/v/ - Video Games

here's_to_you.mp4
(405.6KB, 640x640, 00:12)

>>6917 (OP) 
Nice fucking markup, admin.

1599907208558.jpg
(68.1KB, 640x640)

>Has anyone started making their own game during this time
I was going to make an /agdg/ thread here like a month ago but I felt like I should have progress to post first. You made the thread so 1+1=you know.

not_progress.png
(11.3KB, 681x402)

That said I haven't done nothing. I learned to render bezier curves, and partially how to read font files.

I find bezier curves are very interesting. It seems like some kind of magic but it's actually a very simple concept. You just connect points along all the "control lines" at some position, and you will end up with a position along the curve. Wikipedia has excellent animations that demonstrate it. https://en.wikipedia.org/wiki/B%C3%A9zier_curve

index_2.jpeg
(39.6KB, 376x445)

> Has anyone started making their own game during this time?
Does text games count?

>>7033 (checked)
Do you want to get into the semantics of what constitutes a video game?

>>7036
You control a character's movements and fight enemies until you win.

tired2.jpg
(77.4KB, 534x540)

>>7036
The term "video game" technically includes anything that's meant to be entertaining (or made to be entertaining by the user) and runs on any kind of computing device with a display. You could say Photoshop is a video game if you invented some game related to the buttons on it's UI.

Now the question is; if someone makes a "video game" that's such piece of shit that it doesn't have any kind of entertaining or attractive or interesting or educational qualities and is basically just a scam to waste your time and kill your brain cells, like Gone Home, is it still a video game?

6a0bed68498cc1656a9f6d1264ad3652c3ddc9e1b875ac4827bb29104e2e9097.gif
(947.1KB, 300x300)

>>7038
I was with you until you misspelled "its"

>>7038
>You could say Photoshop is a video game if you invented some game related to the buttons on it's UI.
> if someone makes a "video game" that's such piece of shit that it doesn't have any kind of entertaining or attractive or interesting or educational qualities and is basically just a scam to waste your time and kill your brain cells, like Gone Home, is it still a video game?
There is 0 room for philosophy in this subject. If a video game doesn't have a  2D, 2.5D or 3D environment where you shoot shit up or drive around in race cars or fly with planes, then it's not a fucking video game. 

Puzzleshit isn't vidya, textshit isn't vidya and visual novels aren't vidya. Doom is vidya, minecraft is vidya, flight simulator is vidya. 

Shit like TLOU, Uncharted and Heavy Rain are also vidya, but they're fucking trash vidya that only morons enjoy. 

That is all there is to it.

0b8a156898a41c60d5b7bc2ffee331dd82a4544ed3d2068aafb1cdd337b27a71.png
(150.6KB, 236x292)

>>7040
So in your massively flawed logic tetris isn't a video game? You are insane.

>>7041
Not anymore

>>7040
>textshit isn't vidya
Text based adventure games were one of the first video games created. See: Zork
Then came Sierra and polished the concept. See: early King's Quest.
Defining video game isn't easy. Ideally, a video game has all of the following: lose state, win state and at least some kind of challenge between the two. There are exceptions, depending on genres and game modes. For example, Sim City has challenge, but the freeform mode doesn't have a win state as the game just goes on. An adventure game may not have a lose state (Monkey Island 1) but it certainly has a win state.
Hope this clarifies things for you!

Ukraine_kit.png
(423.8KB, 544x572)

>>7036
>>7038
Shit m8 I'm still to much of a brainlet to figure out 2D/3D engines, hell I tried even tkinter and I made such a abhorrent confusing mess out of it that I don't know even anymore where to start. 

>Now the question is; if someone makes a "video game" that's such piece of shit that it doesn't have any kind of entertaining or attractive or interesting or educational qualities and is basically just a scam to waste your time and kill your brain cells, like Gone Home, is it still a video game?
No. Well to be honest when I look at my current program it is indeed awfully basic at its current stage, despite I wasted several months on this project already.

