I know it does not use navmesh with recast Api, so how AI navigates throught levels ?
I got most of time melee characters npc stuck in objects or fences or some npc not able to reach the player stuck in empty space until the player comes near.

AI characters operate on either the free roming kind (zombies) actions happen based on the distance from the player.
Or ones that follow Waypoints (soldier) follow waypoints till they get into range of the player then they converge on you and attack.
or a combination of both. (fantasy and medieval) These are mostly custom and have limited animation sets.

If you AI is getting hung up on objects then there is not at least 1.3 meters (4ft) of opening or not enough height for them to get under it.

Or the object FPE file is missing the forcesimpleobstacle = 3 command in them.

I understand, AI i tested is from GG demo games, i had npc not able to walk into a wood bridge for example, or not able to find another path ane beeing stuck in some corner. I'll take a look at forcesimpleobstacle = 3.

AI must have terrain almost directly underfoot or it will not respond so bridges are out at the moment sorry ...
However this should be rectified with the AI overhaul after the EBE is finished ..

It doesn't. Not really.

Its surprisingly smart in built games. Both Bugsy's and my latest releases have plenty of people even noting it to be too difficult. Now...if it could only find its way up a staircase

I didn't think it was to shabby in my Die Glocke demo either ... Planning your spawning points and enemy locations is very important .. ..I think of them like attack waves any old shoot em up scrolling game ... Triggered when you get to that particular section of the game

"Its surprisingly smart in built games. Both Bugsy's and my latest releases have plenty of people even noting it to be too difficult."
You mean difficult because AI fire decreases quickly your energy ?
Every GG demo game i tried, the AI is very basic, the npc comes in front of you and fires, it is not able to move around , find a cover or strafe.
If you play a 20 years old game like Star Wars Jedi Knight 2, you'll find challenging AI able to run away, cover or strafe to avoid your fire.

I mean difficult as in semi competent. They can navigate the map and follow you / hunt you down without getting stuck for example.

I'm talking from a perspective of homemade games here, not studio games. The AI is rubbish compared to the AI in (most) commercial releases. Of course it is.
For AI that is coming boxed with the game maker and can be used right away its pretty decent.
In FPSCreator, the enemy AI was even more basic, and we had to rely a lot on crafting an illusion of intelligent enemies rather than having them.
The engine would also lag or crash if scripts had too many if statements or where otherwise too complex. Take a look at what I did with it years ago.

Other engines have AI sets you can purchase from their asset stores by now but you used to have to write them yourself, so basic navigation, coherent animations and so forth was already a big deal.

I'm confident that we'll have open AI and store pack AI in the future but until now, what we got is at least serviceable.
If you want to have enemy AI that is at least somewhat comparable to commercial releases, you will likely have to invest time and money in another engine.

I'm not saying that we don't need a big AI update, we certainly need it when it comes to navigation and customizability (there is a type in there somewhere) all I'm saying is to adjust the expectations to the niche we are in and not to the latest (or old) studio releases.

This set nearly costs 100 bucks and it ain't drag and drop either.

You don't get the point.
I can code AI behaviour that will strafe, find a cover or hide before coming on your back, if there is navmesh navigation, i already done it in Unity, it's easy when you know code.In GG npc they can't walk on a bridge or navigate and use stairs inside buildings, while i can do that in Unity with two lines of code to get npc follow you everywhere, walk stairs or walk across a bridge.

I played your game, and the npc AI was stuck inside walls, unable to navigate to the open side if i was walking back and hidding behind the wall. Or your npc just run in front of the player firing without trying to dodge, strafe or run elsewher, what is very easy to code. GG AI system is disappointing.
Navmesh navigation is used everywhere, it should be another high priority for GG to get at least a perfect navigation throught the level, this will make possible a whole new level of gameplay.

As I said above .... This should be rectified with the AI overhaul after the EBE is finished ..
Lee will introduce some kind of Navmesh System

How so? You asked what I meant by "surprisingly difficult" and I explained where I was coming from.

I later also stated: "I'm not saying that we don't need a big AI update, we certainly need it when it comes to navigation and customizability (there is a type in there somewhere)" Which is roughly the same thing you just said.

There is little you can do until the update is out as I believe a lot about the AI is currently hardcoded and can not be modified. What can be modified is limited and mostly revolves around the distance of the player to the AI.

