lonelygamedev - Tumblr blog

[lonelygamedev]

Pulverizer 14th devlog:

At the start of august I decided to start simple with a little effect for the elevator. Making the rest of the level fade in when the door opens. Afterward I decided to work on the gameplay, trying with experimenting with weapons. I decided to add a few weaknesses to the shotgun enemy to encourage other guns like the M16 and 'Gun Melee'. Allowing you shoot the enemy in the back to stun him for a bit, before moving in to melee him knocking him on the floor. Unfortunately cracks begin to show with the animation being broken, I was sorta stuck for a few days thinking the problem was bigger then it was, that it might require a rewrite of the entire health class. During theses days of thinking and gloom, I played a lot of persona 4 till I finally beat it

Side Quest: RPG In one week If you read my earlier posts, you'd know I once tried to do a challenge of making a JRPG in one week. Well persona really got the best of me and I thought it be good to do the challenge again, with better planning. Plus unlike the first attempt, I now got a cool Tim Cain video talking about how he structured RPG combat systems. A week later it is certainly cool to see a combat system that actually works if a bit lackluster, mostly due to a lack of motivation. Come back next month devlog for another week of development on this!

Side Quest Done

After the RPG I came back to Pulverizer and realized the issue that I was facing was because the enemy AI didn't stop when he was being knocked over, so he try to shoot while he was on the floor. Once I fixed that along with other thing I decided to work on the next enemy, being the rifle dude. When working on him and adding him to the game, I came across a issue where his size wasn't matching with the other characters. It made me realized how convoluted the process was of adding graphics, all sorts of little changes and odd thing's that make it difficult. So I decided to model all the enemies in one day with no planning, and add them to the game at once. Most the enemies where just variations of the shotgun enemy, with the only two unique one being a mech and suicide zombie. Hardest one was the flamethrower enemy due to the complicated rig, spend a whole day fixing his animations. Once that was done I resorted the SpriteRendering scene separating the tools and models, setting the sizes and adding a another variable to the RenderTemplate that modify the Unit per pixel ratio. And with the new tools I did all the animation rendering in one day. But the programming side of it still's need's work, for the next thing on the menu is making the game feel sharp again

[lonelygamedev]

Pulverizer 13th devlog: Off goes the head

After I finished writing part 12 this month, I decided to start work on what I worried to be a complicated mechanic, that being the Execution system. I was mostly worried with how it would look, along with the many technical questions. But first I had some things to do in preparation for that, mainly adding the code and animation’s for melee and charged attack’s, since I gutted the code and they’re required to even start executions. I added back the old function’s for hit detection and damage, cleaned it up for animation’s and added a little charging SFX. With that done I got to work on the actual executions, starting with trying to imagine what I want. I had the idea of making one animation of an entire execution, make the player teleport to the right position and play it. So I got started by creating an animation for the katana and try to add it into the game, but that’s when I came across two problems. First one being the fact that I cannot create one whole animation with two characters, blender just doesn’t allow two armatures in one animation, I had to mix two animation’s and stitch them together. And that’s when I came across my second issue of it just not mixing well with the game, everything else is multidirectional so why shouldn’t the executions be as well? After the failure of the first test I began to have a better Idea for executions The actual 3d animation will be one direction, but it will correspond with the direction both enemy and player are facing. When you get an enemy into an executable state, the enemy plays a looping animation. When the execution start’s the enemy looks at the player as it dashes to the first point, point’s are just the direction from the enemy to the player multiplied by a distance. With that I can play a sequence of animation’s and use a tool to get the position and timing right, thus the execution is done. Implementing it took a bit of time since it all required IEnumator’s, which can be tricky and hard to debug if not done correctly. Bug after bug later I got too see my animation mixed together and It looked quite nice, so I decided to implement it ingame which also became its own can of worms. One issue was trying to figure out how to determine an Execution to play since it required knowing the enemy, weapon and player character. As a temporary solution, I assigned myself to fate and that it will bite me in the ass later. I decided the best course was just add the weapon and enemy and throw out the playable characters entirely for now. After that and rewriting several lines of code I got it to work and was quite happy with the result even if it’s unfinished. I decided to also add a death animation to the player which was quite easy. By then I felt it was time to begin to push the demo out, so I started work on two things in parallel.

First one being a redesign of the one and only playable games in the game, I never liked her original Design and felt off by it. I begin the redesign with coming up with a background for the character with my original idea for her being an assassin/spy/ninja/whatever. Soon after looking at definition’s I decided hired hitman was the best profession, with the backstory that she worked for a lot of illegal companies and one of them snitch for a plea deal. I begin by modifying the colors sense the old design had an issue where the colors of the hand would blend with the body. So I made the body gray while the arms and legs are black to add contrast, and also got rid of the glasses sense they never appeared right when pixelated. Afterward I decided to try out the idea of having little armored plates mainly on the back and chest, I liked the design since it made practical sense while still looking flexible. Afterward I decided to create a backup and experiment with some accessories I came up with that had stiff consequences, mainly getting rid of the mirror modifier. When I added the armored plates, I decided to go into ms paint and add a female reference sheet I had and draw a quick design

The design I came up with added mainly an eye patch, curved the hair a little and made the right arm cybernetic. I liked the idea sense I felt like it communicated her backstory, showing she’s something of an outlaw and had failed hit’s. Slowly I begin to add the cybernetic arm and features and other than a few nicknacks, I think it’s done for the most part.

