I learned that you can dynamically change aspects of a texture in realtime without having to compile it every single time, by using parameters. By adding parameters for size, color tint, and the flatness of the normal maps I can adjust all of it on the fly.
I have a feeling that Blueprints are going to get very complicated very quickly. Some Spanish Brick will look much nicer than the ugly grass from before:
This spanish brick texture is at '0.0' for normal map flatness, let's crank the strength up to '50.0' and see what happens:
Q.) What kind of normal map do you want?
A.) Just fuck my shit up, fam
Let's try something more reasonable, like '5.0':
That's a lot better. We went from "PS2 game" to "Original X-Box game" to "bathroom tile grout at my first apartment".
The next part is dealing with creating Master Materials, and also importing static meshes. Guess what the tutorial mesh is?
Fucking CRAAAAAAAAAAAATES!
Needs... more... crates...
Old man Murray eat your heart out, I've got a StC of negative two years.
The other big thing was different textures for things like metallic and roughness maps, and how its fairly common to do multiple maps in the same texture and isolate them in individual RGB color channels so you can just import one item instead of three.
The next section is on lighting and Lumen, so I think I'm going to call it a night here.
Does anybody here work in the industry? I'm a senior web developer of 12 years giving serious thought to leaving my job and doing something different. I know the pay is much less and that's fine, but my main concern is avoiding crunch culture - I've heard the horror stories and I don't think I could tolerate that. Curious to hear thoughts on working for big studios vs indie devs. I'm kind of leaning towards indie work because I work for a tech giant now and I suspect working on a AAA game would feel too corporate. Plus I like the idea of smaller projects I can have a large impact on. I have a strong software resume but my game dev experience is hobbyist only, almost all 2D so far, so I probably have to do something closer to entry level. Any advice on applications/interviews?
For a while now I've sort of had a hangup about doing this for a living and instead wanted to be a solo dev making a passion project on my own terms - I'm not sure if I have the personality for it though. When I'm responsible for every aspect of the game I find myself always plagued with doubts, and I compare the various merits of my game unfavorably to those that are built by experienced teams. I think having some external responsibility attached to it would be good for me.
I can tell you, I will never have the skills to be a Real Developer™. I do not have the discipline and when stuck will spin my wheels until I burn out. The fact I'm ham-handedly managing two active projects right now is silly with a full time non-dev job.
It's nice that if you have an idea and others get excited about it, it gives a bit of an oomph to push ahead. Then I hire someone if I need the help and who knows.
I can't imagine working full time coding, or ever working for a AAA. This is just fun...
I mean, I just learned about libcurl the other day and I am so in love I don't know what to do with it.
I don't think it takes a uniquely gifted person or anything to be a professional software developer. A relevant education is somewhat important for finding a first job, but making games as a hobby isn't nothing. Web development in particular is always popular and really accessible IMO.
You know, for fun, I think I'll list out my backlogged projects... What they are why they stopped...
Atlantic - A remaster of Monopoly, with extra effort to "hide" that it's the exact same game - Held for architectural issues. battleship - Teaching my nephew how to code games. - Held as he just moved into a new place. blind - a 3d first person action adventure game with only sound. held because sound effects are hard. (Maybe hire a sound person?) bsd - various dabbles into the BSD games and see if something remastered can come from them. (Open source project if something works) - Nothing struck me Desert Bus 3D - someone else did it, but I did make a cool bus in blender! Gaia - remake of sim earth (Open source) - currently active GalaxyNGX - update interface to Galaxy NG PBE 4x game, just found libcurl... very tempting Harmony - Attempt to glue Visual Novel extensions to Inform 7.. Graphics lib was converted to Hikari Hikari - Game to teach Japanese though visual novel/interactive fiction - 6th attempt at "Pineapple Chronicles", but it's not really a Pineapple game so this attempt doesn't count. - currently in work. Inside - A remake of Laser Surgeon: The Microscopic Mission - I have some wormhole code that I need to convert from allegro 4 to allegro 5 Pineapple Chronicles - Pineappple is dead! Long live Pineapple. (See Hikari) RPG Shell - Wonder your filesystem like it's a 2D JRPG? Ehh. I had ideas..... The Guy In the Chair - a game where you played the guy in the chair guiding a Superhero in a city. Text/Unicode/Terminal based - tried to make a Simcity 2000 city loader (For the city the hero would be in) by converting some c++ code I found to C. Failed. I should have just wrote my own loader. Tokaido - A walking sim / treadmill VR game where you walked all 315 miles of the Tokaido Road in Japan cira 1850... Turns out 315 miles is a long walk...
Whatcha guys got on your back burner? Want me to elaborate more on my projects?
haha I do the same thing with my projects, finishing a hobby project is a lot harder than actually doing my job. I've never made a game I would call finished.
Magicario: Magic-based RPG battler, my current project. Considering some major design changes at the moment. https://zek23.itch.io/magicario
Goblio: Incremental game about spawning goblins to kill villagers. Might go back to this some day. https://zek23.itch.io/goblio Final Fantasy clone: An earlier attempt at an RPG battler from the side perspective. I was sort of happy with this one but didn't take it very far anyway. Platformer shooter: Short-lived attempt to make a platformer, didn't really have a strong direction for this one. Zelda clone: Kind of a Binding of Isaac type thing. My first real game attempt. This one turned out okay but I lost interest.
My list of unfinished projects would bust the character limit for a single post but every time I work on something I learn something, I also spent 4-5 year doing monthly/seasonal/yearly gamejams allowing me to learn many aspects of UE4 very quickly.
3 years ago I switched jobs from logistics desktop/web software to flight simulators used in flight training. So I'm still not a gamedev even though I do the exact same things but without the stress and weird shipping deadlines.
So I revisited "The Guy In the Chair" (from above) and that was a blast from the past.
I remember I was having some problems figuring out how pointers worked with structs while deciphering the C++ code template I had to load a SimCity 2000 file. I was just going through what I had done, and WOW it was obvious I had no idea what I was doing. (I have since taken some online classes in C, especially pointers and structs), and there was all kinds of nonsense I obviously was just guessing at to get to work.
First of all, you can't pass by reference in C and that & in the "function" declarations was just kicking my butt. (PROTIP: you pass the pointer). There was junk cut/paste all over the place with no reason what so ever, and I had an overabundance of structs that I ripped from classes not knowing why they were built that way. I had a fun time turning everything into actual C code.
So it's not a new years resolution, but I promised an hour a day of coding... Doesn't matter what. I pick one of my dumb projects and work on it for a bit and see what happens. My GitHub commits look like this now
I never really did anything with GitHub, but now if I have worked on it, and it compiles cleanly, it's getting a repo. I kind of like how that looks.
== EDIT ==
Oh, and also, I was also dead set on using Code::blocks to code as I'm super cozy with GCC, but now... honestly, I'll fire up Visual Studio and just go. I actually try and make all my code cross-platfrom so plopping it into VS has been super easy, barely in inconvenience, and now that the newer versions have dropped TFS for a direct GitHub access. It's a dream to just make a new project, drop in some C files, and get going...
I really... REALLY, wish there were some C templates in there so I didn't have to rename all of new files I create. (And to add templates is a bit of black magic itself.)
Speaking of which...
WHY ON CARL SAGAN'S PALE BLUE DOT IS IT SO HARD TO MAKE A SNIPPIT!
Still plugging away at Nimsters. This week was mostly focused on the map screen, which is more like a Darkest Dungeon-style collection of rooms than the map screen in Pokemon. The rooms can be defined with a number of properties, such as:
- Which directions you can exit the room; this allows you to create one-way corridors or simply have adjacent rooms with a wall between them.
- A trainer present in the room, who will challenge you to a battle when you step into the room until defeated, at which point they may be optionally re-challenged (though for diminishing monetary rewards each time until you go pick on someone else).
- An event that automatically triggers when you step into the room. (This is basically just an incredibly simple scripting block; my test event increases a counter every time you enter the room and eventually triggers a selectable event that questions why you're walking in circles.)
- A set of events that can be presented as options within the room. (i.e. "1. Read the important-looking poster on the wooden sign.")
- A terrain type that affects how the room is rendered and can present optional requirements to enter it, such as having a Nimster of a particular type and level. (I sort of designed and implemented this without really considering whether it really improves the game; the idea is to mimic something like the HM system. But it's there now, so I can always just not define any room requirements if it's too annoying.)
- An encounter table that indicates how often and what kinds of wild Nimsters will attack when you enter that room.
The next item on my checklist is one of the classic banes of RPG programmers: inventory items. I'm working out how I want to represent them in-code and let the player manipulate them in the UI.
Right now I'm sitting around 60 ideas for monsters. Almost entirely out of just smashing a noun and an animal together. Like... "Fossilot"! The skeletal cat Nimster!
Kupi on
My favorite musical instrument is the air-raid siren.
At this point the tutorial was going over the differences in traditional baked lighting and the newer Lumen system.
UE4 style lighting. Old, harsh, shit from a butt:
UE5 lighting with Lumen. New, fresh, global illumination in realtime without using raytracing and melting your GPU into slag:
The bounced lighting giving the sphere a hint on green on the left and red on the right is pretty neat.
If we nuke the light inside the cube, but add a sky atmosphere and sky light, the cube gets a little bit of light from the sky.
But if I add a bigass bright white wall
than I can reflect a ton of sunlight into the cube and light it up that way as well. Neat.
