops to perform.

I ported the FPI scripting system into a clean project to run some tests, 18 2D squares all running a 34 line script the FPS was at 60, but 30 2D squares all running a 34 line script made the FPS drop from 60 to 35 and that's with some of my own optimization added to a few of the commands. It's safe to say a decent amount of the reason that FPS drops when large numbers of enemies are added is because of the scripting system.

Hopefully you get the idea.

Though, I may have completely misinterpreted your use of "efficient" in that sentence, did you mean efficient as in "easy to use"?

Sean

I might have more to say tomorrow but for tonight I just want to say what many others have already said in that this is THE single most coherent and sensible post on this topic that I have read.

Post-script edit: Whoops...I have a lot to say tonight.

As mentioned on the reloaded thread, I have seen my own lab go through the x9/10/11 debate within our engine of choice over the last several years which was OGRE3D. While people like me who are NOT graphic programmers made a fuss about having the latest graphics capabilities, the actual graphics programmers were much more consistent in their view that features from x10 and x11 could be implemented into a x9 based engine.

Again, because I do not know, as I am not a graphics programmer, what is it about FPSCx10 that allows us to have many, many instances of the same model in a level without loss of performance and what is it in FPSCx10 that allows us to have outdoor environments that run at reasonable framerates on modern computers? Because I sure would like to have those features in the next generation of FPSC.

I completely agree about the scripting language. Even though it looks a lot like finite state machines from the 60s...there is nothing wrong with that and the only real limitation here is the coder's creativity.

I used FPSCx10 to make a suite of educational games for middle school kids as part of a much larger National Science Foundation funded project. Obviously, I excised some creativity in using a first person shooter engine for this purpose. Over the course of about 18 months I invested every waking moment of my life into learning and creating in x10 and I woke up EXCITED to start again each day. Since FPSCx10 was cancelled, I've stopped working with the FPSC. I completed and delivered on what I needed to for my project but I miss it dearly and I would love to come back but right now it is unclear what I am coming back to.

Should I put more time into an engine that is "a dead platform" or should I learn to use a very similar engine in FPSCx9 that seems to have become filled with all kinds of features that can't easily be used or used at all without major consequence to performance?

Honestly, I assumed when I saw FPSCx10 was pulled that the whole FPSC series was on its way out with AppGameKit being the new focus. It is very exciting to me that TGC will invest in a new iteration. At first, it seemed like a fresh restart would be the best approach but Wolf is right - categorically - that FPSC needs to retain its character as an engine that anyone can use...yet at the same time it would really be fantastic if it could be something that a developer could continue with, beyond their early learning years, and actually build stable, large-scale games with. Maybe that is not reasonable; I don't know. But I do know that I had more fun with FPSCx10 than I ever had in 15 years of simulation and gaming experience. I still want to play....

I cant help but wonder why you said to code it with C? Surely you would want to code it with Java, its more popular and superior in many ways is it not?

You mean this?

Even Wikipedia has little on it...

I believe that performance wise, C is better than java. And a game written in C works standalone.

Back to the topic.
I completley agree with wolf, the memorycap is FPSCs biggest issue and has to be solved before adding any new things

This is not a problem because the XBox expects/requires DirectX9. Windows however, which is the platform for which FPSC supposedly develops, does not. Windows (out-of-the-box) does not include or understand DX9. What you end up with is an error. Unless and until FPSC is fixed to run natively on computers that are newer than 8 years old, it won't go far.

To me, it's not that DX9 is not capable, it's that it's not supported/present. If you insist on using it, you must include it and install it seamlessly. Making the player guess the problem and find the fix in unacceptable.

Brian.

I don't recall any game that didnt come with an installer of direct X. Steam does it automatically... I suppose its up to the developers to include it then.



-Wolf

I don't recall any FPSC game/demo that did, but I understand what you're saying. I can write my own installer, but it'll require more expertise of our younger members (to whom this suite is actually aimed at) to actually make a game which can be distributed outside the forum.

Brian.