Second Thing I did while redesigning was some modifications to the animation pipeline. One of the first thing’s I did was make a preview camera, add it to a render texture and applied that texture to an image. Next up is render templates sense I had an issue of forgetting the size and camera and texture sizes. Which had occurred when doing event sizes of enemies like mechs and having to increase the texture/camera size to stop cutoff when falling down. After that to speed up the render process, I decided to make a tool where I have lists of animations with names and file locations. So I can press the run button and watch something while waiting for it to be done

[lonelygamedev]

Pulverizer 12th devlog: A catch up

It was late December 2023, I've noticed hints of trouble before with slow down and frequent 100% hard drive usage, but soon small moments became common and I couldn't work or do much of anything for days. When programming halted, I spent my days watching youtube and anime while occasionally reading. This continued for awhile till February, when I finally replaced my hard drive with a SSD. With a back up of almost everything, I can finally get back to work. I know from experience, that it's quite hard to get back into something you left for quite awhile. So I thought I could do a RPG in one week again as a warm up, but I just got annoyed and just wanted to work on the zombie project again. When I got back, one of the first thing's I did was burn all the code that used the new Input system. Because it felt unresponsive and I hated the days when it would just not work. Along the way I also begin work on the idea of items spawning in levels, power ups, guns, etc. After those were done, I realized I also hated the store system. the code was massive and the design didn't work well with co-op. So I went with a simple option and just added healing, ammo and weapon stations around the map Afterwards I decided it was finally time to add graphics, I felt like the game was fun enough and the design was proven. Plus fitting the animations and sfx's in is gonna require alot of code rework, so it was probably best to do it now then later. First thing I decided to model was the shotgun enemy, sense I already had a idea of a design for him

As I begin to animated him, thing's didn't quite work out. Sense when exporting, I notice animations kept disappearing. So I got frustrated and put him aside for a few days After that little adventure, I decided to also experiment with a editor tool. To make the level layout, sense Unity's default tools where a pain. The process of creating it became quite addicting. And I got it done in about a week

After I finished the tool, I begin start on the first tutorial area. When I finished the layout, battles and item stations, I decided to add graphics for the only playable character in the game. I imagined her with a katana and being something of a samurai/assassin, I was also playing RE4 recently so femme fatale was also a possibility. I first started to design a universal rig with IK and FK, with the purpose of reusing it for other characters. Soon I begin modeling the character. before long it soon appeared I didn't thought enough about her design as I should've, sense I wasn't happy with how she came out, especially the colors. But I pushed on and begin adding weapons. Creating a list of everything I needed to make. And start the process of modeling and animating. Along the way I also found out why I had issues earlier with stuff disappearing, basically I thought blender saving button didn't work because the deleting function is so unknown and weird I likely did it on accident. When I got everything finished, I also realized that getting them in game was a different issue. Sense the code for weapons was a mess and never was designed for graphics. So I also had to tear it out and redesign it, from scratch. Creating a function's for shooting, akimbo and melee(which still hasn't been implemented yet) Soon I decided to spend a day adding graphics to the venting machines, health and weapon store(which is fairly basic and needs to be redone later) With the new gun models, I was also able to make them Hang and spin them for drops

After I got those done, I decided to work on the first enemy you see. The zombies, or zombie sense there's only one for now

First design was fairly basic, but later on I redesigned it to fit more with the story of them being stitched together from other parts

When I got the animations in and worked on the part where bits and pieces come off. I realized I also need to work on blood splatters. My plan was to have them be part of the texture of level. so I had to do uv maps and lots of complicated shit for about a week

After I got the blood for the walls and created a fairly simple but good blood SFX and shader, I had to redesign the game health class. Sense it was long and didn't account for the type of death's enemies will have, and many...many other issues. I gutted the hole thing and begin organizing again. Separating each Enemy state into their own class, while also spacing out and categorizing the variables. I also work several new thing's into the enemy health, such as different graphic states for the zombie when it takes damage to the body. After I got the zombie working, I moved on to the other enemy which also appears in the tutorial. The shotgun enemy, so I Recreated the animations and added the Gun guy back in. And making it so you can knock him to the ground with his shield. There are still issues but for the most part it's workable

In here I use a shotgun weapon to blow a big hole in zombie stomach. I'm writing this on the 4th of july, I just finished coding this in. I can hear the fireworks and the smell of gunpowder is in the air. As I lay on the grass and watch the light's with a coke in hand, I feel a great sense of accomplishment. Have a good day people

[lonelygamedev]

Devlog 11: The road to something big #2

Around November the idea of having a demo and getting donations begin to sound really appealing to me. sense for all my time making games I haven't gotten money for it. So I begin to plan for a demo and when I felt comfortable enough that the combat could work. I begin to work a tiny bit on the graphics. First thing I did was experiment with pixel art textures on the tutorial level

While I do need to plan a bit more on the color scheming I enjoyed the look of everything so far. and followed a tutorial that angle everything the higher it goes(basically z + y+z) to fit with the sprites

one of the thing's I also experiment was normal maps. I dropped them later on sense they where quite hard to work with for smaller sprites and I'm not sure about the styling. but we'll see in the future

one I Got the basics of coloring sorted I decided to move on to some gameplay and finally add some graphics to the current list of enemies. First one being the mech which was fun to model. rig and animate. and looked quite cool in game. and It made me realized that I gotta work on the code I use to transform 3D objects to 2D sprites. which I got the basis from a tutorial by Making Stuff Look Good. Basically while it's quite good. I needed to make a few modification. one of the early one's which I don't think I ever mention was the slicing and pivot. you see a whole Animation are actually done in 1 image. Just in a sorta reel from top left to bottom right. each section being a frame of animation. I had to make code to automatically slice it by a specified cell size. for one specific reason: the pivot. basically the pivot is the center of the image. and with pivots it made layering hard to do sense you have to get the height just right to make the top part of the body fit with the bottom. and it's just much better to do this automatically But when adding the mech I was sorta annoyed I had to take several captures. and each one rotate the camera by 22.5 to create billboards and wished for it to be done automatically. it was quite simple to do Now one of the final issue was the normal maps. I wanted to add them to the character for lighting purposes. so the sprites would react nicely and with depth. but for some reason it didn't work. I found out this was because I used Unity Lightweight pipeline so whenever capturing I have to remember change setting's