"I later also stated: "I'm not saying that we don't need a big AI update, we certainly need it when it comes to navigation and customizability (there is a type in there somewhere)" Which is roughly the same thing you just said."
I agree with you on that.

There is little you can do until the update is out as I believe a lot about the AI is currently hardcoded and can not be modified
I discovered GG, and never thaught it would include so much hardcoded animations, AI and other sutff (this is a beginner software conception mistake).
About AI, Lee could expose AI and let us modify it or make our own AI from scratch (vision angle, raycast sensors), and include Recast Api.

Exactly!! I have seen pretty neat behaviour scripts for monsters and animals from smallG but nothing serviceable for "people" so far.
I really hope that opening up the system will be part of the AI update.

This all depend with how far along the EBE is and intends to be. Lee did mention he could spend more time on it to polish it, but that could add 6 months to it. So I think we will be getting a crude version of the EBE and then moving on to AI. I think its safe to say that before May we should be getting into the AI stuff?

This all depend with how far along the EBE is and intends to be. Lee did mention he could spend more time on it to polish it, but that could add 6 months to it. So I think we will be getting a crude version of the EBE and then moving on to AI. I think its safe to say that before May we should be getting into the AI stuff?

How many people will use EBE ? Is this what GG users rally want first ?

I find more important to have a good gameplay and AI before , than bring better lighting and shaders for models and terrain. Because your structures made with EBE will use outdated shaders and still stand on top of an outdated terrain shaders and lighting.
Lee could make better lighting and shaders first , i don't mean making CryEngine graphics but something looking like today graphics at least and it should do lot better promotion for GG than EBE .

Just my opinion.

Hi and welcome to a world with a lot of restrictions.
At the moment I try to optimize AI with some basics and after this I will try to give them some more behaviours. This is the plan and a lot of work needs to be done. To be clear I started this project not even a week ago.
Yes we need raycasts for the AI, this hardcoded AI crap needs to get out ( SetCharacterToRun(e) and so on) and I would like to set animations not with frames(my little wish)(Also if someone could send me a video of all animations and its name that would be great). Also we need that the AI can register his weapon and change it(second little wish)

If you want to create an AI, send me a PM and we can talk about what AI needs, I like to have some input. If you have some experience I would like to talk to you

For all other developers, what are your most annoying bugs(not movement in objects(Physic is not my topic)) you have with AI? What features you want?

My experiences with AI:
https://forum.game-guru.com/thread/213283
http://forumfiles.thegamecreators.com/thread/208051

Also we need that the AI can register his weapon and change it(second little wish)
I am not sure, in AAA games AI will stick with it's default weapon, or able to fire and do melee attack when the player is close, but changing firing weapons is not common if it was what you was meaning.

I already made custom AI using navmesh , raycast and cover points or some path points for the random patrol routine.
This is not features i want , but what features all GG users should also get
I will send you email.

Also we need that the AI can register his weapon and change it(second little wish)
I am not sure, in AAA games AI will stick with it's default weapon, or able to fire and do melee attack when the player is close, but changing firing weapons is not common if it was what you was meaning.

I already made custom AI using navmesh , raycast and cover points or some path points for the random patrol routine.
This is not features i want , but what features all GG users should also get
Until GG gets Recast and Detour Apis integration, raycast functionnalities, you'll only make limited AI unfortunatelly.
I don't know if DLLs integrations and calls are possible or extending physics with Dlls ?

To be honest the EBE is actually important. Maybe not for the majority of us here in the Forums but for all the new comers and lesser experienced game developers. There needs to be some type of built-in structural construction method. For FPSC it was Segments which worked really well. Unreal also had a similar tool its earlier versions. Pretty much like Lee's original "blob tool" concept.

While nobody is forced to use the EBE, I think its an optional element that the engine does indeed need. I was in the process of bringing back Segments for Game-Guru, however I discovered that there is absolutely no way to get Game-Guru to do a punch action for walls. If there was a way to make punch holes in FPE and the grid snapping was improved... I actually wouldnt need the EBE at all. Since I have 100's of decent quality segments from FPSC and some of my own.