True! I didn't bother with it as I thought that everyone already had DX9 installed as a standart. We all learn from past mistakes

@all

I really didn't expect such a positive response to this thread! Lets just hope that TGC can find a way to deliver!

I agree with you on everything I can understand The stuff I can't, well that's my problem.

And the AI thing is, that not many games have actually that good AI. They have good animations and lots of them, which trick the player into thinking that they're really smart (this is very much the case on CoD or even to an extent, F.E.A.R).

But it's a good post, I'm glad I read through it!

Hi Wolf,

Thanks for your thorough analysis of the situation, and you've pretty much pinned down my ambitions for Reloaded. We'll be listing out a lot of new features requested by the community in the Kickstarter which will help sell the idea and get funding. This funding will allow me to switch to full-time FPSC development for at least half a year. My first order of business is to break down the memory and performance bottlenecks of the current engine, and determine precisely where we are spending the memory and frame rate.

For example, the current occlusion system that we call STATIC OBJECTS in DBP, is used to store all static geometry, calculate which parts of the scene should be rendered, handle sliding collision and so forth. If this single system is responsible for slow build times, high memory usage and frame rate drops, my instinct is to replace it with a simpler system which loads in islands of geometry, using a modern occlusion system such as DarkOcclusion (http://www.thegamecreators.com/?m=view_product&id=2311) and relies on a new PhysX physics engine to handle all collisions. In theory this will decrease build time, improve the speed of rendering and thanks to a multi-threaded physics engine, create a quicker and smoother collision system. The upshot of new occlusion and physics means frame rate drops would only happen if you turned a corner and rendered a one million polygon monster.

Hockeykid's analysis of the script system is accurate. It was never designed for the aggressive usage scenarios we now use them for. I created it to provide basic logic for opening doors and operating switches. The same system is now being used to script complex non-linear enemy attack formations, and it's command set has tripled in size! It has never been optimised. There will be some easy ways to improve execution speed of this system without losing any commands, and retaining identical functionality (so important) and will mean very large levels can be populated with very clever entities without worrying about frame drops.

Finally, I have been wrestling with the whole concept of increasing the standard 40x40x20 grid level. My main objection was the potential for users to add MORE content to a single level, even though we are already hitting the memory cap. I was also concerned that the engine, the editor and the whole 'universe system' is largely build on the fact that this size does not change and is fixed. Changing it is the mount Everest of headaches. I can suggest though that if we decide to scrap the old occlusion system, it becomes at least possible, even if not feasible. For kicks, would everyone be happy with an increase to 200x200x100? Does anyone have ambitions beyond this for a single level? I would also add that if we get one of the Stretch Goals which adds real vehicles and roads into the engine, then I would drop the limit altogether so you can create near-infinite landscapes (to about 1 million grid tiles in every direction).

The natural reaction when you see the HUGE list of features for Reloaded is that we'll just keep piling more stuff on top of what we have and hope it does not topple. It would. The opportunity with the Reloaded project is to give me the time to replace the parts that don't work, improve the parts that do and produce a product that delivers on the expectations of the FPSC community. Over the years the product has found it's way into homes, offices, schools and summer camps which means we got something right and made a good game making tool. I think Reloaded will make it a 'great' game making tool!

P.S. I hope I replied with honesty and not a line of crap. I had to giggle when I read that post

This is an excellent size so long as the engine can support the media we will throw at it. As you already stated the Static universe system is an outdated model in terms of technology today and I am sure it was not your intention to purposefully do it. The universe system was the way to go 12 years ago.

Like you mentioned you would re-do the system. The system has to be constantly streaming so we do not hit the memory cap, I don't know how Occlusion works, but if you are confident it will fix this memory issue then I am all for it.

With that being said I am very happy with this, the major issues have been addressed and I am confident now that they will be fixed. You have my vow for some funding of this project.

As I said before 200x200x100 is a huge increase in what we have and if its possible I would do it so long as the memory isn't going to explode.

There you have it from the man who is hogging the awesome.