Next was setting up and going over the differences with archvis lighting in a pre-made room.
Starting off with adding lighting to the environment using Lumen, it looks pretty decent
The next step was to change all of the lighting to static and then bake it. You have options for setting the quality of the lighting, going from Preview to Medium to High to Production. The dude in the tutorial set his to 'Medium', but I figured I'd see what Production looked like. What's the worst that could happen?
12 minutes later I got this:
Much nicer. So it makes sense that you would want to use Lumen in a situation where you were going to have dynamic lighting like a day/night cycle or you wanted something cool like a cyberpunk alley with tons of neon lights or whatever, with the caveat that it is more computationally intensive when being run. On the other hand, you can use static lighting in an environment and get higher quality lighting that is less resource intensive, but with the caveat that your lights and shadows are baked in and can't be altered.
The next section will be on how to utilize the landscaping tool, which looks like a lot of fun.
Also Zek your videos are showing as private for me.
+1
Options
cj iwakuraThe Rhythm RegentBears The Name FreedomRegistered Userregular
Seeing how small(?) teams managed to make visual novels out of both Vampire and Werewolf and somehow get Paradox to back them and release them on Steam makes me wonder how feasible it would be to do one based on Mage.
I've been having second thoughts about joining the industry, though I sent in some applications and already have a phone interview scheduled so we'll see how that goes. I had another thought about a possible "best of both worlds" trajectory - be an indie game designer, keep working my actual job, and pay developers/artists to help implement my game ideas. I live a frugal lifestyle and am fortunate enough to be able to do that without an expectation of making my money back. If I'm happy with what ends up being created I could put it on Steam and see what happens, or do a Patreon or something, but I wouldn't have any expectations.
Is this something I can do totally ad hoc "under the table" or do I need to form an LLC or something? I don't have a clue about legal/tax implications etc. I don't think I'd be paying full time salaries, probably just finding people at an hourly rate on Reddit or something.
I've been having second thoughts about joining the industry, though I sent in some applications and already have a phone interview scheduled so we'll see how that goes. I had another thought about a possible "best of both worlds" trajectory - be an indie game designer, keep working my actual job, and pay developers/artists to help implement my game ideas. I live a frugal lifestyle and am fortunate enough to be able to do that without an expectation of making my money back. If I'm happy with what ends up being created I could put it on Steam and see what happens, or do a Patreon or something, but I wouldn't have any expectations.
Is this something I can do totally ad hoc "under the table" or do I need to form an LLC or something? I don't have a clue about legal/tax implications etc. I don't think I'd be paying full time salaries, probably just finding people at an hourly rate on Reddit or something.
So, I don't make a living off of this, but I do make some small amount of money and spend a larger amount of money. Assuming you are American: To get started, you don't have to do anything special business-wise, you can work as a sole proprietor, and this is fine for working off of Patreon or before you have any actual money coming in. Freelancers can be paid as contractors, but if you pay another American more than $600 in a year you will have to get their tax info and send them a 1099-NEC the next January (I literally have some sitting in envelopes here right now). Also: absolutely find a Work For Hire contract template online and get them to sign it.
Once you get to Steam and other storefronts you probably will want a business bank account and more legal protection, so it's LLC time. Here in Virginia I was able to form a single-person LLC, which is an LLC except my taxes work no differently than as a sole proprietors. As soon as you add a single other full employee though it gets more complicated and that's beyond my experience.
It's impossible to bring in money under the table, every site that handles money will report your income, you could pay people under the table but that means you don't get to write off the expense on your taxes so you're costing yourself money at that point.
Made some more progress in my ugly programmer GUI this past week or so, with my semi-card game style space combat encounters.
First, I've set up the 'battlefield', which has the player's bridge stations set up on one side. Each bridge station may have up to four different actions they can perform, in my Tutorial Mode ship here they all only have a single option, but I will have other encounter types coming down the pipe where there will be multiple action options right from the start.
Each station has a number of turns they require to 'charge up', and turns are consumed by issuing Commands. The 'deck' of commands available to you is made up by the traits of your currently assigned bridge crew, from which you draw a random hand. These commands will allow you to do things like power up actions before they trigger, but there's also trait effects like debuffing the enemy or adding a chance for random resource rewards and things like that. My next task will be actually applying the effects of commands once selected.
I've been having second thoughts about joining the industry, though I sent in some applications and already have a phone interview scheduled so we'll see how that goes. I had another thought about a possible "best of both worlds" trajectory - be an indie game designer, keep working my actual job, and pay developers/artists to help implement my game ideas. I live a frugal lifestyle and am fortunate enough to be able to do that without an expectation of making my money back. If I'm happy with what ends up being created I could put it on Steam and see what happens, or do a Patreon or something, but I wouldn't have any expectations.
Is this something I can do totally ad hoc "under the table" or do I need to form an LLC or something? I don't have a clue about legal/tax implications etc. I don't think I'd be paying full time salaries, probably just finding people at an hourly rate on Reddit or something.
So, I don't make a living off of this, but I do make some small amount of money and spend a larger amount of money. Assuming you are American: To get started, you don't have to do anything special business-wise, you can work as a sole proprietor, and this is fine for working off of Patreon or before you have any actual money coming in. Freelancers can be paid as contractors, but if you pay another American more than $600 in a year you will have to get their tax info and send them a 1099-NEC the next January (I literally have some sitting in envelopes here right now). Also: absolutely find a Work For Hire contract template online and get them to sign it.
Once you get to Steam and other storefronts you probably will want a business bank account and more legal protection, so it's LLC time. Here in Virginia I was able to form a single-person LLC, which is an LLC except my taxes work no differently than as a sole proprietors. As soon as you add a single other full employee though it gets more complicated and that's beyond my experience.
It's impossible to bring in money under the table, every site that handles money will report your income, you could pay people under the table but that means you don't get to write off the expense on your taxes so you're costing yourself money at that point.
Thanks that's really helpful - how do the tax write-offs work? Do I need to be running an LLC and/or intending to sell a product for my payments to freelancers to be a business expense?
I've been having second thoughts about joining the industry, though I sent in some applications and already have a phone interview scheduled so we'll see how that goes. I had another thought about a possible "best of both worlds" trajectory - be an indie game designer, keep working my actual job, and pay developers/artists to help implement my game ideas. I live a frugal lifestyle and am fortunate enough to be able to do that without an expectation of making my money back. If I'm happy with what ends up being created I could put it on Steam and see what happens, or do a Patreon or something, but I wouldn't have any expectations.
Is this something I can do totally ad hoc "under the table" or do I need to form an LLC or something? I don't have a clue about legal/tax implications etc. I don't think I'd be paying full time salaries, probably just finding people at an hourly rate on Reddit or something.
So, I don't make a living off of this, but I do make some small amount of money and spend a larger amount of money. Assuming you are American: To get started, you don't have to do anything special business-wise, you can work as a sole proprietor, and this is fine for working off of Patreon or before you have any actual money coming in. Freelancers can be paid as contractors, but if you pay another American more than $600 in a year you will have to get their tax info and send them a 1099-NEC the next January (I literally have some sitting in envelopes here right now). Also: absolutely find a Work For Hire contract template online and get them to sign it.
Once you get to Steam and other storefronts you probably will want a business bank account and more legal protection, so it's LLC time. Here in Virginia I was able to form a single-person LLC, which is an LLC except my taxes work no differently than as a sole proprietors. As soon as you add a single other full employee though it gets more complicated and that's beyond my experience.
It's impossible to bring in money under the table, every site that handles money will report your income, you could pay people under the table but that means you don't get to write off the expense on your taxes so you're costing yourself money at that point.
Thanks that's really helpful - how do the tax write-offs work? Do I need to be running an LLC and/or intending to sell a product for my payments to freelancers to be a business expense?
No, as a sole proprietor you fill out the taxes under your own name/SSN. Schedule 1 for adjustments to income and C for contract labor and other expenses. If you really never intended to make any money from it and you're just writing off expenses on a hobby I imagine that could get you in trouble, but I'd expect it'd require years of returns and a lot of money before it triggered an audit (just a guess, not an expert).
Unrelated, like Handkor mentioned with flight simulators, my day job involves making training software in Unity. It doesn't have the 'fun' aspect of game design to it, but then programming games isn't the same as playing them anyways. Stuff like that is a good way to get experience with gamedev tools without going into the hellscape that seems to be game dev companies.
I've been having second thoughts about joining the industry, though I sent in some applications and already have a phone interview scheduled so we'll see how that goes. I had another thought about a possible "best of both worlds" trajectory - be an indie game designer, keep working my actual job, and pay developers/artists to help implement my game ideas. I live a frugal lifestyle and am fortunate enough to be able to do that without an expectation of making my money back. If I'm happy with what ends up being created I could put it on Steam and see what happens, or do a Patreon or something, but I wouldn't have any expectations.
Is this something I can do totally ad hoc "under the table" or do I need to form an LLC or something? I don't have a clue about legal/tax implications etc. I don't think I'd be paying full time salaries, probably just finding people at an hourly rate on Reddit or something.
So, I don't make a living off of this, but I do make some small amount of money and spend a larger amount of money. Assuming you are American: To get started, you don't have to do anything special business-wise, you can work as a sole proprietor, and this is fine for working off of Patreon or before you have any actual money coming in. Freelancers can be paid as contractors, but if you pay another American more than $600 in a year you will have to get their tax info and send them a 1099-NEC the next January (I literally have some sitting in envelopes here right now). Also: absolutely find a Work For Hire contract template online and get them to sign it.
