Hey gang!
I am aware that these kinds of threads aren't exactly rare on here and that I will likely not say something that hasn't been said before. The reason why I am writing this is because I believe I can pin point some core issues from an experienced perspective and that I can summarize GG's core issues without bias towards what I believe the engine should be compared to what it is. I will also promise you that I won't bring up tired comparisons to unity or similar useless tangents. This is more an exploration of whats really wrong with game guru, without vitriol and without someone trying to start an argument. I've also spent a lot of time with the software so I can pin point things a bit more accurately than someone who tried it for an hour and then complains on steam about "ps2 graphics lel".
If you've been around this community for a while you might have noticed several projects of mine in the work in progress section... you might have also noticed that virtually none of them where ever released in any playable state.
The only game guru games I ever actually finished are
test games. And those are quite old by now. Even though both of these have been released in 2016 and 2017 respectively, they are still on page 1 of the showcase board. Why is that? Well, there aren't that many finished GG games... and among those that are finished... there might be half a dozen at the most that are actually fairly decent. Thats not a good track record for an engine that came out 8 years ago.
(Test games can be seen
here and
here)
Making games is a long standing hobby of mine and I have created a fair amount of them... but why not in GG? Its the easy game maker after all... but is it? Well, I am currently stringing together a couple of scattered levels from my project
Spectre to send to friends for testing and while doing so, the issues that arose and just the amount of jank inspired me to write this down. To make a list to pin point why GG is quite difficult to work with and why so few game projects worth a lick are ever released.
1. Stability, Memory, Loading and Asset management
This is a big killer, especially for inexperienced users. Now I am aware that max will be 64 bit and a lot has been said about GG being 32 bit and therefore handicapped from the get go. The fact that this isn't exactly true given that the overwhelming majority of games until relatively recently where all running in 32 bit engines. Alan wake, crysis, skyrim, and arma 2 to name a few where all 32 bit game releases. The difference is how these engines handle content.
GG front loads all content within a map, making it not only likely to crash if your map has a certain level of complexity or (god forbid) lightmapping. This is also why loading in Game Guru takes forever.
GG, like its predecessor, isn't too decent in clearing ram, meaning that unless you use the setup command to force it to purge the memory in between level loads games usually crash 2 or 3 levels in. So now we have a game that takes ages to load (I have been reading about people where games took more than 3 minutes to load... this makes everyone who isn't used to GG's excentric quirks assume the game has crashed and tap out)
When creating a standalone game, you are more less required to build all your levels one by one and later stitch them together to avoid crashes, masses of missing media and other stability issues. This isn't obvious or even easy to do for someone who is new to the engine.
You will also find that some media from your map will be missing in the standalone version. You can copy these files in manually, however, now they will not be encrytped and again, a new user will likely have no idea that he needs to do it and release a game with missing props.
Much like fpsc, there is a lot of jank in creating and running a finished game guru game and I dearly hope that this part of the experience will be significantly improved in max.
People's gameplay ideas are also somewhat hindered as its relatively difficult to have values carry over from level to level.
Also, last but not least... Performance. Considering what is on screen in my games, mainly very basic shaders, lowpoly models and bare bones post processing effects its just not acceptable that it sometimes dips below 30 fps despite optimization efforts. I really hope that this will be an absolute PRIORITY during maxs development...but I am not yet convinced that it is.
Rendering (and yes I will go on about lighting again.)
Have you ever noticed that the people who talk about state of the art graphical features the most are usually people with zero art direction skills? Well, its just an aside. I've never been one to insist on having graphical features that I am not skilled enough to use properly nor do I have trouble creating a coherent visual experience using older tech. However, even though I think I can make a game look acceptable in game guru and the quality of the art assets in my games is good enough for me there is always this muddy, blurry ugliness to it in game guru. Given that most of my art collection is adapted to GG and FPSC (darker textures, color corrected materials, compressed .dds files, lowpoly models and so on) there is still just a certain "cheapness" to how everything looks.
This is due to the ugliest lighting engine on the market right now. The lightmapping itself is so unwieldly that I know only 3 GG users that ever gotten the hang of it and the dynamic lighting basically just vomits colors on the models. The normalmapping is riddled with unless you make them by hand the dynamic shadows are much darker on some props than on others... sometimes it won't show on terrain but will then be very pronounced and pitch black on a wall in the same area. Interiors just drown in an onslaught of bad shader rendering, lack of shadows and odd lights. It can really break the players immersion quite easily and you need to do a lot of faking and messing around to hide it well.
I understand that graphics are a complex topic with many different tastes, likes, dislikes and priorities to consider and I don't want to spark another pointless "but its about the gameplay, you philestine!" discussion so let me be clear: A 3d engine needs to have some basic features and lighting is, as I have so often said, a core component. As always, here is my little picture showcasing 2 3d scenes...with and without light:
LINK
I understand that this has been improved in max and I hope to see it iterated upon in the future
Also have you ever noticed how GG was out for so many years and we never got a saturation slider? We got contrast and brightness...no saturation.
AI (and why so few of us ever tried to improve it)
I am aware that this has also been discussed a lot and that the glitchy GG AI that will either snap towards the player and empty a magazine into him, before he had a chance to react or just stand there looking confused is somewhat notorious by now.
However, given the limited amount of commands at hand to have the enemies react to themselves and the environment also seriously limited our ability to improve it all that much. I will make a suggestion here in regards to game guru max.
The GG AI is divided into these rather obtuse packages (modules) and not very approachable. In GG's predecessor, FPSCreator... AI was much simpler. Sure, they didn't have nav mesh either but seemed a lot more responsive than the GG versions. In fact, custom AI by errant AI and butterfinges was quite decent.
In FPSC, there where several simple AI scripts the player could choose from. An enemy could just stand, rotate to player and shoot or an enemy could run towards the player once he was spotted
but nothing else.
I understand that there will likely be a "one fits all" package for the enemies in game guru max but I think it should be possible, and relatively easy to do to also let us have the simple AI scripts back.
Sometimes an enemy should just be a sniper and stay stiff at his spot. Maybe duck when he reloads but thats it.
Sometimes an enemy could just duck down and fire in predetermined bursts.
Sometimes an enemy should just run towards a waypoint, rotate to player and shoot.
If TGC gives us those scripts back, we could have far more interesting firefights in GG.
This video shows FPSC AI, that is usually only a few lines of code that does the bare minimum...however notice how its a lot more reactive than the default GG AI.
Accessibility
I have written a fair amount of tutorials for GG through the years and most of them I've written because I ended up explaining these things over and over again to people in discord. And while explaining how to rig custom weapons, how to optimize props, how to convert props properly for best use in GG and so on I noticed how none of this is easy or particularly obvious.
Underneath the easy game maker facade is a very complex engine where everything is backwards. By that I mean, the workflow to customize stuff or make media from scratch for GG is not only very complicated, requires third party tools that are rather dated its also often very different from what it is in any other engine. Personally I can find some appeal in this but if I where a betting man, I'd put down money that for most people its a deal breaker.
I don't know how this can be achieved but I think a lot more transparency and guidance for new users is needed.
I give you an example: GG's character creator has been around for most of GG's life time. There was zero documentation for users to make their own heads or hair styles until I've written one
this year. Which might have been futile, given that max is about to come out and has a different system. Either TGC or we as a community should put a couple hours aside and jot down how things are done for new members to read up on once Max is out. Something needs to be improved here.
Alright, that was my rant ... and in part also an exploration as to why so few of us ever released a finished game compared to other software I have used and maybe it will also serve as a list of core issues that TGC/Lee can check out.
-Wolf