I also believe, this is the first time I've seen Lee actually post in the forums to a thread that he or one of the other TGC staff didn't start. At least since I've been around.

I would most likely support this ideal Lee is talking about. this sounds reasonable.I to would like a bigger map size if this would ruin the performance in any way then I dont care for it. I am a developer that is about reasonable ideals that will work towards a much more stable and stronger fpsc. i dont even care if there is not a graphic increase in any way. if 200x200x100 is the best fpscR can do out of a 1 million grid tiles then so be it.even if fpscR can only keep 100x100. anything is better then that 40x40x20.I just dont want all this hype upgrades to down size the three most important things. performance, stability and memory.


but also about this ideal. could adding fog to this new system also help the rendering system as well. I believe will work way better. Nintendo have done this with N64 like Mario 64. which it did not render until you was close enough to see it.

Lee, it was very honest and I meant no offence when writing that. Just after reading Wolfs first post I was reminded of all the fustration I had gone thru while trying to use FPSc for some small projects. So that built up and had caused me to type immaturely. Please Note I really do appreciate the all of your hard work and I will be standing here in the end trying out the rebuild - which sounds like the best plan for FPSc.

I will be supporting your funds for the reload and I hope we can get this project started as soon as possible as I can not wait to see what you have to offer for FPSc and i can also speak for many other members here in the FPSc forums.

Dont take what I said as offence I do love you Lee, its just a love hate relationship which I think you could understand.

PAIN!

ok

if you can do a 200 x 200 x 100 hell im in straight away
the speed up and more enemies and vehicles etc then yes
with multi threading spot on.

is it possible?
can you put the in game efects slider like in x10 though, major fast and effective. cool

will wait and see what the list entails, but the area map has won me

dam it!

Would the engine even be capable to handle that? I have doubts

After what Lee has said, no I am definitely more inclined to back this, just need to figure out how to get money on my paypal unless we are able to back this by means of prepaid visa cards.

you probably could use the pre-paid VISA on paypal as if it were a regular credit card.

PAIN!

@Lee

That would certainly allow us more creative freedom concidering health systems, inventory scripts etc. that constantly run in the background. In our current FPSC its tough to equilibrate interactive environments and performance/memory consumption which leads to static design choices.

I hate to add another request to the monumental feedback mount but an ineditor scripting solution would be appreciatable. (maybe with a little list of all the commands )



I certainly don't as everything I ever designed had FPSC's memory cap in mind but this would really open up entire new worlds of possibilitys. ...but then again, only do it if the new optimisation fully supports environments this big. Having this size available but not fully usable wouldn't serve the experienced users and would only fill our boards with "lag/error/crash" threads from the newbies who then throw together zombie-city maps.



...then we would need a terrain editor and AI that supports multi-level interaction in an effective way. This is all very exciting but please don't lose yourself in Sisyphos-work



I'm almost certain that the creative crowd on these boards would have made some righteous titles with FPSC if it wheren't for the discussed issues. I mean, there where some amazing titles in production around here that looked better than the majority of HD engine indieprojects.



-Wolf

Got to say it...if Lee delivers on this you will see some nice things being done. I have always tried to get the most out of this software and hardly scratched the surface as it is.
If free from some of the restraints imposed by current version it would leave more time for some serious development.
Having cut your teeth on this one and having to optimise levels to an extent not required by some of the more 'serious' engines, is to me a benefit which I think I will be grateful for. I learned a lot using this version and expect it to be useful for reloaded when it comes

I look forward to it.

I still think it's funny how I was the one who brought up TGC advertising FPSC to teens.

On the main topic, very well stated, Wolf! I agree with a lot of the stuff you listed. My main hope is that TGC will look more into creating tools for quickly designing custom content. And like you said, I don't need fancy shaders and high poly counts. I just want to make a fun cool game! Look at Minecraft!

You missed something Wolf, dynamic shadows.

I still hope I get to see that in FPSC within my lifetime.