>>7067
>tkinter
Sounds like you're not using a premade game engine. What are you trying to do and why isn't it working out?

yeltsin.png
(38.1KB, 987x836)

19194fad72af767861a419c47d21f22c47f56f94d41c0b33649817e1e120d0ab.png
(458.6KB, 719x718)

>>7076
My first attempt of using tkinter was the idea of trying to make a rpg-ish game, I figured out how to get buttons to display, some numbers, a picture and other minor details. As soon I tried to implement the notification box which displays what action got executed or other events it didn't worked out, I couldn't figure out so that new text string gets appended and the older strings get discarded, I spend around 2-3 weeks figuring out of what went wrong and I found no solution for it so after 2 months working on this program I gave up. 
The codebase is quite messy as its a mixture of OOP and ECS so I guess you could imagine how that one went out, as I don't grasp how a GUI gets structured via ECS instead of OOP, so I opted to make the GUI part in OOP pattern instead.

>>7084
You say you can't figure out an engine, but you're worrying yourself over development paradigm memes like ECS v OOP? You might be putting the cart before the horse a bit.

>>7084
>OOP and ECS
By ECS (entity component system) do you mean DOD (data-oriented design)? ECS is a meme, it adds a lot of architectural baggage that you don't want to be dealing with when you're new to gamedev or possibly at all, and it doesn't really solve any problem that you're going to have unless you're planning to make a really big and complicated game. Nobody cared about ECS until Unity said it'll make Unity faster.

>I couldn't figure out so that new text string gets appended and the older strings get discarded, I spend around 2-3 weeks figuring out of what went wrong and I found no solution
I have hard time picturing how that could happen except maybe you were trying too hard to make the UI system do it instead of doing it yourself, you were overthinking something, or you skipped on your programming tutorials. You just have an array of strings, add a string to the start and remove one from the end (or vice versa). Then loop through the array and print all the strings one by one, or combine all the strings together into one big string (+ a new line in-between them) and then tell the UI message box to replace all the text with that. Of course if you want more fancy lines of text with various colors and such, it gets a little harder, but I have no idea how the UI system you used works.

>>7182
Well I already wrote my program by using classes with a few functions defined here and there. Well then with what should I stick with, by using functions only? 

>>7198
>By ECS (entity component system) do you mean DOD (data-oriented design)? ECS is a meme, it adds a lot of architectural baggage that you don't want to be dealing with when you're new to gamedev or possibly at all
What kind of architectural baggage does it have? 
>and it doesn't really solve any problem that you're going to have unless you're planning to make a really big and complicated game.
I do have a plan for a complicated game but before I want to do that I would like to a couple of much smaller and simpler games first such as a resource harvester game, a coastal defense ones and a few more. 

>I have hard time picturing how that could happen except maybe you were trying too hard to make the UI system do it instead of doing it yourself, you were overthinking something, or you skipped on your programming tutorials.
well I used the delete(0.0, variable) and insert(0.0, variable) functions that is defined by tkinter and it doesn't display the text properly, I need to fiddle around with it again because I forget how it behaved back then.

>>7228
>What kind of architectural baggage does it have?
Unless I'm mistaken, the whole idea behind ECS is to break up all of your entities into multiple parts so you can for example very quickly modify everyone's position and utilize SIMD, and then pull properties from many different arrays depending on what you need at any given time. That's a lot more complicated than just having an array of entity structs for example and passing around a pointer to the entity to whoever you want to deal with it.

I mean if you understand it like the back of your hand and don't even have to think about it, then it's no big deal, but I assumed that's not the case. Otherwise it's just going to increase the amount of things you have to think about when you're trying to solve some problem, which is not what you want when you haven't even got things moving on the screen yet. Also UI can be complicated as is, so trying to use some kind of ECS to build one seems like a recipe for pain.

>I used the .. functions that is defined by tkinter and it doesn't display the text properly
So the problem was the GUI system seemingly not working. As someone who has lost motivation many times by trying to use someone else's shit and struggling to get it to work the way I want (or at all), all I can say is good luck. I'm currently re-learning OpenGL, hopefully getting a better understanding of it than I did in the past.

>>7236
>the whole idea behind ECS is to break up all of your entities into multiple parts so you can for example very quickly modify everyone's position and utilize SIMD, and then pull properties from many different arrays depending on what you need at any given time
From what I have read of the ECS proponents are that utilizing ECS would be much more flexible and in the long run easier to maintain than OOP, giving examples only a single loop function is needed to iterate over all the entities that exist. Entities also can even have more or less attributes added to them without the need of grouping the Entities to a particular group, as defining them a variable or attributes would be enough. The author also has given examples that creating a "Evil Tree" entity there is less concern to what group it belong, as long a function can classify it as "enemy/destructible". Unlike when it would be done as a object form which does need grouping. I have nothing read along the lines of SIMD apart that with the transition to ECS there will be less cache misses, as it would be closer how a CPU handles its data. 
Actually scratch the above, looks like I feel for the meme "OOP bad, ECS good" too, fugg. I just found a nice post that describe those things more in depth here: https://www.gamedev.net/blogs/entry/2265481-oop-is-dead-long-live-oop/ https://archive.vn/dmrEe 