[lonelygamedev]

Devlog 11: The road to something big #1

So around fall I sorta lost myself for a bit. When I went on a vacation I wrote a tiny bit. trying to come up with new idea's. Specifically enemies and thinking of a better score system. the score system I'll come back too but I came up with a few enemies. specifically a gernade launcher guy that shoots gas gernades. And the other being a guy in a mech that shoots homing missiles. also flamethrower's and kamikaze When I came back I decided to work on it a bit. The melee enemy really didn't work so I needed more then 2 enemies. Starting with the mech sense I love me some mechs. It was a bit changeling sense The Pathfinding didn't account for enemies with a size of 2. I couldn't really find a algorithm to modify the a* pathfinding to fix this. So i just Increase the radius for removing a point in the path. it occasionally get stucks but overall it's fine. The mech also as 2 arms that shoots rocket's. another thing about it is the only weakness is in the front so you need to face the mech upfront. Luckily I already had the code for the expanding gas gernades for when I first started the project and copied the splitter from L4D. and quickly when by without much issue. except the shooting. I dunno why but I always had a problem with firing physics object. they always fired where I want it too(I really should fix that) Flame thrower was fun to work on. really fun effects and overall simple to add. Through when making the enemies changing the battles was a pain sense it was incredibility unstable so I begin to rework on it The code was just pure spaghetti. And I started cutting. alot of it fixed by just taking code and putting it in a void. And oddly quickly it was finished and seemed to work fine(Sometimes it takes hours to see if it's stable)

[lonelygamedev]

Devlog 10: Iteration Part 2

So...  been quite busy. I started a trello board for projects I started but never finished. books. programming projects. 3d art projects  video games etc and each month I would try to complete a small amount of projects before I start new one’s and due to trying to finish a few projects I haven’t had enough time to work on the zombie game. after a month or 2 I finally have time and been trying to get into the swing of it. mostly small changes at first but later on some bigger changes. My goal was to add more Depth without adding complexity. Basically Complexity is like the amount of and input’s you have where as depth is the amount you can do with those input’s. which is something I felt like was lacking. The only input I added was a left shift key for healing yourself with a expansive item. Which was a really good idea. allowed for resource management, preparation and longer fight’s. But how to add more depth to the combat? well sorta just thinking I noticed that there aren’t really much choices to the weapons. they sorta just do damage. and other’s might damage enemy armor and while trying to think of a solution to this. I came across the concept of “rock. paper. scissor it“. a piece of advice given out by something called “mental checkpoint“ in a video by “butWhyLevin“. I don’t really watch most of them but it sounded like a good Solution to my game. Basically you make enemys weaker against some attacks more then other’s. Alot of combat system’s do this. from Elemental systems in JRPG’s. Doom’s guns. and DOA fighting system(which if i remember correctly from my brother is very similar). and this means I have more reason to add a variety of guns other then “it shoots really fast“. It allows for experimentation and active thinking when buying. Another thing that I also thought was a good idea was Different Enemy states. so far there is just ‘Knockback‘ and ‘Execution‘. but I wanted more. eventually i came up with 4. ‘Stun‘, stops the enemy for a bit. ‘Stagger‘ which sends the enemy and stops them for a bit. ‘KnockBack‘ which sends the enemy flying. putting all other’s that hits it into a stagger state. and when ended leads to a ‘Knockdown state‘ which is basically just the execution. now the problem comes from adding all theses ideas sense the original health system didn’t allow for attacks with names and was extremely messy. So I had to remove it and Do it all again. adding a new enemy state one at a time and organize it to allow for more variations. that way a Gun enemy responses to a shotgun blast is different then a zombie Right now I got the states finish. With that done I started slowly but surely experimenting Aside from health. Enemies been bugging me with how they moved. They where extremely indecisive. Eventually I created “weights“ which decides why a enemy would go to a point. Like if it as enemys or not. or if it’s close/behind the player. And when standing still it allowed for really cool Behavior. Such as  two gun enemies teaming up for a flank on both sides. But when moving it was just indecisive. 5 feet in a different direction would just make 2 enemies go “Oh I want to be on the other side!“ instead of committing to a plan. and I couldn’t find a way around it for awhile. Around the same time I decided to join the brackeys game jam for another failed project. But for the game I decided to try out Finite state machines for the first time. And I quite enjoy them for how simple compared to behavior trees. and where alot more easy to write. But the tutorial I used also had a method for changing a variable in another class. the person used to change to different states. After that failed I went back to the zombie project and looked at the code. I hated the fact that some behaviors have to contain a reference to a specific class so I can’t reuse the code. So with the new found knowledge of being able to find a variable and change it. I begin to add some new nodes for changing any boolean. int. float etc and see if they’re in a specific state. Another change I made was put movement into it’s own script outside of the AI. with all this I begin changing the entire A.I. getting the zombies to work was easy. The shotgun guy..well that took alot of work. But with a new start I was able to fix alot of the indecisive behavior and flaws. one of which was gun enemies mixing into the crowd. Which I was able to fix by it checking whether or not enemies was nearby and if they where. they move to a point far away. Through I cannot remember how I fix the indecisive. but point is I did it and now the shotguns are actually a danger After that it took quite some time fixing up the enemies(Or I was lazy). during the mean time I still felt like something was missing. Eventually out of random I decided to up the speed of the zombies and the game was instantly alot more fun. It was alot more intense and force me to think and act faster. but thing’s could use a little polish

[lonelygamedev]