Very early test (before I did door frames and doors):


The EBE received a LOT of votes and was at the top of the Feature Voting Board. So there is a lot of people that want it. Though my assumption is most of them are people that do not want or know how to model room pieces themselves. Understandably so, considering "the easy game maker". The product has been taking a pretty big hit on by Steam reviewers... and I think that is because of the limitations currently more-so than the visuals. By the way I value the visuals just the same as everyone else. I'm obsessed with detail but I understand why the EBE needs to get tackled presently.

I will never do a AAA or commen game I play a lot of multiplayer and when the weapon is empty most players change to secondary. How cool that would be if AI could do it too. But yes it is not a must feature.#
No GG is not extendable with dlls. And nearly the complete AI is in DarkAI.dll(A modified version of the standart darkbasic pro lib) But FPSC has a raycast command so it should not be impossible to do. PLEASE LEE, do it and we do the rest

PS: I didnt get a mail sry

It might surprise you but there are hundreds of thousands people with far less experience than you and other artists from this community.
EBE will enable them to create.
Hence, it is certainly wanted and a huge plus for GameGure because it allows people the freedom to create.

It could very well mean a growing user base which is always good.

To be honest the EBE is actually important. Maybe not for the majority of us here in the Forums but for all the new comers and lesser experienced game developers.
I understand your point of view, if you take a look at Unity you will find such plugins to make worlds using Tile maps like Rotorz plugin.
I like FPS Creator segments and rooms quick painting, if FPS Creator had navmesh you could do amazing indoor levels and AI.
Why does it takes so much time to release ? Why not just taking segments painting code made in FPS Creator and bring it to GG ?
After all it's only a 3D tiles placing program.


It might surprise you but there are hundreds of thousands people with far less experience than you and other artists from this community.
I also want a fast painting rooms or building tool to quickly build levels, but without proper navigation like Recast you will still be stuck with dumb AI not able to follow the player everywhere once it detects the player.
Well, you are perhaps right, a painting rooms is what made FPSCreator so enjoyable , and what should make GG more enjoyable before AI re work.

The segment tool was the biggest restriction to people who like to create. After a while in Classic the only segment I used were the floors and stairs so AI could get around. Everything else was costume entities for buildings and contents.

400 x 400 is not enough to create very much without having 100 maps to compile afterwards.

The voting board determined where we are today. (strongly disagreed with it from the start), as menu editing seemed to be a popular choice over working AI and shaders. If you look at the results today, some of the most ridiculous things are over what should be in the core product.

I feel that this is the results of being able to slap things down on a map with little to no thought and then you want to present this garbage as a finished game and make you friends suffer.

MY opinion, not TGC's.

I don't know.
Looking at today games like Overwatch, BattleField, the popular maps are outdoor maps with only few indoor buildings. Why users really need EBE ?
Skyrim or follout 4 are outdoor levels (with interiors maps that could have been done with EBE or modular 3D assets and snapping), but the main part is outdoor. I don't understand the choice to promote indoor game making instead of gorgeous terrain and outdoor shaders and lighting ?
I don't know about EBE, but 3D tiles would make a more optimized level insteda of too much EBE objects cubes.
It's my opinion on EBE only.

The way Lee is designing the EBE to work is, when the construction is complete and you save it, it will convert to a single mesh model. So no, 3D tiles would not be more optimized.

You should actually be able to create "HUGE" indoor levels with the EBE
Its just a case of creating one section ... saving as Dungeon 1 .... edit it again ... Save as Dungeon 2 ..in essence creating your own Module kit that snaps together ...

Looks great indeed, but some times vision is too big, and it should not take the same path as Everquest Next lol

I agree with Myke. The voting board is also a bad idea in my opinion and I have spoken out against it whenever I could but Lee seems to be set on it.
Next thing will be the EBE.

I get where he is coming from, GG is the successor of FPSC which had a very robust segment system. It seems logical that the EBE needs that kind of tool as well. You see, if we don't get that, we'll have to model our interiors ourselves. And that is not an easy game maker, now is it?

You named a few recent games that are mostly outdoor, I could name you a whole lot of recent games that are mostly indoor.

Skyrim had awesome exteriors and giant dungeons. It was also a million-dollar AAA title.