Once you get to Steam and other storefronts you probably will want a business bank account and more legal protection, so it's LLC time. Here in Virginia I was able to form a single-person LLC, which is an LLC except my taxes work no differently than as a sole proprietors. As soon as you add a single other full employee though it gets more complicated and that's beyond my experience.
It's impossible to bring in money under the table, every site that handles money will report your income, you could pay people under the table but that means you don't get to write off the expense on your taxes so you're costing yourself money at that point.
Thanks that's really helpful - how do the tax write-offs work? Do I need to be running an LLC and/or intending to sell a product for my payments to freelancers to be a business expense?
No, as a sole proprietor you fill out the taxes under your own name/SSN. Schedule 1 for adjustments to income and C for contract labor and other expenses. If you really never intended to make any money from it and you're just writing off expenses on a hobby I imagine that could get you in trouble, but I'd expect it'd require years of returns and a lot of money before it triggered an audit (just a guess, not an expert).
Unrelated, like Handkor mentioned with flight simulators, my day job involves making training software in Unity. It doesn't have the 'fun' aspect of game design to it, but then programming games isn't the same as playing them anyways. Stuff like that is a good way to get experience with gamedev tools without going into the hellscape that seems to be game dev companies.
A: re: riding the employee/independent divide, in my country one can be both and simply declare the income. Crucially, one should become a part-time independent while still employed, which means that if you are fired from your daytime job you will get unemployment benefits *and* remain part-time independent.
However, I kind of agree with those people who say that for the part-time developer gig to become a paying job you *just* need to jump in as a full time indie. If you are an MBA with wicked networking skills and your idea of chilling is community managing while simultaneously haggling with peeps on fiver, you could accidentally release a game successfully without actually having developed a game.
If art, programming, game design, writing, QA or a subset thereof are what you do naturally, you can certainly develop a game, but you'll need the above MBA skills to actually launch it and make money of it. If those MBA skills are not part of how you approach this hobby naturally, having to do them to pay rent is a good kick in the pants if there's no-one else around to deliver it.
B: for my upcoming games, I'm going to ask around for someone who is interested in applying their MBA/community/marketing skillset to a game and a game designer/QA person. I've moved along my 'reading up on' skill curve to the point where people are repeatedly saying 'you need more than one founder' for startups. Related post in a second.
C: as a manager in Tech Support (YMMV), what I like in employees is accountability. I don't care if you make mistakes, but I need you to own up to them, be able to listen to me explain what went wrong and what you should have done, and I need to not have to repeat that four times. I have (had) people with less than 12 years of experience who have joined with that attitude, and they were a god-send combined with the young kids, whom I can assign the schedules that are not possible for the 12 year veterans that have to pick up their kids from school.
So I've been working on my current game- days away from a vertical slice- with a guy. He delivers the world which comes from his ongoing development of a stop-motion television series. I've had some artwork from him, but mostly I've recreated physical models that he's gotten over the years from people he works with.
I intended to have a synergy between game and tv series proposal on the marketing side and there has been, but not a LOT.
Recently he got a contact with someone who produces Unity games and wants to look at the vertical slice
Because of my work/family life/development balance, I was intending to do a veeery soft release of the vertical slice as a demo. I don't have anything remotely resembling a pitch. I don't have a good trailer. I don't have a business plan on how to get to fully-fledged game from vertical slice.
Help?
I could use some good instructionals on how to create a pitch deck, as well as a business plan. I do have a more elaborate game pitch that I got from a friendly developer and I could work off, but it's all hands on deck here as my partner wants to talk to the producer *soooon*.
Seeing how small(?) teams managed to make visual novels out of both Vampire and Werewolf and somehow get Paradox to back them and release them on Steam makes me wonder how feasible it would be to do one based on Mage.
Mage: The Ascension would have to be limited to like-- Sphere 2 tops because whoooo boy once you hit 3 in a Sphere shit gets off the hook in most cases. ESPECIALLY with certain combinations.
Don't get me wrong I would buy the SHIT out of it even if it's hyper limited but properly translating the sheer versatility of Sphere Magic is not really doable in a non-dynamic fashion. Too many possibilities.
So I've been working on my current game- days away from a vertical slice- with a guy. He delivers the world which comes from his ongoing development of a stop-motion television series. I've had some artwork from him, but mostly I've recreated physical models that he's gotten over the years from people he works with.
I intended to have a synergy between game and tv series proposal on the marketing side and there has been, but not a LOT.
Recently he got a contact with someone who produces Unity games and wants to look at the vertical slice
Because of my work/family life/development balance, I was intending to do a veeery soft release of the vertical slice as a demo. I don't have anything remotely resembling a pitch. I don't have a good trailer. I don't have a business plan on how to get to fully-fledged game from vertical slice.
Help?
I could use some good instructionals on how to create a pitch deck, as well as a business plan. I do have a more elaborate game pitch that I got from a friendly developer and I could work off, but it's all hands on deck here as my partner wants to talk to the producer *soooon*.
Spoke to a friend who is also a game developer, albeit much more accomplished and running a company. Basically, the problem with my current game design is that it's not designed from the ground up to be free to play with monetization options. I.e. I can pitch all I like, but no publisher would take it. He just pitched 30 publishers so I assume that's indeed the state of the art.
Monetization options are; IMHO:
-hats: updating your character or fighter with cosmetic changes. Doable, but I'd need to revamp the UV mapping on my fighter jet and figure out a way to have high resolution masks work with it. I can picture using meshes in a back room and a camera to snap then save to disk? Or, I use a purely script based texture generator- the problem is shadergraph doesn't seem to do geometry very well. That's about a month or two for just the fighter, without adding characters, animation, textures for the characters and hats in.
-weapons: buying in-game content with functional advantages. Difficult to work out what's needed here. It's a single-player game, with very simple mechanics. Normally, the techtree will be extra storage for (or smaller) 'drop nodes' and better fighter jets. I can imagine selling assistive technology, autopilot etc. but then the game might get too easy as there's only so much I can do to increase difficulty. (That's a design issue.) Main issue here is that it would need more content added to create extra difficulty, for example smart generation of obstacles. For the content on sale, placeholders would likely do for the pitch.
-nagging: get the game to a point where it's fun enough that players will play even if I add recharge countdowns or adverts, but will pay to skip them. It feels counterintuitive to polish the game just to then introduce such crippling. Not the person I want to be, really, even if it is the person that gets to pay the rent.
Battlepasses seem to be the current Free to Play flavor of the month. It also gives you the option to sell level skips or exp boosts
But this is starting to get into the Live Ops world of post release content. Daily/Weekly quests. New characters. Seasonal events.
I was just the other day explaining why, 20 years ago, my web developer friend did offer content updates to his clients, but at steep rates that got steeper the bigger the update. You want to spend your time building new websites (he rewrote a CRM yearly for fun) and not upkeeping websites in a stressful pingpong loop with a client while losing programming skills.
In game dev, afaik, continuous updates lead publishers and developers to drop talent and invest in juniors who get little freedom as data is driving 'innovation'. Because the numbers show what's popular so any investment can go to sure things. Who like sure things? Investors like sure things. Ask Hollywood.
We may all think we like a sure thing, but that's my day job, not the space I want to creatively explore.
I could certainly use fiver to have new content- as long as it paid the bills- and even apply myself to big additional mechanics updates. But the latter always risk breaking the game or not making a splash. I can take that same risk on a new game entirely and have more fun developing it.
Because of that it has a typical static mesh wireframe
But if I enable Nanite for this particular static mesh
than is no longer has a typical wireframe because the number of triangles is changing dynamically based on the distance from the camera, voodoo magic. The effect is significantly more impressive when importing Nanite quality meshes from Quixel instead of doofball ones like this, but still.
"Playtesting":
Using Unreal's built in third person package and the generic model I can adjust the settings for the camera. From here I can enter the environment I built as if it was a third person videogame by using the Play feature...
... which spawned the character beneath the level, falling swiftly into the void. That's no good!
But by adding a playerstart actor...
... I can dictate where the playable character spawns. Cool beans.
Maeking an actual gaem:
It's an obstacle course. Here's the layout. The player spawns at one end, and the statue is at the other. The cylinders rotate at different speeds, each one is attached to a blueprint that, if the player character collides with it, will ragdoll him (effectively player death). Touch the statue at the end and you win
anakinskywalkeritsworkingitsworking.avi
Collision with the cylinders ragdolls the player using Set Simulate Physics
Touching the statue creates text in the viewport and acts as a win condition. Technically it's a videogame. The worst videogame ever, but still, a videogame. Ship it!
I have now officially created more bug free videogames than Derek Smart.
Yeah I played around with nanite too over the holidays and the tech is insane. I can just keep adding these high poly building towers in the background using the same assets as the close up ones with no impact.
As for what is happening in the screenshot, this little droid wants to be pretty. So he runs around wearing wigs stealing makeup out of purses only to run back to his little hidden corner to a mirror where you can paint his face with the makeup.
Seeing how small(?) teams managed to make visual novels out of both Vampire and Werewolf and somehow get Paradox to back them and release them on Steam makes me wonder how feasible it would be to do one based on Mage.
Mage: The Ascension would have to be limited to like-- Sphere 2 tops because whoooo boy once you hit 3 in a Sphere shit gets off the hook in most cases. ESPECIALLY with certain combinations.