I think I mentioned in your thread on the subject that I don't deem this feature to be really necessary. It would be nice but not mandatory

I absolutely agree with this, thank you for raising the issue!

Here are my points on it: FPSC has failed as an engineering tool set. There are two things anyone needs to create a game, even one they'll only use themselves. 1) An efficient, enabling production tool set. 2) Skills in the craft of making your assets/using your tools. The big failure of FPSC over the years, which is why we're all here debating this now instead of toasting to the success of the world famous FPSC 12, is because it fails as a system to produce a product! Forget graphics for now, leave new weapons. Make this what any other major software product is: an enabler to users to create their product. Can you imagine if you had to hack Microsoft Word, or Facebook just to get it to work?

That is what we need, a new FPSC that makes the quality of its production pipeline simple, easy, functional and its first, best goal!

Hey Patrick.

I see that you don't use the gamedev, modder lingo...so I'm afraid we will have to put you down.

I'm kidding While I'm sure you have a point here, could you please define "engineering tool set". As in engine (Unreal Engine, Cryengine, Lithtech engine) or as in development kit (UDK, Unity).

Production can mean anything in relation to making an FPSC game. Do you mean the process of creating your game or building and distributing a final game?



I remember the days of Win95/98 so...yes I can But jokes aside: Facebook is a website and a final product for a broad user mass... word is a text editor.. there is absolutely no comparing here, sorry. You also don't have to hack FPSC to get it to work... you just have to resort to uberminimalist designchoices to evade the memorycap. If you compare other devkit (UDK for example) to finished software it will fall short too... (Imagine if you had to save your document every 3 sentences in word out of fear the software could randomly crash any minute). I mentioned above that FPSC has been marketed as a game and part of that point was that I would like to see it less developed as a standalone software that has to work and be as userfriendly as officesoftware but more as a game development kit... but a newbie friendly and approachable one I'm not sure I get my point across here... so lets use a metaphor: Comparing finished software to a tool to creat it is a bit like comparing a car to its assembly line.



then again...if you put it that way, you are correct

Hey Wolf,

Haha, I've actually been out of the FPSC-user circle for a long time; FPSC R has drawn me back in. I'll try and clear up what I was saying a bit.

My thoughts on FPSCR, in your thread and mine, are about 1) how we get R to fulfil the long awaited 2 goals of the community and TGC: to have software that effectively allows us to make games, that sells well. And 2) Why it hasn't, in my opinion, done so so far. I agree completely with you, that FPSC has been treated like a game rather than a fun, actual tool. And the areas in and reasons why I think its like a game, and not a game making tool are as follows (I sometimes explain myself better in stories, so I'll try and use one here ):

My Dad, a systems engineer, told me a story about how his Dad would look for a new family car, back in the '80s, by looking in the local paper at ads. Those that caught his attention, these being the stand-out good ones, were ones that said 'good starter, starts after only a couple of attempts' etc. To me, it seemed crazy that a selling point of a (fairly) modern car was that it started when you put the key in. Can you imagine a car advert any time in the last 10 years saying 'comes with new, GPS navigation built in! And as a bonus, it now switches on in the first attempt!"? The point being, the people who made cars, somewhere between the 80s-2000 finally had what was needed to make cars that were reliable, functional and thus worked overall as products. Now, once the basic issues of functionality were conquered, once cars actually could behave as cars 99% of the time, they became useful enough and good enough value for money for the common user to consider it a must have, not a luxury, and that actually worked for what they wanted it to do. And it then made it feasible, from a business & science point, for developers to look to the next stage - cool new features.

The point is, is that car makers fulfilled the golden rule of product making: they changed the way the engineering-system of the car worked, so that it could function as the tool it was intended for, satisfying user goals, before making it fancy(though sometimes other products' functions are to look fancy, e.g. as a fashion icons). When it works for what people want it for, people will buy it. Thus is the case with how FPSC R needs to go.