If you keep such a narrow focus on whatever AAA game is currently hot and how smooth you perceive and engine like Unity, you'll just be disappointed.
GG will not be any of that. It will improve, we will be able to make pretty games but it won't ever rival the big engines. To be honest, it doesn't try to either. In this thread you'll find a bunch of slightly older outdoor levels I've made. If we had slightly updated and more convenient shaders/materials...we could already do a lot! Never anything akin to cry engine...but something strikingly pretty. Until then, my games will have this slightly dated look, which is okay by me.

I agree with you on one thing, we really need that AI update and the additional features. That is an absolute necessity.
You've started a lot of these discussions lately, Mouaaa, just know that we all have talked about this endlessly on here. In really long winded never ending threads. For years now. Lee won't abandon the voting board, it seems.



-Wolf

Not 100% agree,
Voting system gives more time left to TGC, And passes the responsibility in the decision making to the users.
This reminds me of a reality show television , which uses a similar system to expel a contestant.
Thus, theoretically the responsibility of a contestant to leave the contest is not of the organizing company, but of the voters; You take a great weight off.

The problem is that time is up.

As for EBE, I'm not a fan, but I understand that it can serve many users with no experience in modeling, and agrees with easy game maker.
EBE is almost the same as a feature demanded by many here, conkit, do you remember?

3com

I have to ask .....

Ok lets assume we have no voting board ..... Whats next ?
Let Lee continue adding features at his discretion ( I'm ok with that if everyone else is )
But the truth is they wont be ... Or maybe a panel of Veterans to guide Lee on his next feature when one is finished perhaps...
Watch that kick off to those who don't agree with the next feature ...

I see many saying we need to lose the voting board but at the same time suggest the features they want or should be next but that's exactly how it was before .. I personally would like the AI and all the third person stuff added but many others don't ... Unfortunately we cannot lose it without an alternative method or we are back to square one ...

One way may be to wipe the voting board after each feature .... Then a week to add your votes so its all fresh and new without any old votes from users that have left or are no longer interested ...

Or we can add a roadmap ..... who picks that ... Do we vote on it ..... LOL lots of talk but no solutions and that's what we need ..

The EBE is almost finished as a first draught anyway so that would be time wasted if we scrap that at this stage ..

I feel like i am watching a dog chase its own tail at the park here

Dave

voting board is fine but it should definitely be reset every so often, maybe every time Lee starts work on his next task it should be put back to 0 votes and announced so that we can all go vote again... that way it will make sure we're still getting what we want

I agree with you on one thing, we really need that AI update and the additional features. That is an absolute necessity.
You've started a lot of these discussions lately, Mouaaa, just know that we all have talked about this endlessly on here. In really long winded never ending threads. For years now. Lee won't abandon the voting board, it seems.

I already has this discussion with someone else.
I am new to GG and i didn't find what is commeon to any 3D game engine ( particle system and editor , physics Api , navmesh navigation).
GG proposes for experienced people a fun modding way of making a game, but i found i waste too much time trying to understand all hard coded stuff to use custom code and models, instead i would have already made an FPS with Unity for example.

I don't want or ask GG to become Unity or Unreal, it must stay the "easy game maker" , it is for beginners or kids that want to make a game without bothering about code or modeling , DLC characters are ready to use with scripts, this is great, like the huge library of DLC models, this way users can get started very quickly and have something playable in no time.
Because it must stay "the easy game maker" my wishes are not appropriate, so i won't vote and i won't ask features anymore.
Lee is perhaps the one to know what is best and what he should focus on.

You're very fast with unity then! Nice!! Not my favorite engine, but some people can do magical things with it!

Do you have any projects you can share??



-Wolf

You're very fast with unity then! Nice!! Not my favorite engine, but some people can do magical things with it!
I am not very fast, there is no magic, if you got good 3D models , level concept and design experience, you can do good small games.
With Unity , If you have already code libraries or plugins like FPS template , AI libraries, you get a playable game fundation in Unity as fast as in G.
The bigger work will be level concept, making all 3D art , animations , effects, balancing gameplay, and correcting bugs .