> As someone who has lost motivation many times by trying to use someone else's shit and struggling to get it to work the way I want (or at all), all I can say is good luck. I'm currently re-learning OpenGL, hopefully getting a better understanding of it than I did in the past.
Thanks. I have myself hard time keeping my motivation up as well, I do try to make some progress everyday but most of the time I end up having no idea what function/object needs to be changed.

7e793a00ce1cd3c434583d06b4060573f6088886eb8b5f4922826a2479ff1683.jpg
(51.7KB, 600x338)

>>7247
Well if it's ECS vs OOP then I'm pretty sure the former definitely comes out on top, though I can imagine it being very easy to do ECS wrong and not actually get the benefit it was made to achieve (or to do OOP in moderation and not suffer the worst of it's problems). It's not so much that ECS is a bad thing, it's more that you're trying to plant a tree sapling and decided to use an excavator instead of a shovel to dig the hole.

The real meme is the whole obsession with paradigms. Both OOP and ECS have things that are useful, rather than thinking about what programming paradigm you should shoehorn everything into, just do the thing that's useful for the problem you have (and that doesn't completely fuck up your CPU cache).

>"Evil Tree" entity there is less concern to what group it belong, as long a function can classify it as "enemy/destructible"
Like a bit flag? I don't know if Python (I assume that's the language you use) lets you do something so basic, but consider this:
if (entity.flags & ENTITY_IS_ENEMY) && (entity.flags & ENTITY_IS_DESTRUCTIBLE) {
    // entity is a destructible enemy
}The above flags take 2 bits (bits, not bytes) of memory. If that's the example someone gave about the power of ECS, you should reconsider the value of the words you read. This "adding arbitrary properties to objects" thing sounds like a very hypothetical scenario to me, what's the actual real world use for the flexibility that lets you add an "evil_tree_root_length" property into a bat entity? That flexibility is useful for engines like Unity though because the developer of the engine has absolutely no idea what the user is going to do with entities (and it's literally their job to all the most advanced complicated tricks to the engine).

I can think of something slightly more useful though, let's say enemies can carry backpacks full of items. You could have an array that has all the backpacks of the world, and an ID of the entity that's holding it. That doesn't sound very "ECS" to me, but the general idea is the same I think. I guess the distinction is how far you go with the concept, I don't take any issue with this backpack example, just the idea that you're splitting your entities into a million pieces when you don't need to.

>I have nothing read along the lines of SIMD apart that with the transition to ECS there will be less cache misses, as it would be closer how a CPU handles its data.
Compared to what? It's true that ECS is good for cache, but so is what I described above. I don't know how much you already know about this whole concept, but basically the most important thing is that the CPU doesn't have to jump to a new location in memory often, and can instead shift next to where it just was because the nearby data was probably already loaded into cache. Problem with languages like Python is that it's harder to tell if the data is actually where you think it is. If you put X into an array, can you tell if X is actually in the array, or just a pointer to X? It's not much use if the pointers are next to each other in memory but the actual destinations are not (which is what happens when you allocate objects in bad OOP).

SIMD just means the CPU operates on multiple values (usually 4) in one go instead of each one separately. It sounds cool but it's kind of annoying to do.

Pic related is my face when I've been staring at this post like half an hour trying to figure out how not to write such long walls of text. I can't remove that thing about tree roots, what if someone reads it and gets something useful from it!?

Glad to see you fags here.
Gospeed and good devving, lads.

>>7254
>The real meme is the whole obsession with paradigms. Both OOP and ECS have things that are useful, rather than thinking about what programming paradigm you should shoehorn everything into
Yeah I agree, which makes me wonder why there is so many new articles about ECS and a large amount of it mentions that its a much better alternative to OOP, which according to the other goy that addressed the OOP flaws that most of those ECS blog poster proponents have posted. I'm not too well versed in either paradigm nor did I used it extensively so I don't have strong opinion about it. 