7DFPS

*I know it's been several months sense the last devlog. Let alone one on the main project. I'll try to get back to it soon. but back to the main subject matter So...I decided to join a Game Jam, the 7 Day FPS game jam specifically. I made a small package that contained a Finite state machine, some sight's and that's about it. Other then that, i'm on my own. I knew what I wanted to make, basically awhile ago I made a prototype that had a really nice movement system. And I wanted to create a little arena/wave system out of it Halfway through I sorta knew that it won't end up like how I wanted it too. I learnt that game jams aren't really about making what you hope for and more about learning new thing's from a quick prototype. So in the spirit of that, what did I learn? 1: A fact about scoping I've alway's had a issue with scope. But this game jam taught me something. While making a "big idea" may sometimes be the case(especially depending on the jam). But it goes a bit deeper. It isn't just what sounds big or if it as alot of deep mechanics. you really have to consider alot of core elements about it. I learnt this when I realized just how time consuming 3D is. Especially for the type of game i'm making. I Spend several hours on especially breakable movement for this game. When If i was making a 2D game I would have already done it in a hour or two. a 3D game takes everything difficult and rises it to the power of 3. Movement? well you gotta account many forms of rotation. types of movement along with gravity etc. Enemies? well now you gotta deal with advanced pathfinding and annoying broken physics along other aspects. Art takes far more time now especially if your making textures. Animation is harder now. etc etc etc I now really now what I should try to look for when considering scope. I'd imagine if I designed a 2d game first then added 3D. I would likely have time to add sound and art 2: Writing on paper works wonders It's quite nice being force to come up with a list and quickly come up ideas. In 2 days I manage to get 4 enemies and 6 weapons. which is quite nice if I say so myself 3: There's a difference between Coder art and No art Almost everything in this game is cubes, including all the enemies. It made alot of thing's hard to tell. especially if I hit a enemy. and if they're attacking. Hell if It would've been alot better if I just took a bit of time. Coded a stun system along with some simple animations just to make it slightly satisfying and easier to concentrate on what's going on 4: Test it alot more and make sure to purposely try and break it. The movement felt good at first. But eventually there came a point where the more I play the more I saw more issues and it became unstable. It became really annoying when I try to play the game and test the design. When the movement goes against me when there was little time. I should've purposely tried to break it so I got the core of the game alot better And I think that's about it. Overall i'm glad I got a game that have even the coding foundation for level progression. roster of weapons and guns. But alot of the potential is wasted due to time mismanagement. quite a miracle that I even finished it You can play it here. good luck getting pass level 2 lol https://lonelygamedev.itch.io/bad-fps-game

[lonelygamedev]

Ludum Dare 53 Devlog

So I decided to do the recent Ludum dare that was coming up sense i’ve haven’t done one in awhile. I practice a bit of coding before hand and while doing that I was trying to think of a solution to a typical problem I have in game jams where I get easily distracted by youtube or something, so I thought turning the internet off would help before hand I wrote out a few style’s and do and do not’s 1: No 3D unless if it involves simple models and textures 2: Have game finished on Saturday 3: Sunday be polish And Menus + sound and some graphics 4: Almost No writing 5: Little Music 6: NO COMPLICATED MENUS! UI CODE IS THE WORK OF THE DEVIL! 7: NO RPG'S I also had a few ideas of style’s I could do or gameplay types that I might be able to fit the theme into Idea 1: Temple OS style FPS Idea 2: City Builder Idea 3: NES Zelda Idea(Warning Too Big! most likely) The Jam eventually started at friday on 7 PM with the theme being Delivery The First Idea was A NES zelda where you deliver packages and I immediately had a bad feeling about it sense it’s like many of my other game jams “Oh I have this fantastic open world game idea!“ 7 days later: “oh shit I have 3 hours left and it’s nothing that I wanted!“ But in the end I couldn’t come up with a Idea and I wanted to get started so I went with the Zelda idea where you deliver packages to people I should also mention I will occasionally be using the internet but only for a hard problem and to have a change in Music considering I don’t have much on my computer

Chapter 1: Friday So the first thing I did was made a Sprite sheet for getting some basic graphics by 9:45 PM I got the camera and Movement sorta done On 10:21 PM I created the MapManager where it manages the world and each screen you move too. I also created some Gizmos to tell me where a area is Chapter 2: Saturday 1:21 AM. well it turns out the back end of the code was actually quite complicated. I had to find a way to convert a world position. To a location of a room. then that room as a 2D array with no negative int’s that I also have to convert it too. Made it Incredibly annoying. I needed to do this for enemy AI to move tile by tile so I used a method that was likely incredibly Inefficient and horribly complicated but only did it because it was what I was similar with I also created a Test AI that just Moved Randomly just to Move it can’t go out of a room and collide with walls. Tho some enemies spawn into walls due to the fact that unity’s tile colliders made it hard to tell what is and what isn’t a collider sometimes. I then went to bed 12:49 PM After a Incredibly long sleep I made the Health system like zelda where you have half hearts. Also created the Item interface so I can have a sword with a attack now 2:55 PM The Inventory system. My fucking god what a pain, like I said UI code is the devil. And in the end it didn’t even entirely work, just hard to just convert a 1D array to a 2D grid. Should’ve write it on MS paint or paper to help Now this is where thing’s get a bit Iffy sense I forgot to record alot of stuff I might’ve spend the rest of the day on a Quest sytem + some NPC’s and the entire map Chapter 3: Sunday 12:52 AM So Here I thought up some enemies and coded them. one of them being a Bull that charge’s at you until it crashes into a wall, the other is some kind of zombie that throws at you and the other is a Bear that charges at you. stops and slashes at you and I mostly added him as a excuse to add bear traps. sense while thinking of gameplay I thought of luring and trapping enemys as a gameplay thing 1:00 AM Added money and made enemy flash when you hit them 2:19 AM got the basic UI for the store. sense alot of it was just Reuse inventory code so it’s not too hard. alot of the hard stuff will be the buying code and such 2:44 PM finally made it able to buy items and also equip them

