I have been working with the GG source code since 01/22 and have come
to the conclusion that it may not be feasible to fix the major problems
with GG. The reason I have come to this conclusion is based on what I have
accomplished so far. I have discovered that there was a slowdown with the physics
implementation. To correct this I had to completely remove the Bullet physics
interface with GG. I had to also remove the version of the Bullet physic engine
GG was using as it had been heavily modified which was causing many problems
in GG. I then added a newer version of the Bullet physics engine and started building
the base of a new interface with GG. This is where the recoding runs into a snag.
The main GG engine is a spaghetti mess of global variables and very little object
orienting. The other problem is that parts of the core have been dumped into LUA
scripts which makes it more obfuscated to debug or add too. In conclusion anything
can be fixed and I know how to fix it but, the question remains is it feasible( time & money)
to recode most of GG to fix these issues. At this time I am running out of both and
I am contemplating giving up. As I am one of the few( or maybe the only) one who understands the
physics code and the DBPro source code I believe this would not bode well for GG.
I am looking for some feedback from the community before I make a final decision.

To be honest what do you want anyone to say ?
Are you intending to do this via GitHub or for your own use ?
Anything via GitHub has to go through Lee and if its for your own benefit then its up to you I guess
If it is personal Its not like the community will benefit after all your hard work .. Its not like we can all have a copy .

It’s your choice realy ,do you want to spend the time it will take to complete it . All improvements are great but only if you can spare the free time to work on it .

Dave

Just fix the stuff that's broken, I'm actually finding that there is far less wrong with the physics than I first thought. Sure the legacy from DBPro is obvious (I am assuming that's where the spaghetti comes from) and doesn't conform to modern standards but I haven't yet found anything terminal about the patient, I've seen far worse believe me and some of that (terrifyingly) is responsible for getting you to your holiday in Tenerife safely!

(btw, wtf is 01/22? Have you come back in time of something. {i.e. Jan 2022})

I am glad it is working for you but I still insist that the core of the Bullet Engine was butchered by Lee.
You will run into problems eventually.



Did you look how elapsed time is handled?



Not sure what this means exactly.Are you talking about GPS systems?
I do not vacation there.



Easy now, here in the United States this means Jan 22 of this year.

I was going to point out that we don't say, "the 22nd of January," we say, " January 22nd," so why shouldn't the numerical notation follow suit(?), but I think he already knew that. I think he was just giving you a hard time, lol.

I am not sure how I should answer this, because of course I want a proper and working implemented physic engine, but if you dont have the resources to do it, what can we do?

I hope you do it! Do it one time right and it never bother you again

This is not something for the forum, it really is something which you have to discuss with Lee.
If a newer version of the physics engine mean more flexibility and better performance, i'm sure everyone is in favor of implementing it, i certainly am.
But you should contact Lee with the question if it is desired and of course the required compensation for your work.

I think this is the main point ,if you want paying just say it’s a lot quicker up front Ime sure some people would pay for it.

Dave

I feel that this is a matter between you and TGC's.
Like Earthling45 says, we all want everything to work, but this sounds like a major undertaking.

Your time is worth something and I would be sure to tell them what your offer is to do the changes you describe.

Let them ponder over it or even set a date down the line when the funds would be available to get this done.

Since GG is not open source I could not ask the users for compensation.
TGCs would need to fund it like Lee said with contracts that's why he exposed the source code.
At this point I can only assume that there will never be any contracts for coding.
I am not sure why some think this is not the place to discuss it as all other discussions about
the source code are being posted in the forums.

Yes but not funding to do it . Like Myke says its really between you and TGC

Why not Lee posted about funding contributors in the forums and the mods have been so quick to post
statements that Lee is handing out contracts to fix GG as if they had inside knowledge.
Actually it is between all of you and TGC if you hope to keep using GG.

Ok how much do you want and we will discuss it .. Sounds pretty daft doesnt it ..


That's right ... go and discuss it with Lee then ?