> If you put X into an array, can you tell if X is actually in the array, or just a pointer to X? It's not much use if the pointers are next to each other in memory but the actual destinations are not (which is what happens when you allocate objects in bad OOP).
I honestly don't know how I can go about that with Python as the documentation of handling variables/arrays doesn't mention anything about pointers, Python does do cases when there is 2 variables (A = 10, B = A) then referring to B is the same as using A memory wise, though I don't think you meant that one. 

> you should reconsider the value of the words you read. This "adding arbitrary properties to objects" thing sounds like a very hypothetical scenario to me, what's the actual real world use for the flexibility that lets you add an "evil_tree_root_length" property into a bat entity?
Man, I tried to find that particular site that had this "Evil Tree" example but I cannot find it anymore. I think you raised a good point too that it's not really such a big advance. 

>I can think of something slightly more useful though, let's say enemies can carry backpacks full of items. You could have an array that has all the backpacks of the world, and an ID of the entity that's holding it.
When a backpack is placed in the world, and a enemy can carry a backpack then it would make more sense that the enemy has a inventory variable that contains the backpack full of items, thus the world part shouldn't concern itself if a enemy entity has a backpack or not it should only concern what entity exist and in what position. The way you described implies the array should have 2 dimensions one for the list of backpacks placed in the world and the second of the entity that owns this backpack which is placed in the world. 

>Compared to what? It's true that ECS is good for cache, but so is what I described above.
I guess compared to the amount of memory it would take when it would be done as a "traditional" OOP way.

>>7254
Good post.
I think everyone and anyone trying to write their first game really should be reading through other peoples' code for games that are finished and work. I would like to recommend the source code for Brogue to everyone here. It's a decent blueprint for writing a fairly complex game in C. It uses arrays, a few structs, and bitfields. 

https://github.com/tmewett/BrogueCE

>>7330
The paradigm obsession is a kind of occupational hobby or disease (I don't know which) with software engineers and their management. Think about Agile, Waterfall, structured programming, Scrum, and on and on and on. I don't know what it is about software development that encourages or selects for this behavior, but it's been going on for a very long time in a kind of ouroboric tail-chasing way. I'm not saying it's good or bad, and you get frameworks for things rarely in pretty much every discipline. I wish I understood it better, honestly, even though I have no professional connection with software development. I'm just kind of fascinated at what forces inside the industry drive that thinking.

>>7818
It really is a plague in programming. My guess would be that it's a result of programming still being badly understood. These "schools of thought" with their "gurus" and what not are very common in pseudoscientific fields.

>>7818
>>7976
When you work a hundred hours a week for assholes who don't understand shit but take all the credit (and profit) after bloating everything up, repeatedly changing your contract, hiring for diversity instead of skill, skipping bug testing and throwing away 90% of what your team made for the project you might feel pressured to find a better way of doing things. Since devs are psychologically abused and conditioned to be dependent on publishers, they try out different coding paradigms instead of murdering their bosses.

>>7818
This is standard middle-management behavior, just applied to software development.

Good software specs and good software developed to these specs could be done. However, management and customers "believe" the cost in time and money is far too high. So they latch on to any snake oil bullshit that promises to "solve the software development problem".

I've seen this in related field, so it is quite easy to recognize this arrogant and childish behavior in software dev.

>>7330
>why there is so many new articles about ECS
It's the latest programming fad so everyone must do ECS now, a year later it'll be something else and you should replace everything with that. They always start from a good place, ECS is good at what it was made for, but then people see it's good and think everything can be made good by applying the same method to everything. Same with OOP, except OOP was such a big fad that it basically defined modern programming and all new languages.

>documentation of handling variables/arrays doesn't mention anything about pointers
Yeah I don't think Python lets you know that pointers exist. There's a few reasons to believe that any object you put into an array in Python is actually just a pointer or reference of some kind. I'm not super well versed in Python, but in Javascript numbers are set by value and objects/arrays are set by reference.

>When a backpack is placed in the world, and a enemy can carry a backpack then it would make more sense that the enemy has a inventory variable that contains the backpack full of items, thus the world part shouldn't concern itself if a enemy entity has a backpack or not it should only concern what entity exist and in what position. The way you described implies the array should have 2 dimensions one for the list of backpacks placed in the world and the second of the entity that owns this backpack which is placed in the world. 
This is getting into implementation details that aren't very meaningful since we don't know what backpacks are supposed to do in the game, but the backpack itself could have a value that defines who's holding it. Or do the opposite and put the backpack ID onto the entity. Or you could just put the backpack itself onto the entity, though then an entity can't carry more than 1 backpack, and every bullet and door will have backpack related values always present. That is, unless you make a union (another concept that I don't think Python has) so only entities that can hold a backpack have backpack related values, and for door entities the same data will represent door related values etc.