So then I ran out of footage I know I wrote most of the quest. then created the character’s and graphics which took awhile. also the enemies. The Sound. by then it was Monday. I missed the combo deadline but I might be able to do the jam. so I just wanted to get it finished as soon as possible. I sorta felt like it was pointless considering you just deliver pointless packages to NPC’s. I wasn’t able to do the dungeons and I didn’t feel all that invested. Didn’t even did the soundtrack though luckily I didn’t forget the sound. The graphic looks alright. the Main menu was....lacking to say the least and I enter early on monday after spending all night on it. and went to sleep halfway though I came up with a Idea for a Pizza delivery Crazy taxi ripoff that I felt would’ve been better After the game jam I decided to watch a few streams of people playing games. one of them being someone called LittleDoctorGames. and while watching them I noticed that I was able to submit my game and I did just for the hell of it. Eventually I realized I had to create a twitch account and be in chat for them to play. Not really planning on using it much other then lurk and occasionally have fun in chat. I actually had a pretty good time in there with nack’s(I think that’s the person who run’s it). we enjoyed talking about metal, a few movie’s and was just a good place to have joke and chill. honestly too the point where I didn’t really mind that it took awhile for them to get to my game, next night actually. Was sorta worried sense i’m not exactly used to see people looking at my game’s and most of the time it’s a somewhat bland reaction. but it was actually really great somehow they enjoyed my 5 AM quickly slapped together dialogue where I stopped caring. and the gameplay didn’t turned out bad. It was actually fun for some people which is sorta a foreign concept to me. that people actually enjoyed it. I guess I was lucky to pick zelda sense I learnt from that game that even if the AI can be annoying with their random movement. it can be enjoyable from the choices you make and exploring. Which some people seem to enjoyed. being impressed by the size of the map for a game made in Ludum dare In all honestly the more I look back at my abandoned projects I was amazed by how well this came out. Even simple projects sometime can just be left in the dust but this one turned out fine despite it being ludum dare Which is often way too short for me. I guess the turning off the internet work amazingly and sorta made me sad I didn’t gave my all on the game by the end to resolve some of the issues making it a solid ludum dare entry. But I guess that’s more next time A few more fun fact: Most of the music are half recorded song’s for meme’s. 3 rammstein album’s. a tool album and 2 tear’s for fear’s album. and when I had the internet on I was really into Biting Elbows. I think on sunday and monday I listed to the elemental album 4 or 3 time’s alone another fun fact is the funny glitches. one of which happen after the game jam finished which was when I was showing a family member and realized bear traps where Infinite but yeah Ludum dare 53 was surprising fun. I’ll definitely be revisiting and I’ll definitely be using that no internet rule for more of my project so expect more progress soon here’s the game I made https://ldjam.com/events/ludum-dare/53/fetch-quest-simulator

[lonelygamedev]

Devlog 9: Iteration Part 1

Have you watch many videos of devlogs? if so, how many of them are actually finished or still ongoing? In my experience looking for new devlog’s on youtube that number is always high. I sorta always have a feel for what the issue might be, especially after seeing a title like “Remade my games 3 TIMES!!“. I knew it might be on what you focus on or the scope of the game but I couldn’t quite put my finger on it till now So I actually made a demo for my zombie game and released it to itch and the TigForums, to no feedback whatsoever. I knew I’m missing something at the game so I put out a demo as a sorta “please help“ to give me idea, so I was quite disappointed by the feedback. Honestly didn’t really want to work on it but I didn’t want to abandon it especially after it was nearing a 1 year anniversary like all my other projects. So I played it again and see what was up, at this point I was just throwing shit at a wall and seeing what sticks and I refused to do anything all that big and meaningless to the gameplay like graphics I thought the Melee was lacking(all you really do is just swing and enemy knock back, at best it’s a cheap way for crowd control) so I decided to work on that and add a Melee specific enemy. Eventually I begin to get annoyed by something with the melee, it was never usual for any enemy other then the normal zombies. It’s damage was so weak it didn’t make the reward worth it considering it got me a face full of shotgun. But it gave me a great Idea, have a attack right after you dash where it’s charged and do’s alot of damage but can only kill one enemy. so the reward is worth the risk. Despite it only working 50 percent of the time, it felt pretty cool killing a bunch of zombies then switching to your sword and quickly dashing for a smooth kill. After telling my brother about it he advised me to take out all the characters and just focus on the one with the dashing and melee and build the game around that and focus on balance later. So I did and once again while playing I gotten a new Idea, a score system like a platinum game where you get rewarded for how “stylish” you are. I was quite surprised that I never came up with this before considering how natural it seemed, you’re on a TV show so you’re obiviously gonna score for how well you do. Then once I got a basic score board ready for reach room that gave me a Issue, what else am I going to add to it? I just got the amount you killed with will always be the same after each room. well after thinking I’d thought of a system where some enemys have armor with some weapons doing kinematic damage. and when that armor reaches zero it will knock down the enemy so you can go up and execute them adding to their separate category called “executions”. Another one is combos which might need some work etc That one iteration opened a whole can of possibility! and made a entire game loop. And now I got that done I got alot more iteration’s to do with those systems that I can talk about when I make a part 2. But I think this is why alot of devlog’s just die out, alot of people focus too much on new big feature’s like graphics, co-op/multiplayer, a new weapon etc, and either ignore or don’t know about theses little iteration’s that could make or break everything. Honestly I was a bit heartbroken realizing that after a year of development. the co-op/controller support, graphics and other thing’s that I worked hard on adding was sorta useless and sorta blind me to the real issues that lied in my game. all those thing’s are one’s I should’ve worked on later but I learnt a valuable lesson in the end. That I need to stop bitching and just get to work

[lonelygamedev]