Maybe it is just me but I am sensing a lot of animosity about this thread.
I am just trying to have a discussion about the future of community contributions
to GG because with out them GG may not last much longer. Since the source was posted
others have expressed there doubts about the future of GG depending on community
contributions. I have read all over the forums how users want this feature and that feature
and how they have no idea how to code it but hope someone will step forward to do it right away.
As far as compensation from TGCs I have had this discussion in the past with Lee and the answer as
always been not enough funding due to poor sales which is understandable and I doubt there
has been any change seeing as how he released the code. I guess I am answering my original question unfortunately
it seems the answer is all too plain. I guess I was reluctant to accept it as I had really wanted to make
GG better than it is. As posted in another thread TGC can not except people to contribute for free while they still sell the product. I apologize for any disturbance this thread has caused as I never should have posted in the first
place when I already had the answer.

That should not be.

Of course it is ok to open this thread and discuss this matter, but the question is not answerable by us, only Lee can answer it.
We can discuss, but we do not decide if a contract is handed out for this work and that is obviously what you seek and need in order to continue with it.
Therefore you really should contact Lee and see if he thinks it should be done and hand you a contract for it.
Maybe he already has someone working on the physics engine.

Judging by how well your relationship went with the agk physics I'll say good luck.

As for physics being broken... Isn't everything broken? Name one feature that works fully... The engine is not supposed to be a professional tool, I don't think it needs to work 100%, if it does the job then it's good enough for a beginner (sure Lee says he wants GG to be used by all but let's be honest, it's still very much going to be 99% hobby users or beginners).

Ah it's that culture thing, my English brain just couldn't parse any sense out of it. So what it meant was "after almost a month of working with the code ... ".

I must admit that the way GG interfaces with BE is to say the least quirky, I'm thinking that the progression was probably along the lines of:

1) Original DBP code did dynamic stuff itself with in-house coded solutions, bespoke (i.e. furlongs/fortnight) measurement units.
2) Changed to use an out of the box solution (ODE?) with cobbled together conversion of units to allow existing games to still work.
3) Conversion to using BE with yet another set of cobbled together conversion code.

I may be missing a few steps there but the point is that there is a lot of legacy code in there to make old levels and models function as intended, any major re-write would be a major pain in that regard as it will likely break a lot more (in terms of legacy stuff) than it will fix and in any case the users are not going to appreciate it whatever you do.

So really all I'm saying is do the minimum required to fix or replace the really bad bits rather than a complete clean slate solution.

Ohh no stab , i just added a lot of physics to GG Loader so i did not get way behind when you released all the new physics stuff. link. , just kidding


I surely do not hope they are doing anything behind closed doors and wasting our time ? everything should be made in GitHub so no one waste time working on the same things, I would be pretty pissed if i worked on something and found out that it all was wasted because TGC had another plan.



I kind of fell the same as you , i spent way to much time on fixing bugs in GG then i can justify, i know this is free work and i did not have to do any of this, but it was needed so i can continue to use GG as a level editor. GG in its current state (PP2) work perfectly as a level editor, so personally i will take a break from GG and see what the future is bringing.

Some peek and pokes (on the forum) from TGC is needed to keep the interest for GG up, lately they have been pretty absence with GG, not sure why ?

Anyway time for a GG break, lets wait and see what TGC plans "really" is

Just you I hope ..
When I say you should speak to Lee I mean it .. If you can do the job and it gets you a contract then that's awesome.
By all means share your plans ( we will love the updates ) but we are just mere minor pawns when it comes to your time and money.

At most the comments are trying to encourage you to contact them .. What have you got to lose

@Preben.

What do i know, i'm just like many other members on the other side of the fence and thus do not know what is the plan or if there is a roadmap at all for future development.

Your contribution has been huge so far, i really thought you were being paid for your services.

Sorry for my late reply, this whole thread sprung up in the last 14 hours I don't think there is any animosity either, just mild confusion perhaps. StabInTheDark was the very first collaborator on the GitHub system, quickly followed by Preben and AmenMoses, and as I was also still finding my feet with GitHub I just wanted to take a breath to see how it unfolded. I will certainly be hiring coders to help out but wanted to get the groundwork right first before I focus on specific features (as I will need to manage them through which takes time).

@StabInTheDark : Thanks for taking a deep look at the current Bullet Engine implementation, I agree it's definitely not a physics coders dream to wander around inside, not least because we tried to wrap Bullet around the previous ODE way of working, and then a few years to hack in features as required. It's not pretty, and I understand you not too keen on fixing the existing 'system' and I agree to rewrite a new system will definitely need funding and time (not least to ensure backward compatibility with physics functionality that will likely change under a newly improved simulation). If you want to email me directly, we can discuss the specifics of what is broken and go from there.