Don't get me wrong I would buy the SHIT out of it even if it's hyper limited but properly translating the sheer versatility of Sphere Magic is not really doable in a non-dynamic fashion. Too many possibilities.
Yeah, it would have to be super limited, but I'd love to think of a way, because I love the lore and worldbuilding in Mage(Awakening) to pieces.
Mini-update: I decided that the roguelite progression where you have a long run during which you accumulate gear wasn't really working. Partly because I'm not confident the combat would be fun for uninterrupted runs of 10-20 minutes which would be needed for the equipment acquisition to build up to something meaningful. So instead I decided to focus on quick runs, just one area at a time, with permanent progression in the form of the equipment you find every 5 battles. Then on the menu you can equip or salvage it. For some reason I'd been dreading building the equipment UI but I simplified the design to make the work more palatable.
Depends on what you're trying to do I guess. Looking at your profile, text adventure games /interactive fiction?
If so, depends on what you're trying to achieve.
-Getting lots of feedback from players? Look up which platform gets a lot of players; might even be a browser-based one.
-Develop your own world? Use existing tools for ease of use and support: https://itsfoss.com/create-interactive-fiction/
-Grow your skills? You can start from scratch learning programming, or you can start with simple games in an editor that has options to do a lot more, such as Game Maker.
soo... talk to us!
I skipped the report last week because it was AGDQ week, and while I was poking at a little side-project, I didn't have enough completed to be worth typing up a report. And, for that matter, I was too invested in watching AGDQ to bother typing anything up on Friday. Six of one, half-dozen of the other.
My latest project was more a matter of personal curiosity than anything truly useful; nothing new in that respect. Though I've used some ready-made randomization methods in my engine in the past, I wanted to explore alternatives and/or just build my own. The practical upshot of that work is that each individual entity that wants to use randomization carries its own randomizer state around with it, and there's a central "randomization seeds" component which exists solely to issue seeds to new entities' randomizers as they're created. Though the method I use is unorthodox (using addition and treating the randomizer state as a stream of bits instead of multiplication, bit-shifts, and XORs and consuming the entire word at once), the way the seed generation and the individual randomizers combine produces sufficiently differentiated random results with a distribution that resembles the ones produced by the Xoroshiro+ and Mersenne Twister PRNGs that I'm happy with the results.
On the Nimsters front, my Git log is showing that the majority of my work this week was on inventory items, as I mentioned last time. In Nimsters, there are four different types of items, all of whose effects are now implemented in the game:
- Restoratives: One-time use items that can restore health, stamina, or status effects, alone or in combination.
- Accessories: The Nimsters equivalent of held items from Pokemon; equipped to a Nimster to produce a passive effect of some kind. No accessory currently produces an actual effect, but hey! They can be equipped, which is all the requirement doc called for.
- Treats: Have two different functions; used on a wild Nimster, it restores its health and adds to its "respect" value, which makes it more like that Taming nodes will show up on the timeline that allow you to tame the Nimster and add it to your party if you land on them. Used on a Nimster in your party between battle, they function like Rare Candies and their derivatives, granting variable amounts of EXP.
- Supplements: The equivalent of Pokemon's Protein/Carbos/Iron/Zinc items, except that they modify the Nimsters equivalent of IV instead of EVs.
I also implemented the wild Nimster encounter tables, so you can be attacked while wandering around the map. The next feature on the roadmap is actually getting the taming mechanics working, so Nimsters can be a proper monster-collection game at last.
My favorite musical instrument is the air-raid siren.
So despite the ever tempting allure of shiny new project, simearth in UE5 go make a planet-generator (still doing research for that one), I'm doing more logic game stuff, and I've finally decided to bite the bullet and do my long put off graphical overhaul. As much as I do like the pure schematic look I accept that I'm going to need to properly show port values, it was something I quickly hacked in for my own use after all, which entails a redesign of the basic structure of the entities, they're too small to show everything and I recently confirmed that yes scrolling does still work even through it's not hooked up to a button yet so I'm no longer confined to a single screen. For my thing with 20 outputs... well I'll figure something out. I'm probably also going to make the number the key "port" that people interact with, currently it's centered over the name, and also distinguish between clicking on a port to draw a wire and clicking on the entity to move it or whatever. I'm also using a bunch of kludgy stuff like 3 hardcoded shaders I just switch between as needed and the UI objects themselves are getting too clunky and tied to the bits I want to be independent - the thing that manages the simulation is for some reason also in charge of where you're allowed to move things on the grid! I'm also running into more and more disontinuities - render objects that need to exist but have no direct connection to actual simulation entities so my current 1:1 design is no longer really workable
So naturally I also decide to move all the UI bits from C++ to lua (currently main menu is lua but in-game is C++) which as a bonus makes things Moddable and enforces Separation of Concerns, but it's the main game that needs the special shaders and I can't just accept whole shader source straight from lua oh no that won't do at all
Now I'm sitting in my "engine" project defining a complicated shader generation model that can have 8 textures of varying types and accept arbitrarily formatted vertex streams. Why? Nobody else will ever use this, even if I put it up on github. I don't even need it, if I need lighting and shadows and fancy materials and all that I can just go fire up unreal engine. I think I am actually incapable of scoping any personal project appropriately for a solo developer
Kupi's Weekly Friday Game Dev Status Report
My latest project was more a matter of personal curiosity than anything truly useful; nothing new in that respect. Though I've used some ready-made randomization methods in my engine in the past, I wanted to explore alternatives and/or just build my own. The practical upshot of that work is that each individual entity that wants to use randomization carries its own randomizer state around with it, and there's a central "randomization seeds" component which exists solely to issue seeds to new entities' randomizers as they're created. Though the method I use is unorthodox (using addition and treating the randomizer state as a stream of bits instead of multiplication, bit-shifts, and XORs and consuming the entire word at once), the way the seed generation and the individual randomizers combine produces sufficiently differentiated random results with a distribution that resembles the ones produced by the Xoroshiro+ and Mersenne Twister PRNGs that I'm happy with the results.
I might have mentioned this before, but I created a small (0-9 squared) scrambled lookup table. I don't really need 'good enough' randomization as much as I need random-looking complexity. In fact, for my procedural generation I wanted the exact same outcome for every map tile or prefab combo, so that even though the city I generate is endlessly complex, it does not change between plays.
So for that purpose, a lookup table does the trick!
Kupi's Weekly Friday Game Dev Status Report
My latest project was more a matter of personal curiosity than anything truly useful; nothing new in that respect. Though I've used some ready-made randomization methods in my engine in the past, I wanted to explore alternatives and/or just build my own. The practical upshot of that work is that each individual entity that wants to use randomization carries its own randomizer state around with it, and there's a central "randomization seeds" component which exists solely to issue seeds to new entities' randomizers as they're created. Though the method I use is unorthodox (using addition and treating the randomizer state as a stream of bits instead of multiplication, bit-shifts, and XORs and consuming the entire word at once), the way the seed generation and the individual randomizers combine produces sufficiently differentiated random results with a distribution that resembles the ones produced by the Xoroshiro+ and Mersenne Twister PRNGs that I'm happy with the results.
I might have mentioned this before, but I created a small (0-9 squared) scrambled lookup table. I don't really need 'good enough' randomization as much as I need random-looking complexity. In fact, for my procedural generation I wanted the exact same outcome for every map tile or prefab combo, so that even though the city I generate is endlessly complex, it does not change between plays.
So for that purpose, a lookup table does the trick!
Oh, it gets worse in my case. I went through a couple drafts of the original post and decided that the internal workings of my new randomizer weren't interesting enough, but the exact definition of "good enough for my purposes" here is kind of funny, so here it is:
Because of the way my game's engine is structured (farming operations out to worker threads to be performed in parallel, but with an expectation that re-runs of the game with the same input will produce the same final state), I can't use a single source of randomness for all random events in my game. Every Entity has to have a Component which stores the state of its randomizer so they can update independently. Object creations are single-threaded, so ultimately what happens is I have a "randomization seeds generator" that increments by 1 every time you get a new seed, and a "randomizer" Component which holds the state.
My previous randomization implementation used the well-known Xorshift algorithm. But then I said to myself, "you know, really what's sufficient for my purposes is some algorithmic walk of the bit-space of a 64-bit word". It turns out that if you add a number with no 2s in its prime factorization something that fits in a power-of-2 data field (like an ulong in C#), it'll walk the entire available bit space. So I put together a small program to find a number which is a multiple of many prime numbers other than 2 which is just over half the bit space (reasoning: a large number of bits will change each time) (result: 3 * 5 * 7 * 11 * 13 * 17 * 19 * 23 * 29 * 31 * 37 * 41 * 43 * 1423 = 9308384687483226345) and added that number to the previous state after consuming all the bits in the previous state. After fooling around with it, I found out that this produces distributions that very often subtly favor 0 for reasons I wasn't able to determine analytically, and also have jagged distributions favoring some numbers above others.
That sent me back to looking at Xorshift, and when I looked it up on Wikipedia, I found that it had a period of 2^64 - 1. I went and read the paper where the genius who's been coming up with half the random number generators in Wikipedia's list of PRNGs originally published Xorshift, and one of the things in there was a proof that, so long as the seed state isn't 0, Xorshift performs a full walk of all the entire set of bit patterns in a 64-bit word. In other words, the thing I'd already used had the attributes I was already looking for (minus one value, which is no real loss).