On space games and  the issues you might come across when developing them

So for a week I decided to try and develop a game where you travel across a universe, just to see if I could even do it in unity(Bear in mind alot of this info comes from a GDC from the kerbal space program developers) The main issue with computers is that they cannot handle big numbers so you will experience issues while really far away from the center, because of that graphics and physics will jitter alot The most well known and talked about solution is called Floating origin where if you’re player moves far enough from the center, It will move the player back. In simple terms the player is always in the center, the universe just moves But I came across a 2nd issue. The Camera(both editor and in game) have issues rendering at far distances. It was actually really hard to find a planet at sizes this big Another less known solution is called Scaled Space, something came up by the kerbal space program devs. It works like this: Say you’re on earth and you want to render the other planets in the solar system. So you created a miniature verison of the solar system and put a camera in that. that way it can see really far away. You render that first, and then render the camera on earth over that miniature scale of the universe While you can find your own scripts on floating points scaled Space was a issue sense it was only talked about for a brief bit in the GDC talk and I had to make it myself It was sorta easy at first. have 2 scenes. create a layer for the two and cull each camera to each scene. have a global variable for how scaled down the scaled space should be and copy the camera position(scaled down of course) + rotation But how do we do the rendering? eventually after experimenting I found the solution. We have the camera in the scaled see render the skybox + scaled down planets. while the camera in the original scene only render to the depth on top of the scaled planets. that way it will render objects on top of the scaled space without the skybox blocking the other planets in scaled space The biggest issue so far is combining all that(along with other features) together. here’s a video of the current state of the project

[lonelygamedev]

Devlog 8: H E L  L

So sense October i’ve been busy in life, homework and a game jam through november and half of december. Around christmas to the end of the year I decided to get back into the zombie game. So now with a second level I thought it’s time to start on a unity tool If you want to know more about what the tool Is and how it works, here’s a video I made

https://www.youtube.com/watch?v=v-QUZSGJ7Bg&ab_channel=LonelyGameDev So how was Development? hell to say the least. Serialization is a massive pain in the ass and honestly i’m not in the mood to explain all this shit. In the end While it does need a little more work(and maybe some optimization because unity tells me that SerializeReference is inefficient) it’s for the most part is done Another thing I work on a getting a build of the game to work so i can have a little demo for playtesting and I finally got that working. after another game jam maybe i’ll finish up the demo. polished it up so beginners can understand how it works easily without having the knowledge of the person behind it

[lonelygamedev]

Devlog 7: Squares be gone!

so here’s the thing. I now put in the infrastructure to have more then 1 level. which mean’s more enemy types. but having alot of types of moving objects in a fight is a bit hard when they’re all squares and there can only be so many differently colored squares, so I have to add characters

So what method will I be using? well i’m gonna try my hands at Billboarding with 3d to 2d sprite animation. First what is billboarding? well billboarding is just having the sprite point towards the camera, but that’s not it, i’m using a specific type of billboarding where I use differently angled sprites to have a illusion of a character turning(a methiod you likely seen in many old FPS’s like doom)

so now that we know what billboarding is. what exactly are we using to make the sprites? well as i mention in the first devlog i’m using a method used by deadcells where it takes 3d models and turns them into sprites. it also allows for smooth animation and normal maps for lighting. I got the base code from a github linked from this video https://www.youtube.com/watch?v=iNDRre6q98g&ab_channel=Makin%27StuffLookGood which explains the process better then I could

when I was getting the code to work and stuff I was thinking through what issues I might come up with, One was splitting the body into 2 with the legs on the bottom and the body on top. so the legs can be animated separately from the upper body. in order to do that I had to do a few thing’s to the script. for one when creating different sprites for animations it just puts it all into 1 image where I’d have to manually split it into several sprites manually. aside from the obvious problem of this being boring it also as a issue where the pivot is not exactly between the feet. this is especially a problem for splitting the body in half sense the pivots need to match perfectly in order for the top part to fit with the bottom luckily through a few unity forum posts and dealing with one very annoying error I was able to split the image into several sprites and set the pivot with the Camera.worldToViewport I was able to get it all working. Then I got sick with a stomach flu. After that i started continuing with the project again. this time i need to get my second challenge through of making a custom sprite animator. the In engine Animator competent for unity can slow the game down when there are alot of objects base on my earlier test. and it be f**king horrible to once again assign every single sprite to a sprite render. so I have to create a custom animator for sprites. it was surprising easy and I was able to get the character looking at an object in no time https://twitter.com/LonelyGameDev1/status/1579777368000204800 now a even bigger challenge is making it walk. sense you can move in 8 directions(horizontal, vertical and diagonally) that means I have to create 8 walking animations right? well now. sense the character faces towards stuff, if it turns 22.5 degrees the walk to the right now is at a angle of 112.5 instead of 90 so because of that I have to create several 3d animations of each direction that rotates by 22.5. Then after finishing all that and rendering the results. and assigning them to a 2d sprite animation scriptable object. and using some code i finally got walking working https://twitter.com/LonelyGameDev1/status/1584057486809927680 and with that in character graphics are done. i don’t have the footage but i manage to get the zombies working with it as well this time with a running animation. obiviously all this is extremely choppy when changing directions the player character looks like he’s teleporting their leg’s and there’s no option yet for blending. this along with many other graphical issues will come next time

[lonelygamedev]