@Preben : Saw your AppGameKit physics work for GameGuru loader, pretty cool, thanks for sharing! We are not working on anything secret here, and no freelancing has started on GameGuru, so what you see on GitHub and the branches are the very latest of everything right now. As always your contributions have been amazing and if anyone deserves a rest it's you! Just for the benefit of everyone, Preben spent 20-hours straight searching for a memory overrun that resulted in 3 lines of code which fixed a crash that was almost impossible to trace.

I will get getting to the end of Feb for GameGuru sales and I can then figure out how much money I can spend in March on doing some freelance work. It might not be enough to replace the entire physics engine in one go, but we can certainly identify the issues and make a start in the right direction. If anyone has any physics engine related issues, please do post them to the GitHub issues tracker so we have a shortlist we can work to: https://github.com/TheGameCreators/GameGuruRepo/issues

This may rub some up the wrong way but I would be happy to click a " GitHub Paypal Donations " button now and again when I have a few spare quid if I saw it on the forum

Just for those who may want to contribute and help pay for contracted work
Call it the money I spent on fags .. ( For Americans that's a cigarette to us )

That’s actually not a bad idea for those who want to add to the pot.

Dave

I highly appreciated this! I know i was complaining for quite sometime, two months here on forums and in emails but that fix has allowed for the stability of the public previews of which ive grown attached to and dont think i could ever go back to DX9. So thanks for that Preban. It also cleared up a lot of confusing on the store from steam users who were using the PP and experiencing crashes with store items.

I'm in support for a donation system if it ever comes to be.

In terms of physics Its not really that bad as is. My only issues are getting things to move at all with collision data on in a smooth matter without a novel of LUA.

I think what would make a lot more sense here in terms of remuneration for coders doing intensive tasks would be a bounty board where people could donate $$ towards a goal that would remain in escrow until Lee approved it as a fix.

I would figure it'd be something like 'pay to prioritize development'. Think about it. We all have a pretty good sense that Game-Guru is extremely low priced. I think overall that's why we keep coming back to it. When I purchased it many years ago as FPSC:R I specifically came over from acknex because the value to feature benefit (having shaders alone was worth it) was saving me many hundreds of dollars and hundreds of hours of development time.

Of course the downside of that is a slow development cycle. Shy of people making adjustments based on their good graces I'm all in favor of having something to help stimulate work (rewards, benefits, etc for coders). I personally won't be touching the code anytime soon because of the book I'm writing but I may join the fray eventually. Admittedly I'm pretty rusty with my C/C++ code but it was my bread and butter for a long time so I guess we'll see how that pans out.

I digress though; bottom line is I can see no reason why we wouldn't have a way to directly contribute towards development, as users. I don't see it as something that should be required but rather something that would help speed along fixes of certain features.

Thoughts?

I would donate some $ thru PayPal . I guess little steps to repair the physics would be the way to go .

Lance

Yeah, Game Guru has always been a rip-off for TGC, like two or three coffees! I really feel that TGC have under-priced this tool from day one.

That of course means less revenue, less cash for hiring others.

Performance is\has been Game Guru biggest problem, and efficient code can only contribute to making that better, so If the code needs cleaning up\ fixing\rewriting then it is only in the right way to go about making this tool even better.

"To Recode or not Recode" : I don't think it is a matter of should it be done, but rather how do we fund the work.

I've said a few times before on these forums that I would back GG again if it was an option. It came out of crowd-funding and onto Steam way too early imo.

I really feel that at a minimum Stab, Preben, and AmenMoses should all be on the payroll already - those guys can\have\will just add so much to this awesome tool Lee has created.

Also what about a button for those that add media to the engine for free also..... I mean this product has an awesome community behind it, and it would be nice to be able to reward them for their efforts (even thou they don't expect it). In the end it may only buy them a coffee now and then, but it is something.

So yeah where is the donate button?

Reliquia....

"Yeah, Game Guru has always been a rip-off for TGC, like two or three coffees! I really feel that TGC have under-priced this tool from day one."