Now I'm sitting in my "engine" project defining a complicated shader generation model that can have 8 textures of varying types and accept arbitrarily formatted vertex streams. Why? Nobody else will ever use this, even if I put it up on github. I don't even need it, if I need lighting and shadows and fancy materials and all that I can just go fire up unreal engine. I think I am actually incapable of scoping any personal project appropriately for a solo developer
Either that, or like me, you've got the problem where the coding is the fun part. I struggle with art and design, but code comes easily and gives me that "you accomplished something!" feeling when something compiles and does what it's supposed to. Thankfully I've well and truly run out of projects for my engine, so it's kind of forcing me into the design space, but it might help to reflect on your motives.
My favorite musical instrument is the air-raid siren.
Oh I'm absolutely having fun, it's just a very good thing I will never need to actually make a living doing this
As I sit here designing a cache-based system to draw text because I can't now assume that everything lies in a predetermined font I can blit at load time...
Do I draw the text to an intermediate buffer? Try to force glyphs into the same cache textures? How do I handle invalidation vs fast re-render? So many choices
Depends on what you're trying to do I guess. Looking at your profile, text adventure games /interactive fiction?
If so, depends on what you're trying to achieve.
-Getting lots of feedback from players? Look up which platform gets a lot of players; might even be a browser-based one.
-Develop your own world? Use existing tools for ease of use and support: https://itsfoss.com/create-interactive-fiction/
-Grow your skills? You can start from scratch learning programming, or you can start with simple games in an editor that has options to do a lot more, such as Game Maker.
soo... talk to us!
Thanks for the reply.
I was thinking of doing a turn-based puzzle thing, possible like Dragon Quest/Pokémon or even a Fire Emblem (I’m leaning more that way now, as it has that puzzle element that’s more a game to me).
Depends on what you're trying to do I guess. Looking at your profile, text adventure games /interactive fiction?
If so, depends on what you're trying to achieve.
-Getting lots of feedback from players? Look up which platform gets a lot of players; might even be a browser-based one.
-Develop your own world? Use existing tools for ease of use and support: https://itsfoss.com/create-interactive-fiction/
-Grow your skills? You can start from scratch learning programming, or you can start with simple games in an editor that has options to do a lot more, such as Game Maker.
soo... talk to us!
Thanks for the reply.
I was thinking of doing a turn-based puzzle thing, possible like Dragon Quest/Pokémon or even a Fire Emblem (I’m leaning more that way now, as it has that puzzle element that’s more a game to me).
Definitely cheap laptop territory. Also, this might be too ambitious for a first game (assuming this is your first). Do you have any computer science background or programming skills?
Depends on what you're trying to do I guess. Looking at your profile, text adventure games /interactive fiction?
If so, depends on what you're trying to achieve.
-Getting lots of feedback from players? Look up which platform gets a lot of players; might even be a browser-based one.
-Develop your own world? Use existing tools for ease of use and support: https://itsfoss.com/create-interactive-fiction/
-Grow your skills? You can start from scratch learning programming, or you can start with simple games in an editor that has options to do a lot more, such as Game Maker.
soo... talk to us!
Thanks for the reply.
I was thinking of doing a turn-based puzzle thing, possible like Dragon Quest/Pokémon or even a Fire Emblem (I’m leaning more that way now, as it has that puzzle element that’s more a game to me).
Definitely cheap laptop territory. Also, this might be too ambitious for a first game (assuming this is your first). Do you have any computer science background or programming skills?
Oh no! Not a bit! But I might manage an FE1 grade of game.
Might do a SHMUP first once I figure out what coding language to learn.
Depends on what you're trying to do I guess. Looking at your profile, text adventure games /interactive fiction?
If so, depends on what you're trying to achieve.
-Getting lots of feedback from players? Look up which platform gets a lot of players; might even be a browser-based one.
-Develop your own world? Use existing tools for ease of use and support: https://itsfoss.com/create-interactive-fiction/
-Grow your skills? You can start from scratch learning programming, or you can start with simple games in an editor that has options to do a lot more, such as Game Maker.
soo... talk to us!
Thanks for the reply.
I was thinking of doing a turn-based puzzle thing, possible like Dragon Quest/Pokémon or even a Fire Emblem (I’m leaning more that way now, as it has that puzzle element that’s more a game to me).
Definitely cheap laptop territory. Also, this might be too ambitious for a first game (assuming this is your first). Do you have any computer science background or programming skills?
Oh no! Not a bit! But I might manage an FE1 grade of game.
Might do a SHMUP first once I figure out what coding language to learn.
A loooong long time ago when I was learning and trying all sorts of game type I had made a small 1 screen boss rush SHMUP in BASIC, the older kind with line numbers that you either retyped your code or saved it to an audio cassette.
The rest of my path was Logo -> QBasic -> direct x86 assembly -> C/C++ in DOS with interrupt calls to the hardware -> C++ OWL in Windows 3.1 -> C++ OpenGL -> C++ DirectX/HLSL -> C# XNA and finally Unreal Engine where I no longer type any code and just do everything with Blueprints. It's like chill mode for weathered programmers.
Use any language that works for you but I suggest aiming for the ones that have the easiest path to display stuff on screen and apply logic to things. At the beginning you want to be managing non gaming/system stuff as little as possible.
Depends on what you're trying to do I guess. Looking at your profile, text adventure games /interactive fiction?
If so, depends on what you're trying to achieve.
-Getting lots of feedback from players? Look up which platform gets a lot of players; might even be a browser-based one.
-Develop your own world? Use existing tools for ease of use and support: https://itsfoss.com/create-interactive-fiction/
-Grow your skills? You can start from scratch learning programming, or you can start with simple games in an editor that has options to do a lot more, such as Game Maker.
soo... talk to us!
Thanks for the reply.
I was thinking of doing a turn-based puzzle thing, possible like Dragon Quest/Pokémon or even a Fire Emblem (I’m leaning more that way now, as it has that puzzle element that’s more a game to me).
Definitely cheap laptop territory. Also, this might be too ambitious for a first game (assuming this is your first). Do you have any computer science background or programming skills?
Oh no! Not a bit! But I might manage an FE1 grade of game.
Might do a SHMUP first once I figure out what coding language to learn.
A loooong long time ago when I was learning and trying all sorts of game type I had made a small 1 screen boss rush SHMUP in BASIC, the older kind with line numbers that you either retyped your code or saved it to an audio cassette.
The rest of my path was Logo -> QBasic -> direct x86 assembly -> C/C++ in DOS with interrupt calls to the hardware -> C++ OWL in Windows 3.1 -> C++ OpenGL -> C++ DirectX/HLSL -> C# XNA and finally Unreal Engine where I no longer type any code and just do everything with Blueprints. It's like chill mode for weathered programmers.
Use any language that works for you but I suggest aiming for the ones that have the easiest path to display stuff on screen and apply logic to things. At the beginning you want to be managing non gaming/system stuff as little as possible.
I started my current run in 2016, and I'm still learning things. Early on, I was most busy with figuring out how to make things move and interact, and that was cool and rewarding. But those early successes delayed me learning the most useful things. I'd like to post a pseudocode 'quick guide' to those useful things that others can comment on.
A: learn Variable types:
-a bool is an off/on value,
-an int is an integral number (0 or 3 or 10555),
-a float is a number with a decimal point (0,125 or 456,74),
-a vector2 is a combination of two floats,
-a vector3 is a combination of three floats.
-a string is a text value ("name" or "error message")
B: learn to declare a class:
-ex. 'magic_type' contains an int 'identifier', a string 'name', and an int 'damage'.
-ex. 'character_class' contains a string "name", the bool 'isplayer', the vector2 'characterposition', the int 'health', and int 'magic_identifier' and the int 'power'.
C: learn to create lists:
-for example to 'magic_types_list':
--add new magic type (0, "water", 30)
--add new magic type (1, "fire", 50)
-for example to 'character_list':
--add new character_class ("paul", true, new vector2 (0,0), 100, 1, 5)
//this is a character named "paul", who is the player (positive bool), who is standing at the exact center point of the world, who has 100 life, uses fire magic, and has a power of 5.
--add new character_class ("enemy", false, new vector2 (0,5), 80, 0, 4)
//this is a character named "enemy", who is not a player, stands 5 squares away, has only 80 health, uses water magic, and has a power of 4.
do something!
if (kyboard.key.P is pressed)
{
.for (every character_class NowActing in the list 'character_list' ) do this: [go down the list of characters]
.{
..set string actorname = Nowacting.name;
..for (every magic_type magic in magic_types_list) do this:
...{
....if (magic.identifier == NowActing.magic) [set variables according to the magic the actor uses]
....{
....set int damage_amount to magic.damage times NowActing.power, divided by 10, rounded to an int;
[for example the player uses fire magic, with a damage of 50. multiplied by the power level, that's 250.]
[That's too much, so we divide by 10. However, that may end up with a float, and we're setting an integer. So we round the result before saving!]