Devlog 6: Saving and Co-op

So after the RPG post I’ve gotten sick with covid and could do every little during those weeks. The most I did was making the shields actually useful in defending against zombies and getting you out of a corner, The second was making projectiles a interface so I can have grenade launcher with little fuss. And with that a 3 character is created, Demoman(i’ll give him a name later. just a quick and easy name. plus you can tell where I got the idea from) After I got through covid I decided to tackle two big features, A save system and Coop. The reason why I thought it was a good time to tackle them was that I was at a point where I knew my game would be fun with a few sound effects and some polish, So I thought it be good to get in some pretty important features in so it doesn’t become too much of a nuance to code other stuff later on Saving was fairly simple after a brackey’s tutorial and one tutorial by a channel called ‘Game Dev Guide‘, that taught me how to do the slightly more complex stuff like having a List of saves for a load menu. The real issue was trying to save a Scriptable Object, you see I was using something called a Binary Formatted for saving. Which just turneds stuff like floats, bools and other thing’s into 1′s and 0′s, but it can only turn a few thing’s into 1′s and 0′s and stuff like Vector3 cannot be turned into 1′s and 0 without some manual code. And Scriptable objects are the same way, and I needed to use them sense it’s how I store which character i’m playing as. So I came up with a pretty cool trick, just store the filepath to the scriptable object and when loading the save just do Resources.load add the path in and you’ll have your current character loaded after that. And after the menus for loading a save, the fountain for saving is done I think Now here’s where it fucks everything up. for Coop or main menus to work. I had to make it so the player are spawned when a level is loaded and not apart of the level. sense alot of the code and scripts and Inputs reference player(or the player references a object in the level) it then creates a null error sense it’s trying to access a object that doesn’t exist because it didn’t exist until the level was loaded. So I turned everything from the camera, stores, weapons and shooting off. My goal now was too get the co-op menus working and get to spawn the player characters at the start of a level. Then slowly fix everything. But what do i mean by Co-op menus? well I just mean menus where two players press a button for when they’re ready, select a character and start the game. I never tried Co-op before so all of this was new to me. plus i had to create like a wheel for what character the player selects so that was somewhat complicated. After that I changed the save system a little bit to 1: have a static SavaData class that represents the current save, so I can access many players there are and other data stuff, And 2: store multiple characters and spawn them when a level is Loaded. and after that I got two players that can move with each respective controller Now all that’s left is to get everything else working. I’d decided to add a script called PlayerHandler that handles stats and UI about the player. And also set’s up it’s connections to all the other objects. But listing All the bugs i’ve obiviously fixed by getting their competent and setting the variables is boring so instead i’ll just talk about the interesting design issues I’ve come accross. First thing was the camera, for co-op should the camera be split screen or single screen? well split screen could be easier and I wouldn’t need to change my camera script. But i’d lose the feel of the original Smash TV style gameplay where it’ just one screen and two players. In the end I pick a single screen to prevent more buggery and to keep that feel. If the end result later have problems i’ll switch to split screen. The second was store’s, How do I handle having multiple menus open for a store? with split screen the problem could be fixed easily but now that I won’t be going with that I need to come up with a hack solution that(again) if it doesn’t work. I can change later. for now having each store be on different sides of the screen(left and right) will work. I learnt through looking at other coop games that I likely shouldn’t think too much about what if scenarios where a player decides to be annoying because it’s not my fault he keeps the menu one to bug the other player. Again another thing I might change later. Another design thing was status. such as the health or current amount of ammo etc. Decided to just have it be a UI element on the bottom of the screen. So that’s all the basic design choices I’d made. And while there are a fair amount of bugs with the co-op and save system. for the most part I think i’m done for now and should move on to other parts of the game like the sound and an actual graphic for the sword. Anyway that’s all for now

[lonelygamedev]

Making a crappy RPG in 7 days

Around july I was sorta burnt from working on my current project for awhile, and looked for a game jam to work on something new. There wasn’t much game jams that interest me, so I decided to do a sorta challenge. Make a RPG in a week which i thought was doable. I mostly wanted to do a RPG because they’re complex games to program and therefore made for a good challenge. I also decided to do random generation for the world so i won’t be spending my time on On the first day i worked on tiles, and generation of the map. got everything done in the morning, went to bed and got a algorithm for the dungeons done when I woke up Often time in game jams I have a issue where I spend more time thinking about a how I’d implement a mechanic then actually spending the time to implement it so this time I decided to adopt the “don’t think just do it“ mindset

With the 2 night coming i decided to work on the Combat System, one thing that I guess was stupid of me to not think about was the fact that a final fantasy style RPG system would require alot of UI programming. So after i got the change from map screen to battle screen working it was time for the actual battles and stuff. The reason why i have to do UI is that i want everything to be dynamic for multiple attacks, enemys, and party members. And Unity UI system as components for that, such as a layout components that allows you to add a new gameobject and sorts it’s position dynamically base on it’s parent. For example let’s say in my case you want to add a button that selects a specific attack. Well instead of instancing the button and setting it’s position in code, by getting the button for a attack before it and increasing the position for that for a certain amount. you just instantiate the button and the layout component for the parent it’s under just set’s it’s position under all the other button’s for you. Still instancing each button for each party member, enemies and attacks with their own menus is still a massive load. But i got them all done fairly early. After I finished that during the night, I drank some coffee and spend the rest of my day slowly grinding at the other stuff that comes with combat. which is mostly interfaces for Health, the role for party members, attacks, and enemies. after that was just smaller stuff like how each area(forest, plain lands, mountains) increase your chances of getting attacked and a battle returning to the overworld when you win the battle Now it times for day 3. With day 3 I decided to add stemia for each party member. It was my way of not spamming the most powerful attacks. and you regain your stemia after each turn. It was the only thing I did that day(except maybe fixing some bugs i guess sense I don’t list them and so I don’t remember). If I remember correctly it was a pain to implement sense I had to redo alot of the code for the battles just to fit it in. Other then that my slowness must be due to burnt out from the second day. And with that day 3 is done Onto day 4. I got a pretty big goal for day 4. I have to create a inventory menu that allows you to equip certain items and store all of them. which not only means i need a item interface that takes into account the properties of weapons and armor, but also menus that opens and closes, plus for polish sake i need to have it so it turns off all it’s sub menus menus when one closes as well. And for all that i had to create a fair amount of structs for the menus and a algorithm to close all submenus by searching all it’s children. And I had to make equipment menus that allows a player to select armor and weapons. And buttons for inventory’s for each item. Though most of the time was spend on making interfaces generalized and i didn’t even made the equiptment menu actually equipt items. Last thing I added real quick was money On Day 5 I now needed to do towns where you can shop and buy weapons, armor and items. the menus wheres simple enougn. already wrote alot of the code on day 4. the programming of the store and buying was sorta hard to program sense I wanted to be able to buy multiple items, I’d also needed to spawn the towns. Another thing that also took a bit was making the actual items have functionality in actually defending and attacking the enemies Ever sense day 3 I was already thinking of quitting the challenge like all my other game jam attempts. But i kept moving on until pass day 4 I’d realized that the project was not going to turn out how I wanted it too. I’d fell into the same issue I’d always go into with my game jam attempts. I dream too big. so at that point I felt pretty demoralized