I agree , it should be going for at least double the current price and still be in the price range that most users can afford .

Lance

For the DX11 release, TGC should think about a DX11 high quality model DLC pack.
With new great models. I would pay 50++ € up for a good pack.
TGC need to release more DLC packs on Steam. More quality packs, more money.... yes they also need more 3D artists then...

Actually it resulted in more moaning than anything with comments like ..

" Stop spending so much time on DLC's instead of fixing the problems "
" Money grabbing with DLC's "
" Why don't you spend time fixing GG "

In fact we had a moan just last week ..
https://forum.game-guru.com/thread/219238?page=4#msg2595809

I know its not Lee that creates the DLC, Models etc but you just cant tell that do some people ..
It just doesn't sink in ..

What also is needed are some show cases of what can be done currently with this game engine.
So much was added last year but the focus has been and is on dx11 and problems.
While there certainly are problems, there is also the positive side like EBE and the new shader possibilities for terrain and water-entities.

There have been several quality updates for the existing DLC packs.

The real issue is, i think, the negative sphere that surrounds GG.
Messages about AI scripts which you have to buy from a thirdth party in order to have a better working AI does not help.
Also a faulty script in a DLC pack (mission to mars) does not do much good for the general view of GG.

I’m talking about really good, outstanding DLC packs...
I guess nobody would complain, if they are really awesome, high quality.

I vote for recode but not only the phyiscs engine but the whole thing.

How about start from scratch, implement Lua the proper way so we have full control over everything and anything by using Lua in a way that makes sense and not only tweaking unrelated features to make magic happen. Add a plugin system, make GG modular meaning everything is a plugin:

-physics
-terrain
-AI
-weapon
-UI
-FP camera
-TP camera..etc

And each plugin must be hooked up to Lua so we have some level of control in-game and also plugins could also make calls to other plugins by using Lua.
It would allow people like Stub in the Dark develop and sell their own plugins. Not mods like with FPSC, no source contribution like what is planned with GG but plugins. How about that?

This way, TGC don't need to fund contribution, don't need to maintain the code of others. If anyone interested to add a new feature to GG the community can also fund the development by purchasing the plugin and it would be maintained by the creator of the plugin. In case TGC want, can fund the plugin, make it official being maintained by TGC but it is going to be up to TGC and the creator of the plugin. The point is, if TGC can not fund something really important and useful, the community could still fund the development by purchasing the plugin when it done and functioning as expected and it would also motivate talented coders to contribute.

In my opinion this is what would boost the development of GG significantly. Yes it means we won't see shiny new features for a year or two. Definitely not this year. But then, we would see rapid development in all areas.

I have already mentioned this, but in case TGC is interested, personally I'm willing to consider to "pre-order" GG:Reloaded to fund it development.
My suggestion to TGC is do a survey and see how many people willing to fund the development of GG: Reloaded. No new features would be promised this time but only to improve the core and add plugin support to allow developers to add improvements and new features of their own... Not even need to guarantee all the existing features, only a solid core that we can build up on....

Unfortunately, yes people would complain, and it has happened recently, Lee added some new good quality PBR assets to a couple of the existing DLC's, and the response was literally "stop wasting time on assets and fix the engine". Over the last year or so the community (especially the Steam community) have made it very clear they don't want more assets, they want things fixed, and core features added.

It's a mixed bag really, we do get a lot of requests from Steam users wanting more assets, and the other half saying all DLCs are evil, so I don't think there is a way to win this one. We will be doing some DLC's this year, and of course our regular DLC updates to keep them fresh, but mainly we will be working on the core engine functionality and outstanding issues. Now that DX11 is mostly out the way, we can get back to core work.