....set string magicname = magic.name;
....}
..for ( every character_class Receiving in the list 'character_list') do this; [again go down the list of characters]
...{
...if (Receiving.listindex is not NowActing.listindex) [characters should not attack themselves]
....{
....set string receivername = Recieving.name
....print actorname + " hit " + receivername + " with " + magicname + " for " + damage_amount + "damage!"; [show the attack]
....set Receiving.health = Receiving.health - damage_amount; //work out the effects of the attack
....if (Receiving.health < 0) [if the receiver died]
.....{
.....print receivingname + "died";
.....NowActing.power = NowActing.power + 1;
.....print actorname + " leveled up!";
.....}
.....else [if the receiver did not die]
.....{
.....print receivingname + "was wounded";
.....}
...}
..}
.}
}
And that's the framework you can add a game onto, playing around as you see fit! An easy one is use keyboard input to move players around. Then, check if the difference between the positions is lower than a minimum distance. You can add 'vulnerabilities' to magic types, so a water user is impervious to fire, but weak against earth. etc. etc.
Posts
Created a massively multi player version of Lights Out
http://geometricgames.com/lotw/lightsouttogether.html
I made a game, it has penguins in it. It's pay what you like on Gumroad.
Currently Ebaying Nothing at all but I might do in the future.
I have a feeling that Blueprints are going to get very complicated very quickly. Some Spanish Brick will look much nicer than the ugly grass from before:
This spanish brick texture is at '0.0' for normal map flatness, let's crank the strength up to '50.0' and see what happens:
Q.) What kind of normal map do you want?
A.) Just fuck my shit up, fam
Let's try something more reasonable, like '5.0':
That's a lot better. We went from "PS2 game" to "Original X-Box game" to "bathroom tile grout at my first apartment".
The next part is dealing with creating Master Materials, and also importing static meshes. Guess what the tutorial mesh is?
Fucking CRAAAAAAAAAAAATES!
Needs... more... crates...
Old man Murray eat your heart out, I've got a StC of negative two years.
The other big thing was different textures for things like metallic and roughness maps, and how its fairly common to do multiple maps in the same texture and isolate them in individual RGB color channels so you can just import one item instead of three.
The next section is on lighting and Lumen, so I think I'm going to call it a night here.
For a while now I've sort of had a hangup about doing this for a living and instead wanted to be a solo dev making a passion project on my own terms - I'm not sure if I have the personality for it though. When I'm responsible for every aspect of the game I find myself always plagued with doubts, and I compare the various merits of my game unfavorably to those that are built by experienced teams. I think having some external responsibility attached to it would be good for me.
It's nice that if you have an idea and others get excited about it, it gives a bit of an oomph to push ahead. Then I hire someone if I need the help and who knows.
I can't imagine working full time coding, or ever working for a AAA. This is just fun...
I mean, I just learned about libcurl the other day and I am so in love I don't know what to do with it.
Atlantic - A remaster of Monopoly, with extra effort to "hide" that it's the exact same game - Held for architectural issues.
battleship - Teaching my nephew how to code games. - Held as he just moved into a new place.
blind - a 3d first person action adventure game with only sound. held because sound effects are hard. (Maybe hire a sound person?)
bsd - various dabbles into the BSD games and see if something remastered can come from them. (Open source project if something works) - Nothing struck me
Desert Bus 3D - someone else did it, but I did make a cool bus in blender!
Gaia - remake of sim earth (Open source) - currently active
GalaxyNGX - update interface to Galaxy NG PBE 4x game, just found libcurl... very tempting
Harmony - Attempt to glue Visual Novel extensions to Inform 7.. Graphics lib was converted to Hikari
Hikari - Game to teach Japanese though visual novel/interactive fiction - 6th attempt at "Pineapple Chronicles", but it's not really a Pineapple game so this attempt doesn't count. - currently in work.
Inside - A remake of Laser Surgeon: The Microscopic Mission - I have some wormhole code that I need to convert from allegro 4 to allegro 5
Pineapple Chronicles - Pineappple is dead! Long live Pineapple. (See Hikari)
RPG Shell - Wonder your filesystem like it's a 2D JRPG? Ehh. I had ideas.....
The Guy In the Chair - a game where you played the guy in the chair guiding a Superhero in a city. Text/Unicode/Terminal based - tried to make a Simcity 2000 city loader (For the city the hero would be in) by converting some c++ code I found to C. Failed. I should have just wrote my own loader.
Tokaido - A walking sim / treadmill VR game where you walked all 315 miles of the Tokaido Road in Japan cira 1850... Turns out 315 miles is a long walk...
Whatcha guys got on your back burner? Want me to elaborate more on my projects?
Magicario: Magic-based RPG battler, my current project. Considering some major design changes at the moment. https://zek23.itch.io/magicario
Goblio: Incremental game about spawning goblins to kill villagers. Might go back to this some day. https://zek23.itch.io/goblio
Final Fantasy clone: An earlier attempt at an RPG battler from the side perspective. I was sort of happy with this one but didn't take it very far anyway.
Platformer shooter: Short-lived attempt to make a platformer, didn't really have a strong direction for this one.
Zelda clone: Kind of a Binding of Isaac type thing. My first real game attempt. This one turned out okay but I lost interest.
3 years ago I switched jobs from logistics desktop/web software to flight simulators used in flight training. So I'm still not a gamedev even though I do the exact same things but without the stress and weird shipping deadlines.
I remember I was having some problems figuring out how pointers worked with structs while deciphering the C++ code template I had to load a SimCity 2000 file. I was just going through what I had done, and WOW it was obvious I had no idea what I was doing. (I have since taken some online classes in C, especially pointers and structs), and there was all kinds of nonsense I obviously was just guessing at to get to work.
First of all, you can't pass by reference in C and that & in the "function" declarations was just kicking my butt. (PROTIP: you pass the pointer). There was junk cut/paste all over the place with no reason what so ever, and I had an overabundance of structs that I ripped from classes not knowing why they were built that way. I had a fun time turning everything into actual C code.
So it's not a new years resolution, but I promised an hour a day of coding... Doesn't matter what. I pick one of my dumb projects and work on it for a bit and see what happens. My GitHub commits look like this now
I never really did anything with GitHub, but now if I have worked on it, and it compiles cleanly, it's getting a repo. I kind of like how that looks.
== EDIT ==
Oh, and also, I was also dead set on using Code::blocks to code as I'm super cozy with GCC, but now... honestly, I'll fire up Visual Studio and just go. I actually try and make all my code cross-platfrom so plopping it into VS has been super easy, barely in inconvenience, and now that the newer versions have dropped TFS for a direct GitHub access. It's a dream to just make a new project, drop in some C files, and get going...
I really... REALLY, wish there were some C templates in there so I didn't have to rename all of new files I create. (And to add templates is a bit of black magic itself.)
Speaking of which...
WHY ON CARL SAGAN'S PALE BLUE DOT IS IT SO HARD TO MAKE A SNIPPIT!
Still plugging away at Nimsters. This week was mostly focused on the map screen, which is more like a Darkest Dungeon-style collection of rooms than the map screen in Pokemon. The rooms can be defined with a number of properties, such as:
- Which directions you can exit the room; this allows you to create one-way corridors or simply have adjacent rooms with a wall between them.
- A trainer present in the room, who will challenge you to a battle when you step into the room until defeated, at which point they may be optionally re-challenged (though for diminishing monetary rewards each time until you go pick on someone else).
- An event that automatically triggers when you step into the room. (This is basically just an incredibly simple scripting block; my test event increases a counter every time you enter the room and eventually triggers a selectable event that questions why you're walking in circles.)
- A set of events that can be presented as options within the room. (i.e. "1. Read the important-looking poster on the wooden sign.")
- A terrain type that affects how the room is rendered and can present optional requirements to enter it, such as having a Nimster of a particular type and level. (I sort of designed and implemented this without really considering whether it really improves the game; the idea is to mimic something like the HM system. But it's there now, so I can always just not define any room requirements if it's too annoying.)
- An encounter table that indicates how often and what kinds of wild Nimsters will attack when you enter that room.
The next item on my checklist is one of the classic banes of RPG programmers: inventory items. I'm working out how I want to represent them in-code and let the player manipulate them in the UI.
Right now I'm sitting around 60 ideas for monsters. Almost entirely out of just smashing a noun and an animal together. Like... "Fossilot"! The skeletal cat Nimster!
UE4 style lighting. Old, harsh, shit from a butt:
UE5 lighting with Lumen. New, fresh, global illumination in realtime without using raytracing and melting your GPU into slag:
The bounced lighting giving the sphere a hint on green on the left and red on the right is pretty neat.
If we nuke the light inside the cube, but add a sky atmosphere and sky light, the cube gets a little bit of light from the sky.
But if I add a bigass bright white wall
than I can reflect a ton of sunlight into the cube and light it up that way as well. Neat.
Next was setting up and going over the differences with archvis lighting in a pre-made room.
The next step was to change all of the lighting to static and then bake it. You have options for setting the quality of the lighting, going from Preview to Medium to High to Production. The dude in the tutorial set his to 'Medium', but I figured I'd see what Production looked like. What's the worst that could happen?
12 minutes later I got this:
Much nicer. So it makes sense that you would want to use Lumen in a situation where you were going to have dynamic lighting like a day/night cycle or you wanted something cool like a cyberpunk alley with tons of neon lights or whatever, with the caveat that it is more computationally intensive when being run. On the other hand, you can use static lighting in an environment and get higher quality lighting that is less resource intensive, but with the caveat that your lights and shadows are baked in and can't be altered.
The next section will be on how to utilize the landscaping tool, which looks like a lot of fun.
Also Zek your videos are showing as private for me.
Is this something I can do totally ad hoc "under the table" or do I need to form an LLC or something? I don't have a clue about legal/tax implications etc. I don't think I'd be paying full time salaries, probably just finding people at an hourly rate on Reddit or something.