eventually I’d realized something. While I do want to make a good decent game for once in my life, more then anything I want to go the distance. I just wanna get to the finished line even if the game is shit, even if it’s not what I’d image. More then anything I want to make it to the end and say “I made it“, no matter how flawed it is. So I kept going with the goal of slogging my way to the final day. I might not be a winner but at least I tried and stick with it through the end On day 6 I wanted to do different Biomes on the map which was a pain sense I’d have to make graphics for each Biome, all of which where slapped together The algorithm that generates it was a pain and took a bit of messing with to make it look meh. it had a very grid like look and some lakes disappeared when the biome changes. But it was good enough for my needs. After that I’d have to make the actual menus for equipping items...well work. I’d thought this was the hardest part sense In the way the inventory is coded Items are counted as Stacks, so instead of 5 buttons of the same sword. it’s just 5x sword. And I’d thought that It might make thing’s difficult with the way dropdown’s are done to make the actual number go down and re add the item when it’s empty. While it did took some fiddling(there where specify alot of errors with the fact that I copied code because Weapons and Armor are not the same thing). But other then that it went pretty well well it’s finally here. The last day and it was hell there where still alot of stuff I needed to do so i’m just gonna do whatever and call it done. One thing was some of the graphics. many of them still used the default unity UI. So I downloaded a free asset for pixelated fronts and used that. then I created my own sprite for the menus it self. Next on the list was the fact that the enemy and party member sprites where all white rectangles so I’d had to make sprites for the 3 classes I’d added(knight. wizard and archer), and now comes the next big issue. I have alot of enemies to create and i need to make sprites for them, and i had to code the biomes to make it so different types of enemies and battles spawn in different tiles. So I’d just create a bunch of dumb enemies classes. throw basic default AI with shit balancing. and downloaded alot of pictures vaguely related for their sprite(yeah it was lazy but I did found it funny). after all that I’d added them all to their respective biomes and called it quits and what’s the end product? well horrible. I’d completely neglect the weapons and armor. so shops and buying are useless. there’s a health potion that as no function to heal. and I’d completely forgot to make battles work for more then 1 enemies(sense when testing it I’d just had one) so it gives out a error when it’s not the first enemy turn and breaks the game. there’s also no leveling up in the game because I’d forgot halfway through development and in the end I’d didn’t have the time to add a big feature. But I’d stuck with it and honestly I think that’s good enough for now

[lonelygamedev]

I have a twitter now

follow me at @LonelyGameDev1 for more gamedev stuff more content soon to come

[lonelygamedev]

Devlog 5: slowly but surely

So I been working on a lot of stuff, let’s talk about it. The first thing i’ll talk about is the flowfield and pathfinding, sense I sorta overshadowed those in devlog 3. A flowfield is just a grid of Nodes with a direction to them that is set by whatever algorithm you want. Most of the time it’s used for simulations of weather, fluid’s etc. I wanted to use them for pathfinding sense they’re really good and efficient for alot of enemys with the same path. So to create it I just followed the tutorial by Turbo Makes Games https://www.youtube.com/watch?v=tSe6ZqDKB0Y&ab_channel=TurboMakesGames The only real issue so far is that it causes a dip in the framerate in one the bigger rooms with a lot of Nodes. I could solve this issue by threading it, making it more efficient or whatever. But we have other issues to content with

The second issue is Behavior trees. With it I don’t have to rewrite code for every new enemy script. I just followed this tutorial and I have a framework for behaviors trees up and running quickly https://www.youtube.com/watch?v=F-3nxJ2ANXg&ab_channel=GameDevChef I got normal zombies and gun zombies up and running as well. I would make more enemy's but I want to keep my scope small and make at least one level fun with minimal amount of enemy's before I add more content. I took a brief break from development at that point Until I saw a random video that was recommended to me by Jonas Tyroller (the guy who made Islander’s and that wingsuit game). The video was about making a good decision in a game, which he describe as a pyramid https://www.youtube.com/watch?v=J_qQ76ouWXI&ab_channel=JonasTyroller So after watching the video I decided to look at my game again and see what I could do to make anything better or interesting. And I notice there was a obvious issue that I was oblivious too for months....The guns have infinite ammo, there is no interesting decision making when it comes to shooting except which gun you buy. And once you get the best gun you are practically invisible. So I added reloads (sorta similar to the game SYNTHETIK), ammo and clip. With that done I also need a way to buy more ammo and an upgrade to akimbo mode. So i had to redesign the store menu system again (this was like the 4 time lol). Sense it’s menu’s where not coded in a way that supported multiple menu for guns in your inventory And that’s about all I added. There are still some thing’s I need to get done, Gun enemys don’t exactly work well with flanking and don’t avoid zombies. And the combat in the first level still isn’t exactly “fun“ but hey I’m getting there

[lonelygamedev]

Devlog 4: update

yes i’m still working on the game despite no update after awhile. i still need to get some more work done. i mostly got gun enemys to work with behavior trees and flanking behavior. i just need to do some modification to the design and such. life been busy as usual with chores and homework so yeah