This is why I compared Facebook to MW, which are, as you rightly pointed out, very different in the type of product they offer, but are similar because they both are tools with complex, interconnected parts, that are extremely functional for the average user in delivering their certain products, be it chatting or typing.

In my opinion, an FPSC user wants it to do 2 things: allow her to make a relatively good game, but work in a way that you don't have to be a coding genius to be able to make the game, so all she has to worry about is her creativity and skill in the crafts of texturing, level design, etc. But this is where I think FPSC has failed.

As you are of course aware, like any vet user, a new guy could pull up to FPSC and think it will be easy to create his game; just switch it on, drag in all the characters I imagine, and hit play. Of course, its never that simple. Firstly, if you've got an original idea, its original, so different to what's there already, so you will need your own media to populate the game. So you get some model packs; but you would need an infinite amount of model packs so that everyone gets what they wanted. And what if you want the characters to behave in a slightly different way, you don't like the textures? Etc. The point being, that even with making compromises, of course anyone who wants to make a few games will at some point need to bring in their own media, ai, scripts, etc into FPSC, and if you want to even tweak several characters (let alone import several complete new ones, as is often the case with the upper-class of FPSC games) this becomes a hugely important, and time consuming effort.

So this feature of the game-making process, user generated content, should be recognised and made to work on a large scale in FPSC, like say importing songs into itunes en mass, or copying folders of files around a PC is. How do you address this? Well maybe something say like a wizard that allows you to import many files at once, specificy that all these are character files say, that automatically finds the animation file, textures, etc, that you've made, applies them, then makes them FPSC, and alerts you to errors that you can manually fix, etc.

My point in mentioning this area of game-making was not to say this is the most important thing FPSC needs to change, but as an example that there are big areas of indie game-making not well addressed in FPSC, and where the tools needed to fulfil them aren't provided adequately or referenced to enough (no one software can do it all of course, but FPSC at least should be compatible with the 1 or more tools that provide the features it can't).

In my opinion, game making even on a small indie level is a classic systems engineering project. FPSC is the tool set that allows us to complete the project, who's basic functionality should be to adequetly address the main phases of game-making. When I said engineering tool set before, I just meant the overall system of tools in FPSC that allow you to go from taking your media from other 3D software, to dragging it into the levels, testing and compiling into an executable, and the process in between.

My point summed up, of why FPSC doesn't work as a game-making tool, isn't because it doesn't provide enough features like physics, graphics, weapon variablity, etc (which, although could be improved and will be fun to have are not what kills it). It's because it lacks the crucial 'behind the scenes' tools and solutions involved in the design, construction, adaptation phases - the development process - that are all present in even a very average indie game's development cycle.

Linking to what I said before, TGC are the car designers, we are the car users, and FPSC is the car. FPSC is the product that has yet to develop its primal key aspect: it works as the intended by the user, being able to carry out its job. What is its intention to us? Be a real, but newbie/friendly game-making tool set, that approaches the systems-engineering area of game-making effectively, giving us the tools we need without having to rely on the mods of extremely generous vets like yourself to get it to work, after we've bought it.

Sorry about being very verbose on this, I think our main points are the same. FPSC isn't really an approachable game-making tool that cuts out the huge learning curve associated with e.g. coding, and my opinion on why this is so is because it simply doesn't work. It doesn't provide proper solutions or tools that are adapted to the way game making works. And in turn, this is why I think it hasn't reached as broad an audience as it could. If its basic functionality as a tool set was solid it could do what YouTube did for video: suddenly condense what was needed to make good entertainment into something everyone could approach, so that indie became the new standard. Then when it worked for the large market of video makers, it became very successful. If FPSC R really works in meeting the requirements of a game-dev process for indie sized games, it could be big.

Few! Now I think I've reached the end of this post! Apologies again for its length.

What are your thoughts?

I only will buy it, when there is a better Menu/multiplayersystem.
Now its epic but with a good Multiplayer with CtF and TDM it would be the best, and i would buy it.

It would be cheaper (I would think) to pledge the amount needed for a full copy (upon release) now, rather than wait for the full price later.