Your GG games reflects that you got really good level composition skills and you choose carefully models and textures, this is what matters, if you used Unity with some models packs , and some game template plugins, you would do some fantastic looking games

For beginners Unity like Unreal 4 needs a learning curve, but this is an investment you make, once you got the basics , you learn very quickly new things and you are at ease.
There is game templates for beginners that you can entirely "mod" in the same way as GG
https://www.assetstore.unity3d.com/en/#!/content/62569
https://www.assetstore.unity3d.com/en/#!/content/42204
https://www.assetstore.unity3d.com/en/#!/content/31067
https://www.assetstore.unity3d.com/en/#!/content/31419
https://www.assetstore.unity3d.com/en/#!/content/18793
https://www.assetstore.unity3d.com/en/#!/content/19644
https://www.assetstore.unity3d.com/en/#!/content/27438
The advantage is that you benefit Unity graphics , animation blending, Ai and particles. And modding is drag and drop to change or make new prefabs.


Do you have any projects you can share??
It's commercial work only and private on small games, it's more about exotic graphics like Dota 2, OverWatch, Rime for example so there is nothing exceptionnal. I do realistic modeling and texturing only as practice, for fun with teammates for some games without commercial goals, or for some paid works.

That's an eye opener .. I wasn't aware of that ..

I've not seen any of the big engines that don't offer this. The examples provided are of course editable, wouldn't be much of an engine if they weren't. If you have the cash, Unity or Unreal dev is nowhere near as hard as people think. In fact if you had enough cash, you can practically buy your entire game in various add on's out there. Of course, you may have very little idea what is exactly going on and could be asking for trouble when releasing a game if you go this route, but it is entirely possible to make a game pretty quickly if you have the cash to throw at the various plugins you can get. Some people even release those add on's as full games, and those my friends, ARE asset flips.



Woof, woof! Yeah, how many threads have there been about this sort of thing? I realise to newcomers this is all new, but to us long termers (lol, sounds like prison) we've heard it a hundred times.

Still a couple of points

If the EBE takes until May to get finished in a rough state then GG is not going to get any new positive reviews on Steam for some time. I would hope it is finished this month or next at longest. There really doesn't seem much left to do with it as far as I can see to get it operational and out there. If Lee is adding a few more bells and whistles than I have seen then fine, but as it stood when it was on test can't see much to do beyond fix a few bugs.

The voting board. I hated it from the first, thought it had far too many options (designed to confuse) and was a recipe for chaos. Personally, I think Lee should listen to obvious users who have spent more time than himself actually using GG and know the real issues when making a game. Although, now the voting board is here and been here for some time it's difficult to go backwards.

Still, lets be honest, it doesn't take a rocket scientist to work out we need more speed, far better AI and FAR better lighting. Get those 3 in the right ballpark and negative comments would practically evaporate.

Still, lets be honest, it doesn't take a rocket scientist to work out we need more speed, far better AI and FAR better lighting. Get those 3 in the right ballpark and negative comments would practically evaporate.

I agree,beginners tend to choose 3D software based on graphics they've seen first. Better lighting and a particle system would help indeed.

Fear not, I plan to weave core features in with voted items, so we get a little bit of everything we want. The last update included a core tweak to improve the mouse tracking and rendering speed which was a big no-no on some levels (MMS) and would never have appeared in any voting list (amongst other essential fixes). The next one will include the EBE, which while a controversial choice, does reflect a LOT of votes and releases those votes back to the users so they can vote again. Despite the excitement surrounding the viability of the voting board, it did vote AI in as number two so it definitely speaks to core requirements from time to time, and I also agree Menu Editing will make customising of screens and HUDs much more transparent which can currently only be done by LUA scripters or those familiar with how they work

I am not averse to performing some kind of reset on the voting board (if indeed the votes are reflective of users from a year ago who no longer have an interest in GameGuru) but I have to acknowledge that at some point a user deemed the feature selected as their most important item and so it could be equally controversial to go anywhere near the reset button. My gut feel is a combination of core updates and voting board updates will produce a workload that reflects the sensible need to finish what we started, plus the need to give the community a real say in the direction GameGuru development takes in the long run.

I use the same engines Bethesda does for making extended content for their communities. The way they construct interiors is actually pretty simple. It just uses pre-made pieces you snap together. However it would not ever punch holes into anything like FPSC segment editor did and did well. You also cannot paint a texture on interiors in any of Bethesda's tools. So in a way the AAA developers are using an incredibly simplistic method for building their interiors. I could do a whole video on it if people were interested. The way they are able to build places so quickly is thanks to their very advanced and easy to use snapping tools and wide arrange of interior assets.