So, I don't make a living off of this, but I do make some small amount of money and spend a larger amount of money. Assuming you are American: To get started, you don't have to do anything special business-wise, you can work as a sole proprietor, and this is fine for working off of Patreon or before you have any actual money coming in. Freelancers can be paid as contractors, but if you pay another American more than $600 in a year you will have to get their tax info and send them a 1099-NEC the next January (I literally have some sitting in envelopes here right now). Also: absolutely find a Work For Hire contract template online and get them to sign it.
Once you get to Steam and other storefronts you probably will want a business bank account and more legal protection, so it's LLC time. Here in Virginia I was able to form a single-person LLC, which is an LLC except my taxes work no differently than as a sole proprietors. As soon as you add a single other full employee though it gets more complicated and that's beyond my experience.
It's impossible to bring in money under the table, every site that handles money will report your income, you could pay people under the table but that means you don't get to write off the expense on your taxes so you're costing yourself money at that point.
First, I've set up the 'battlefield', which has the player's bridge stations set up on one side. Each bridge station may have up to four different actions they can perform, in my Tutorial Mode ship here they all only have a single option, but I will have other encounter types coming down the pipe where there will be multiple action options right from the start.
Each station has a number of turns they require to 'charge up', and turns are consumed by issuing Commands. The 'deck' of commands available to you is made up by the traits of your currently assigned bridge crew, from which you draw a random hand. These commands will allow you to do things like power up actions before they trigger, but there's also trait effects like debuffing the enemy or adding a chance for random resource rewards and things like that. My next task will be actually applying the effects of commands once selected.
Thanks that's really helpful - how do the tax write-offs work? Do I need to be running an LLC and/or intending to sell a product for my payments to freelancers to be a business expense?
No, as a sole proprietor you fill out the taxes under your own name/SSN. Schedule 1 for adjustments to income and C for contract labor and other expenses. If you really never intended to make any money from it and you're just writing off expenses on a hobby I imagine that could get you in trouble, but I'd expect it'd require years of returns and a lot of money before it triggered an audit (just a guess, not an expert).
Unrelated, like Handkor mentioned with flight simulators, my day job involves making training software in Unity. It doesn't have the 'fun' aspect of game design to it, but then programming games isn't the same as playing them anyways. Stuff like that is a good way to get experience with gamedev tools without going into the hellscape that seems to be game dev companies.
A: re: riding the employee/independent divide, in my country one can be both and simply declare the income. Crucially, one should become a part-time independent while still employed, which means that if you are fired from your daytime job you will get unemployment benefits *and* remain part-time independent.
However, I kind of agree with those people who say that for the part-time developer gig to become a paying job you *just* need to jump in as a full time indie. If you are an MBA with wicked networking skills and your idea of chilling is community managing while simultaneously haggling with peeps on fiver, you could accidentally release a game successfully without actually having developed a game.
If art, programming, game design, writing, QA or a subset thereof are what you do naturally, you can certainly develop a game, but you'll need the above MBA skills to actually launch it and make money of it. If those MBA skills are not part of how you approach this hobby naturally, having to do them to pay rent is a good kick in the pants if there's no-one else around to deliver it.
B: for my upcoming games, I'm going to ask around for someone who is interested in applying their MBA/community/marketing skillset to a game and a game designer/QA person. I've moved along my 'reading up on' skill curve to the point where people are repeatedly saying 'you need more than one founder' for startups. Related post in a second.
C: as a manager in Tech Support (YMMV), what I like in employees is accountability. I don't care if you make mistakes, but I need you to own up to them, be able to listen to me explain what went wrong and what you should have done, and I need to not have to repeat that four times. I have (had) people with less than 12 years of experience who have joined with that attitude, and they were a god-send combined with the young kids, whom I can assign the schedules that are not possible for the 12 year veterans that have to pick up their kids from school.
I intended to have a synergy between game and tv series proposal on the marketing side and there has been, but not a LOT.
Recently he got a contact with someone who produces Unity games and wants to look at the vertical slice
Because of my work/family life/development balance, I was intending to do a veeery soft release of the vertical slice as a demo. I don't have anything remotely resembling a pitch. I don't have a good trailer. I don't have a business plan on how to get to fully-fledged game from vertical slice.
Help?
I could use some good instructionals on how to create a pitch deck, as well as a business plan. I do have a more elaborate game pitch that I got from a friendly developer and I could work off, but it's all hands on deck here as my partner wants to talk to the producer *soooon*.
Mage: The Ascension would have to be limited to like-- Sphere 2 tops because whoooo boy once you hit 3 in a Sphere shit gets off the hook in most cases. ESPECIALLY with certain combinations.
Don't get me wrong I would buy the SHIT out of it even if it's hyper limited but properly translating the sheer versatility of Sphere Magic is not really doable in a non-dynamic fashion. Too many possibilities.
Spoke to a friend who is also a game developer, albeit much more accomplished and running a company. Basically, the problem with my current game design is that it's not designed from the ground up to be free to play with monetization options. I.e. I can pitch all I like, but no publisher would take it. He just pitched 30 publishers so I assume that's indeed the state of the art.
Monetization options are; IMHO:
-hats: updating your character or fighter with cosmetic changes. Doable, but I'd need to revamp the UV mapping on my fighter jet and figure out a way to have high resolution masks work with it. I can picture using meshes in a back room and a camera to snap then save to disk? Or, I use a purely script based texture generator- the problem is shadergraph doesn't seem to do geometry very well. That's about a month or two for just the fighter, without adding characters, animation, textures for the characters and hats in.
-weapons: buying in-game content with functional advantages. Difficult to work out what's needed here. It's a single-player game, with very simple mechanics. Normally, the techtree will be extra storage for (or smaller) 'drop nodes' and better fighter jets. I can imagine selling assistive technology, autopilot etc. but then the game might get too easy as there's only so much I can do to increase difficulty. (That's a design issue.) Main issue here is that it would need more content added to create extra difficulty, for example smart generation of obstacles. For the content on sale, placeholders would likely do for the pitch.
-nagging: get the game to a point where it's fun enough that players will play even if I add recharge countdowns or adverts, but will pay to skip them. It feels counterintuitive to polish the game just to then introduce such crippling. Not the person I want to be, really, even if it is the person that gets to pay the rent.
But this is starting to get into the Live Ops world of post release content. Daily/Weekly quests. New characters. Seasonal events.
I was just the other day explaining why, 20 years ago, my web developer friend did offer content updates to his clients, but at steep rates that got steeper the bigger the update. You want to spend your time building new websites (he rewrote a CRM yearly for fun) and not upkeeping websites in a stressful pingpong loop with a client while losing programming skills.
In game dev, afaik, continuous updates lead publishers and developers to drop talent and invest in juniors who get little freedom as data is driving 'innovation'. Because the numbers show what's popular so any investment can go to sure things. Who like sure things? Investors like sure things. Ask Hollywood.
We may all think we like a sure thing, but that's my day job, not the space I want to creatively explore.
I could certainly use fiver to have new content- as long as it paid the bills- and even apply myself to big additional mechanics updates. But the latter always risk breaking the game or not making a splash. I can take that same risk on a new game entirely and have more fun developing it.
This static mesh has Nanite disabled
Because of that it has a typical static mesh wireframe
But if I enable Nanite for this particular static mesh
than is no longer has a typical wireframe because the number of triangles is changing dynamically based on the distance from the camera, voodoo magic. The effect is significantly more impressive when importing Nanite quality meshes from Quixel instead of doofball ones like this, but still.
"Playtesting":
Using Unreal's built in third person package and the generic model I can adjust the settings for the camera. From here I can enter the environment I built as if it was a third person videogame by using the Play feature...
... which spawned the character beneath the level, falling swiftly into the void. That's no good!
But by adding a playerstart actor...
... I can dictate where the playable character spawns. Cool beans.
Maeking an actual gaem:
It's an obstacle course. Here's the layout. The player spawns at one end, and the statue is at the other. The cylinders rotate at different speeds, each one is attached to a blueprint that, if the player character collides with it, will ragdoll him (effectively player death). Touch the statue at the end and you win
anakinskywalkeritsworkingitsworking.avi
Collision with the cylinders ragdolls the player using Set Simulate Physics
Touching the statue creates text in the viewport and acts as a win condition. Technically it's a videogame. The worst videogame ever, but still, a videogame. Ship it!
I have now officially created more bug free videogames than Derek Smart.
As for what is happening in the screenshot, this little droid wants to be pretty. So he runs around wearing wigs stealing makeup out of purses only to run back to his little hidden corner to a mirror where you can paint his face with the makeup.
Yeah, it would have to be super limited, but I'd love to think of a way, because I love the lore and worldbuilding in Mage(Awakening) to pieces.
Is my best bet simply to buy a cheap laptop? iOS hates downloading and running anything that isn’t an app.
If so, depends on what you're trying to achieve.
-Getting lots of feedback from players? Look up which platform gets a lot of players; might even be a browser-based one.
-Develop your own world? Use existing tools for ease of use and support: https://itsfoss.com/create-interactive-fiction/
-Grow your skills? You can start from scratch learning programming, or you can start with simple games in an editor that has options to do a lot more, such as Game Maker.
soo... talk to us!
I skipped the report last week because it was AGDQ week, and while I was poking at a little side-project, I didn't have enough completed to be worth typing up a report. And, for that matter, I was too invested in watching AGDQ to bother typing anything up on Friday. Six of one, half-dozen of the other.