Speaking of core work, thanks for your ideas so far on the PayPal Donate button, which has the potential to create a flame war the "likes of which even Bob has never seen!". I like the theory that community donations go directly to the freelancers who work on the features the community want to prioritize, and as far as I can see it's a win-win if the user who donated feels the money went directly towards a feature they want to see done quickly. I will start a new thread on this (after donning my asbestos armor) and get the reaction from the whole GameGuru forum community

Finally, for Zigi, the idea of writing GameGuru from scratch gives me the shivers. There is cumulatively about 10 years of coding inside GameGuru right now, amounting to probably over a million quid in time and costs to create. I dare say we could shave this figure down with the experience we have gathered, but it would still take many years and by the time we finished it, there would be no-one left here to use it. The only course we have is to fix what is broken, add what is missing, and slowly reform the code base so it does not resemble the work of a complete madman. Over the years I have considered both Unity and Unreal as the technology base for a rewrite (i.e. the power of Unreal, the ease of GameGuru Classic) but when you got closer you realized it was messy when it comes to licensing, lack of freedom with the product, creating executables, depending on a black-box system that constantly shifts the sand and the fact we would be throwing away plenty of technology I.P and hitching up to someone else's train, where-ever that leads. If someone wants to rewrite GameGuru from scratch and work with us to publish it, I am all ears

i love the idea of donation funding to specific goals, like better physics, skydome etc etc.
like kickstarter for game guru extras.

when reach the money goal for that feature it can either get a contract out or fund lee to do it next. the money only comes out when it reaches its goal so not to have floating money where people complain...

if that came in i would be a bugger! expect high funds on night day and stuff like that.

bang goes my managers monthly wage lol

im happy to chip in. but i want to chip in for what i want not pay for multi players etc. i waited long enough and time for people to put their money where their mouth is.

Well I'm less in favor of a donate button than I am of a bounty-board style bounty list. The way it would work would be:

Board has specific objective
(Example: Physics engine rework)
People can donate to the objective to fill a sort of pot (escrow)
(Example: over 6 month pot exceeds 500 USD)
Someone finishes the work, submits it to github, it gets approved and included as release
Funds are released to person who finished the work via paypal.
Bounty is removed and unlisted.



I don't know how that would ever be implemented aside from some sort of escrow and PHP system.

But that's what *I*'d like to see. I wanted to make a site like that a while ago but couldn't spare the time for it. In the meantime, intellectual capital is free today so have a bite

There is an old saying , "You can put lipstick on a pig but it's still a pig " . GG is not a pig but , not fixing the core of the engine first 'in my opinion' would be heading in the wrong direction .

Lance

I can't imagine that you seriously suggest rewriting all of Lee's work done over half a decade. I'm familiar with impressive physics related stuff you have shown us in the past but nothing much past that and I just know that this would take ages and probably dissolve in sand. This leads me to believe that you wish to recode all physics related elements of this engine which seems like a more reasonable undertaking.
Okay.
I'm fine with that. Personally I am just okay with the way physics are handled in GG. Enemies ragdoll, objects react to impact and can be thrown around. That's all I need. Everything else is entirely trivial to me. I know of people like T-Bone who are working on advanced things related to vehicles so I can see an improved system benefit him an users like him. However, with the memory related issues we are currently facing I think improved physics are something not worth poking at. At least not now.
Personally I don't like where this whole discussion is headed. GG is small and a niche interest. I don't think we attract a crack team of mercenary coders with this whole bulletin system that is suggested. I would much prefer the talented people who are familiar with TGC's stuff to take a swing at it than hiring people looking for a quick buck. I just doubt that this whole pay pal thing will go anywhere.
I also have the feeling that minor flaws are being blown out of proportion here.



-Wolf

Sounds awesome! Popcorn at the ready.

This would look great on the Product page

But it is the work of a complete madman .. isn't it?

Edited to add, I think Lee's "Million quid" is a low estimate, 10 years ago I quoted my main client 2 million to replace one of their systems and got turned down, they went with another system entirely and so far the cost of that system has risen to 400 Million quid with no completion date in sight. The system in question is way less complex that GG (although the implications of failure are a lot higher ).

I had resolved myself to not post anymore but I now feel I must clarify somethings
for historical sake.


First let me say I have not really done any thing that impressive with physics.
All I have done with physics was learn as much as I could about the Bullet Physics engine
and learn thru trial and error how to properly hook it up to a rendering engine(DBPro/AGK).
On top of that I took some other peoples (Open Source)work for example the V-HACD library
and used it to automatically create compound collision shapes. The only thing I have actually impressed
myself with is finding and fixing some bugs in the core of the Bullet physics engine but this was
done out of necessity while trying to use Bullet. Bullet is open source and I did contribute these fixes
back to the Bullet code.

