If you are going to skim, better listen to it instead.
A few days ago I had told you that I'm stopping with making the police station in my game Dani's Race because it had a Spaghetti Code Problem. The main() function in the Main_Update.py file was 1683 lines of code long and contained way too many things in it. At the moment, the same function is down to 641 lines of code. This is still way too much stuff in the main() but this is a hell of a lot of reduction.
After the article I worked on cleaning the main() function by myself and documented some of it on Dani's Race Matrix Chatroom. This is when I moved pretty much the majority of the racing code out of main() and into Racing.py.
I did not simply dumped the code from main() into an equally stupid spaghetti code thing in Racing.py. Instead I've broken it up into 13 individual functions, which are now useful to do various interesting things that weren't possible before.
There is a point in the first story mission where Moria sits into a car you bring her and goes on practicing the racetrack. To do this in the Script.py I had to manually fool the system to think that Moria is now a part of the race.
if Story["Moria"].get("driving"): racepos =[-61.7856, 1028, 0]# Moria drive the car to the raceif Story["Moria"]["driving"].getDistanceTo(racepos) > 2 and "Moria-Practicing" not in Story:
Story["Moria"]["driving"]["npc"] = "story"
Story["Moria"]["driving"]["target"] = racepos
Vehicle.TargetedNPC(Story["Moria"]["driving"])# Moria racing the raceelif"Moria-Practicing" not in Story:
Story["Moria"]["driving"]["race"] = "RacetrackRace"
Story["Moria"]["driving"]["npc"] = "racer"
Story["Moria"]["driving"]["racing"] = True
Story["Moria-Practicing"] = True
But now I have a specific function StartRace() which can start any race even without Dani being a part of it.
def StartRace(racename, dani=None): race = Races()[racename]
race["started"] = True
race["start-time"] = bge.logic.getRealTime()# Deleting startersfor starter in race["starters"]:if starter["cylinder"]:
DeleteCylinder(starter, ignoreTimer=True)# Making the racers racefor racer in race["racers"]:
racer["racing"] = True
racer["launchtime"] = 100
racer["race"] = racename
# Testing
racer["npc"] = "racer"
racer["active"] = False# Beam of energy
racer["beam"] = Reuse.Create("Racer.Indicator")# If the race includes dani.if dani:
dani["race"] = racename
dani["checkpoint"] = 0
dani["lap"] = 0
# Make the racing UI visible scene = bge.logic.getCurrentScene() pointer = scene.objects["PointingArrow"] posindicator = scene.objects["Dani_Pos_Indicator"] timeindicator = scene.objects["Dani_Time_Indicator"] lapindicator = scene.objects["Dani_Lap_Indicator"] posindicator.visible =True pointer.visible =True timeindicator.visible =True lapindicator.visible =True# Play the start the race sound.
bge.logic.globalDict["SoundDevice"].play(bge.logic.globalDict["sounds"]["active"]["sound"])# Play a Ding
This theoretically gives me an ability ( with a minor modification ) to make racing in the city happen in the background. Like once every so often you would be able to stumble upon street racers doing their racing by themselves. Which sounds like it could be a lot of fun and could give the game a bit more life.
In the same time making the functions smaller gave me an ability to look at them more carefully and it made me fix two long lasting bugs: Bug #26 and Bug #29.
The thing causing both of those bugs was a rather sloppy implementation of the car's spawning. For example, you may drive pass a racing event, for which the race driver cars would spawn. And then you may continue a little bit further, making one of the cars de-spawn and re-spawn again as a car on the road, which is no longer a race car, but which still was recorded in the metadata of the race as a race car in the race. Which you can probably guess did some problems.
Also if a car spawned for the race it would not spawn ever again, for some reason, if you go further from the race. Forcing you to restart the game if you want to play the same race twice.
One friend of mine was playing the story mode and one mission required going to a race. But he chose to pass through the race starting zone when going through the previous part in the story. The cars spawned and then when the story needed it to be a race, refused to spawn again, making everything break. So cleaning the main() actually helped the game be a bit less buggy, since I saw all those mistakes and fixed them along the way.
Here is a recording of my PeerTube livestream where I finished moving the code of the racing. And where I discovered and fixed some of the bugs.
By the way, subscribing to the RSS of this website will not only give you new articles, but also notify you when I am streaming on PeerTube.
Further I had to move all kinds of other code, such as optimization code and code for doors and elevators. Some of which I did during the livestream that I linked above.
Later I tried doing another livestream where I tried cleaning the code, but I was tired as hell and so it became a stupid ADHD-fest where I ended up working on the Police Station instead, in the end, because I couldn't concentrate on the code.
But, then even later I actually streamed again, this time with proper energy and concentration and moved elevators, doors and a lot of optimization code out of main(). Here is the recording of that stream:
In the same time I broke main() itself into a few functions, separating it into main(), Init() which runs in the first frame of the game and OnExit() which runs when you close the game. The Init() itself is also not just a copy-paste from what there was in the main() but rather I broken off some of the initialization into Racing.py, Doors.py, Elevators.py and Opt.py. Leaving only just a few if statements in the Init():
for object in scene.objects:if"spawn" in object:
Vehicle.RegisterSpawnPoint(object)if str(object.name).startswith("Crate"): crates = bge.logic.globalDict["garage-crates"]
crates.append(object)if"LightSpawn" in object:
Opt.RegisterObjects(object, spawnAtDistance, "LightSpawn")elif type(object.blenderObject.data)== bpy.types.AreaLight:
Opt.RegisterObjects(object, spawnAtDistance, "AreaLamp")elif"Tree" in object.name:
Opt.RegisterObjects(object, spawnAtDistance, "Tree")elif"Palm" in object.name:
Opt.RegisterObjects(object, spawnAtDistance, "Palm")elif"Elevator" in object:
Elevators.RegisterElevator(object)elif"Door" in object:
Doors.RegisterDoor(object)elif"ambiance-sound" in object:
bge.logic.globalDict["sound-ambiances"].append(object)
I could have made functions to precalculate all those things in some sort of way that would look like this in Init():
But I decided against that, because it would have required looping over all those objects multiple times and the game has a lot of them. Even if it only happens on the first frame. The first frame is already a bit too slow, so to optimize it a bit I made one loop in the Init() and call functions based of the object type from there.
I just want to remind you that there is a petition going on for the game and it is taking a bit long. Please, if you have a fediverse account, make sure to sign that petition so that this "already old" version could be released and a new petition with all those improvements could be started.
The Spark is great in all the things that graphical artists value. It looks jaw-dropping. But it fails for me in the story-telling aspect of it. What is the story of The Spark? A character falls down a hole and finds there a clue. And because of how utterly bored he is, he decides to follow that clue until reaching a place where he finds grass. A thing that he was programmed to find. Yes, it is not as straight forward as him just walking there by himself and finding the grass. There are obstacles along the way, but non of them make any impact.
I want to document the way you might have a possibility to use the same software to make a similar website. @Madiator2011 already done that with blog.madiator.com. Lets go over: where you get the code, how do you set it up, how do you publish, how do you manage accounts, and most importantly, how do you modify everything, so it will look like your own thing.
Something along the lines of 75% of all humans are tasked with mindless work, which is often pretty much decorative and is there only to facilitate a certain governmental goal of employment. Basically non of us are doing anything actually useful and we are all struggling because of it. And I often feel depressed doing such jobs, yet there is seemingly no obvious road to escape them.