My latest project was more a matter of personal curiosity than anything truly useful; nothing new in that respect. Though I've used some ready-made randomization methods in my engine in the past, I wanted to explore alternatives and/or just build my own. The practical upshot of that work is that each individual entity that wants to use randomization carries its own randomizer state around with it, and there's a central "randomization seeds" component which exists solely to issue seeds to new entities' randomizers as they're created. Though the method I use is unorthodox (using addition and treating the randomizer state as a stream of bits instead of multiplication, bit-shifts, and XORs and consuming the entire word at once), the way the seed generation and the individual randomizers combine produces sufficiently differentiated random results with a distribution that resembles the ones produced by the Xoroshiro+ and Mersenne Twister PRNGs that I'm happy with the results.
On the Nimsters front, my Git log is showing that the majority of my work this week was on inventory items, as I mentioned last time. In Nimsters, there are four different types of items, all of whose effects are now implemented in the game:
- Restoratives: One-time use items that can restore health, stamina, or status effects, alone or in combination.
- Accessories: The Nimsters equivalent of held items from Pokemon; equipped to a Nimster to produce a passive effect of some kind. No accessory currently produces an actual effect, but hey! They can be equipped, which is all the requirement doc called for.
- Treats: Have two different functions; used on a wild Nimster, it restores its health and adds to its "respect" value, which makes it more like that Taming nodes will show up on the timeline that allow you to tame the Nimster and add it to your party if you land on them. Used on a Nimster in your party between battle, they function like Rare Candies and their derivatives, granting variable amounts of EXP.
- Supplements: The equivalent of Pokemon's Protein/Carbos/Iron/Zinc items, except that they modify the Nimsters equivalent of IV instead of EVs.
I also implemented the wild Nimster encounter tables, so you can be attacked while wandering around the map. The next feature on the roadmap is actually getting the taming mechanics working, so Nimsters can be a proper monster-collection game at last.
So naturally I also decide to move all the UI bits from C++ to lua (currently main menu is lua but in-game is C++) which as a bonus makes things Moddable and enforces Separation of Concerns, but it's the main game that needs the special shaders and I can't just accept whole shader source straight from lua oh no that won't do at all
Now I'm sitting in my "engine" project defining a complicated shader generation model that can have 8 textures of varying types and accept arbitrarily formatted vertex streams. Why? Nobody else will ever use this, even if I put it up on github. I don't even need it, if I need lighting and shadows and fancy materials and all that I can just go fire up unreal engine. I think I am actually incapable of scoping any personal project appropriately for a solo developer
So for that purpose, a lookup table does the trick!
Oh, it gets worse in my case. I went through a couple drafts of the original post and decided that the internal workings of my new randomizer weren't interesting enough, but the exact definition of "good enough for my purposes" here is kind of funny, so here it is:
Because of the way my game's engine is structured (farming operations out to worker threads to be performed in parallel, but with an expectation that re-runs of the game with the same input will produce the same final state), I can't use a single source of randomness for all random events in my game. Every Entity has to have a Component which stores the state of its randomizer so they can update independently. Object creations are single-threaded, so ultimately what happens is I have a "randomization seeds generator" that increments by 1 every time you get a new seed, and a "randomizer" Component which holds the state.
My previous randomization implementation used the well-known Xorshift algorithm. But then I said to myself, "you know, really what's sufficient for my purposes is some algorithmic walk of the bit-space of a 64-bit word". It turns out that if you add a number with no 2s in its prime factorization something that fits in a power-of-2 data field (like an ulong in C#), it'll walk the entire available bit space. So I put together a small program to find a number which is a multiple of many prime numbers other than 2 which is just over half the bit space (reasoning: a large number of bits will change each time) (result: 3 * 5 * 7 * 11 * 13 * 17 * 19 * 23 * 29 * 31 * 37 * 41 * 43 * 1423 = 9308384687483226345) and added that number to the previous state after consuming all the bits in the previous state. After fooling around with it, I found out that this produces distributions that very often subtly favor 0 for reasons I wasn't able to determine analytically, and also have jagged distributions favoring some numbers above others.
That sent me back to looking at Xorshift, and when I looked it up on Wikipedia, I found that it had a period of 2^64 - 1. I went and read the paper where the genius who's been coming up with half the random number generators in Wikipedia's list of PRNGs originally published Xorshift, and one of the things in there was a proof that, so long as the seed state isn't 0, Xorshift performs a full walk of all the entire set of bit patterns in a 64-bit word. In other words, the thing I'd already used had the attributes I was already looking for (minus one value, which is no real loss).
Either that, or like me, you've got the problem where the coding is the fun part. I struggle with art and design, but code comes easily and gives me that "you accomplished something!" feeling when something compiles and does what it's supposed to. Thankfully I've well and truly run out of projects for my engine, so it's kind of forcing me into the design space, but it might help to reflect on your motives.
As I sit here designing a cache-based system to draw text because I can't now assume that everything lies in a predetermined font I can blit at load time...
Do I draw the text to an intermediate buffer? Try to force glyphs into the same cache textures? How do I handle invalidation vs fast re-render? So many choices
Thanks for the reply.
I was thinking of doing a turn-based puzzle thing, possible like Dragon Quest/Pokémon or even a Fire Emblem (I’m leaning more that way now, as it has that puzzle element that’s more a game to me).
Definitely cheap laptop territory. Also, this might be too ambitious for a first game (assuming this is your first). Do you have any computer science background or programming skills?
Oh no! Not a bit! But I might manage an FE1 grade of game.
Might do a SHMUP first once I figure out what coding language to learn.
A loooong long time ago when I was learning and trying all sorts of game type I had made a small 1 screen boss rush SHMUP in BASIC, the older kind with line numbers that you either retyped your code or saved it to an audio cassette.
The rest of my path was Logo -> QBasic -> direct x86 assembly -> C/C++ in DOS with interrupt calls to the hardware -> C++ OWL in Windows 3.1 -> C++ OpenGL -> C++ DirectX/HLSL -> C# XNA and finally Unreal Engine where I no longer type any code and just do everything with Blueprints. It's like chill mode for weathered programmers.
Use any language that works for you but I suggest aiming for the ones that have the easiest path to display stuff on screen and apply logic to things. At the beginning you want to be managing non gaming/system stuff as little as possible.
I started my current run in 2016, and I'm still learning things. Early on, I was most busy with figuring out how to make things move and interact, and that was cool and rewarding. But those early successes delayed me learning the most useful things. I'd like to post a pseudocode 'quick guide' to those useful things that others can comment on.
A: learn Variable types:
-a bool is an off/on value,
-an int is an integral number (0 or 3 or 10555),
-a float is a number with a decimal point (0,125 or 456,74),
-a vector2 is a combination of two floats,
-a vector3 is a combination of three floats.
-a string is a text value ("name" or "error message")
B: learn to declare a class:
-ex. 'magic_type' contains an int 'identifier', a string 'name', and an int 'damage'.
-ex. 'character_class' contains a string "name", the bool 'isplayer', the vector2 'characterposition', the int 'health', and int 'magic_identifier' and the int 'power'.
C: learn to create lists:
-for example to 'magic_types_list':
--add new magic type (0, "water", 30)
--add new magic type (1, "fire", 50)
-for example to 'character_list':
--add new character_class ("paul", true, new vector2 (0,0), 100, 1, 5)
//this is a character named "paul", who is the player (positive bool), who is standing at the exact center point of the world, who has 100 life, uses fire magic, and has a power of 5.
--add new character_class ("enemy", false, new vector2 (0,5), 80, 0, 4)
//this is a character named "enemy", who is not a player, stands 5 squares away, has only 80 health, uses water magic, and has a power of 4.
if (kyboard.key.P is pressed)
{
.for (every character_class NowActing in the list 'character_list' ) do this: [go down the list of characters]
.{
..set string actorname = Nowacting.name;
..for (every magic_type magic in magic_types_list) do this:
...{
....if (magic.identifier == NowActing.magic) [set variables according to the magic the actor uses]
....{
....set int damage_amount to magic.damage times NowActing.power, divided by 10, rounded to an int;
[for example the player uses fire magic, with a damage of 50. multiplied by the power level, that's 250.]
[That's too much, so we divide by 10. However, that may end up with a float, and we're setting an integer. So we round the result before saving!]
....set string magicname = magic.name;
....}
..for ( every character_class Receiving in the list 'character_list') do this; [again go down the list of characters]
...{
...if (Receiving.listindex is not NowActing.listindex) [characters should not attack themselves]
....{
....set string receivername = Recieving.name
....print actorname + " hit " + receivername + " with " + magicname + " for " + damage_amount + "damage!"; [show the attack]
....set Receiving.health = Receiving.health - damage_amount; //work out the effects of the attack
....if (Receiving.health < 0) [if the receiver died]
.....{
.....print receivingname + "died";
.....NowActing.power = NowActing.power + 1;
.....print actorname + " leveled up!";
.....}
.....else [if the receiver did not die]
.....{
.....print receivingname + "was wounded";
.....}
...}
..}
.}
}
And that's the framework you can add a game onto, playing around as you see fit! An easy one is use keyboard input to move players around. Then, check if the difference between the positions is lower than a minimum distance. You can add 'vulnerabilities' to magic types, so a water user is impervious to fire, but weak against earth. etc. etc.