Allow me to clarify, I am not suggesting rewriting all the work done on GG that would be insane and
probably not possible. For those that have not looked at the GG source let me explain GG is made
up of Libs mostly the original DBPro libraries and some extras for things like terrain and physics.
Code from these libs are called in the main engine. When the main engine was moved form DBPro
basic syntax to C++ it was not done correctly by C++ standards in my opinion. This makes it very difficult
and very time consuming to hook up to and debug. Some of that code would need to be put into proper
C++ classes and not left as a global soup of variables. The real slow down was with the way the Bullet Engine
was modified forcing it to fight itself for positioning of objects which would slow down any game engine.
The code that wraps the Bullet engine was also contradicting itself by trying to move kinematic objects with forces.
Enough about the errors, Over the past month I have actually solved this by removing the damaged Bullet
engine and replaced it with a newer version I have created the wrapper needed to call the basic physics
and I have it running in GG with a better terrain collision and correct dynamic and static objects.
Also I have kinematic objects which GG did not have before and a character controller which is not
slowing down the FPS when bumping into objects. Unlike most other users of GG I have been able to tell
that the physics was not right and was slowing down the whole engine. Until seeing the source code I could
not determine the exact problem. Over the past several years I have given Lee advice on how to use the
Bullet physics engine but without actually working on the code it was the best I could do.


For those who may not be aware I was working on and using the Bullet physics engine in DBPro before
Lee was even using physics in Reloaded. I had already developed the ragdoll interface before ragdoll was
talked about in Reloaded. The current ragdoll in GG is the code I gave Lee for Reloaded and he was able to drop it in
and get it working.



GG currently does not reuse collision shapes and you might say it is not a big deal but they use memory.
Every bit counts.



Maybe they are but from my knowledge and point of view GG is limping along.
I do not like being the bearer of bad news but Lee was well aware when he exposed the source code
that someone like me would look at the code and easily see the problems. Forget any physics you think
are fancy features a FPS game engine needs a properly working character controller for the player
and enemies and a properly working static terrain and static objects that do not slow down physics.
Good memory conservation and this intern makes it easy to have more enemies navigate the scene.
I am not asking any members for money, what I have done so far I have done without any promise
of compensation( more than 160 hours). I was just wondering like all of the community what is the future
of GG development and thought the community should be aware of my research of the source code.
I could have just looked at the code for 15 minutes and posted the problem but instead I put in the
time to prove what the problem was and I have a faster working GG.

I understand what you are saying but at the end of the day ,do you want to help with GG .there is never going to be a lot of money available.not trying to start a argument here .

Dave

That is the problem Stab, you know what you talk about like maybe a few more members of this board, but to the rest of us, it is completely unknown what it all means.

There are some easily repeatable slowdowns accompanied with a huge drop in fps when a player is close to a building which contain multitude collision boxes?
I've seen it happening when close to buildings from the megapack, but remove collision from the FPE and the problem is solved.

I'm quite curious how it runs performancewise, do you want to share a video of the new physics engine in GG?

Lee said:


If I read correctly - Stab knows how the bullet system works, and had a hand in getting it into GG.

Game Gurus biggest problem is memory in stand-alone games, and surely if the current Bullet implementation is costing GG memory, and will never work properly, then why would we not want it sorted....

I know Lee wants to maintain backward compatibility, and I agree with that to a point. But if something is broken then it should be fixed regardless of whether it breaks old projects or not imo.

So anyhow, it seams to me it needs to be addressed and Stab is able to do that. I can't think of any reason why I would not want that personally.

All I can say is I hope TGC and Stab can come to a workable solution for getting this sorted.... let the negotiations begin ....


Reliquia....

An interesting read Certainly it would explain GG's odd speed quirks. I for one am all for GG's physics to be improved, It could be that there are several areas it could improve on beyond the basic player control.

If Stab in the Dark manages to improve on this he deserves a beer or two.

@Stab, is there any chance you could provide us with a built game or show us a demonstration which compares your GG build vs the current GG build? Would love to see something like that.

from my side i hope everything works out so that Stab in the Dark can fix everything that he thinks is necessary .
It is now the time to make sure we do it wright , i support Stab in the Dark , lets make sure we try and get to a point were we can support the people who can fix GG in any way that we can .

Thanks to everyone who as made any changes up to now to get GG to be the best Game Engine