Bethesda does not have tools for making shapes, punching holes, painting textures for interiors. Its actually incredibly basic and relies heavily on pre-made assets. Since GG does not come with a lot of pre-made assets for walls and alike, the EBE only makes sense. To be honest... its more advanced than what Bethesda uses. As again... all they do is snap pieces together the same way you would entities with the snapping tool.

Please refer to 5:57 within this video of the development of Oblivion which runs on the same engine as Fallout 3. The process is no different int he Creation Kit which is used for Skyrim and Fallout 4. I use both engines.



Since I can make my own interior assets... this is kind of why I requested improvements to the snapping methods instead of the EBE tool kit. However this would then only benefit people like myself.

Thank you!
, however, I just don't think this is the case, simply because all the good looking unity projects I follow have a huge timespan between conception and the first demo. :S I do think that GG is faster in this aspect. I've only released a test game with it so far, so I have miles to go here.

I'd never use Unity but I've done a bit of freelance work back in UDK and taught myself how to use the UE4. Despite all the bells and whistles, I just noticed that the UE4 is ticking way too much time out of my evening and is a lot more demanding in "cerebral energy" than GG so I'll probably stick around in here.



You got some shots of that just for fun stuff?? I'm always interested.
I've done freelance before, I know how paranoid customers are about you showing around their "precious" concepts so I get that you can't show that here.

I guess the ultimate questions would be: Why use GG if you are able to use unity with relative efficiency? Why stick around? For me its this community, time efficiency and being used to "how the engine ticks". What about you?



In another way, they really are not. I know that there is a trend among developers to view their art and assets as an "investment" they'll have to make simply because there is so much out there on the market but it really is not. Its one half of your game and should have some touch from the developer himself (or his art department if given).

While the final result can be easily snapped together for the game logic to be added, these have been rather painstakingly made in not-so-easy software by the art department before.

And thats the problem with this approach as the suggestions of simply buying prefabs by artists like wizard of id. You'll end up with tons of games using the same room styles and very few people having the tools and knowledge to customize them. That was less the case with FPSC and its segment editors and we will eventually need a tool like that rather than a few stock- room sets and a dozen sets you can purchase.

Bethesdas Method works because they can pitch concepts to 3d modelers that than create their dungeons for them. Most of us here dont have that luxury.

EDIT²:

Is what I'm saying. Didn't see that line below tha vid! Sorry.



-Wolf

I guess the ultimate questions would be: Why use GG if you are able to use unity with relative efficiency? Why stick around? For me its this community, time efficiency and being used to "how the engine ticks". What about you?

Why someone uses this package or this other, this is customer business and preferences not yours, a customer buys a product, he uses it or not as he wishes, he paid for it. I am using GG as an hobby and for fun only. For a commercial game possibility, i'll wait for GG to get some features that has been already dicsussed thousand of times already.

Do you really think Unity is the best game software ? i make also 2D games and i find better ones than Unity like Game maker 2 and many others better for some tasks and faster to use. For 3D you choose Unity if you need better looking game and better features, otherwise there is lot of game libraries, and other free and commercial 3D engines that will do the job. If you work with different teams projects you'll work with more software than Unity only as a lone developper for example, Team up with people has lot of advantages compared to solo work.

I just noticed that the UE4 is ticking way too much time out of my evening and is a lot more demanding in "cerebral energy" than GG so I'll probably stick around in here.

I think this is the main selling point of GG and why many of us picked it up.
You got a running game in 10 minutes using pre defined assets and scripts, and the editor is very fun to use , this is almost like making a level in a sand box game because GG tries to keep it simple in the editor, then you can start polishing it with your assets and modifying scripts.
There is more and more people interested in software like that, easy to use like Game Maker, but able to produce as good results as software that needs more skills. This has been done for 2D, and there is more and mode coming for 3D games.

I think the ultimate GG version would be something like the new kickstarter, something with modern graphics, with all games templates and assets ready to use, you have fun making something playable in some minutes, than advanced uses can bring better models or make new scripts.

Alright buddy, I won't bother ya again then



I think I've mentioned a couple of times that I really don't like it. I feel like it falls short compared to most other current engines.

I agree with everything else you say otherwise.