Maybe none of this even makes sense in Python's terms since there's no concept of where the data is, only where the "variable" is. The goal is to group the data in some way so that it's as close together as possible, in as linear order as possible when you loop through things, and with as little unnecessary data as possible. How you connect things together and find the data is maybe of less importance.

Ironically it seems like people realized that not knowing the data is bad, since both Javascript and Python now have some kind of buffer objects that let you define chunks of data, and then get a structured "view" into it (for example reading a part of it as 16bit integers).

>>7380
>anyone trying to write their first game really should be reading through other peoples' code for games that are finished and work
Do you find that helpful? I hear people recommend this kind of thing often, but personally I find it very hard to learn by reading a big system someone else made. I can see it being useful when you already know how to do all the basics and you're just looking for different/better ways to structure things, but if you haven't made a game yet then I imagine it's too much get a grasp of when all the concepts about rendering and entities and scenes and sounds and everything come at you at once.

9b43ea5a88acf1b787454227f86db4e9e1b83b24b7fae70feedb74854864f15b.jpg
(89.1KB, 700x788)

>spend literally an entire hour debugging since opengl is just showing weird white/black triangles no matter what I do
>forgot to call glUseProgram

working_GUI.png
(21.8KB, 1029x796)

working_GUI_2.png
(35.9KB, 1033x796)

how_do_you_like_the_buttons.png
(107.5KB, 1014x450)

After around 3~ days of fiddling around with Tkinter I finally managed to have a much less messy codebase to work off from, I also managed to get tkSnack to work so that buttons play funky sounds when its pressed with the downside when a sound plays longer than 3 second or so (I haven't measured it exactly.) it's possible that a earrape can occur which happens with the varmmove.wav file ripped from Total Annihilation. 
So obviously I need to integrate the other remaining buttons with all the functions I have written previously, it should make managing several blocks/structure much easier now because then the user has overview of all the values now instead of having to press the "info" function manually all the damn time. I did skimmed the tksnack documentation if its possible to limit a sound playback to 1 or so which doesn't seem to be the case, hmm yeah I need to try a different approach and hope it works I guess about that. After all the functions are integrated and some more being written I might try to figure out if I can add some more custom graphics to it, like spinning turbines or something. 

>>8120
>It's the latest programming fad so everyone must do ECS now, a year later it'll be something else and you should replace everything with that. They always start from a good place, ECS is good at what it was made for, but then people see it's good and think everything can be made good by applying the same method to everything. Same with OOP, except OOP was such a big fad that it basically defined modern programming and all new languages.
Given how long it took for the ECS meme to take foot I think it might be at least another decade for that to happen, unless somebody managed to make a "brand new" breakthrough in programming paradigm I suppose, which could start the whole debacle from anew again.

>>9327
>spend hours investigating why the test for a function fails
>it's the test that was buggy

dorfRTS_comms_test1.gif
(7.6MB, 600x338)

Got a lot of work on my RTS done recently.

>>9564
Oh shit you came back!

>>9415
The fucking worst, sometimes even if it passes i'm wondering if i'm testing what i actually need to test.

>>9564
Nut

>>9564
>>9569
>>9577
Didn't we have this conversation before? Deja vu is hitting me somehow.

>>9578
it'll be a real deja vu when this site goes down in about 2 weeks from now

>add new variable into my shader
>opengl freaks out and shits weird "undefined" errors out of the ass
>it's literally identical to another variable
>move things around and rename both variables, the old variable always works, new variable always breaks, even if I remove the other so the function always has only 1 of them, the new one never works
>spend like half an hour like this ready to throw my computer out the window
>it's because the new variable name had í in it (instead of i ) and despite almost completely renaming the variable multiple times I never replaced the í on it

motorworks_test1.gif
(7MB, 525x338)

>>9569
>>9577
>>9578

I made a little entry in the /agdg/ wiki for the game http://8agdg.wikidot.com/dorf-rts

Also I pretty regularly update the twitter account for the project if you want to follow me there https://twitter.com/dorfrts

I was thinking of setting up an IndieDB page for it but I'd need to set up an actual studio and publisher page, and I'm not sure what I should do in that regard since I'm just one guy.

>>9655
Only use a US keyboard when writing code anon.

>>9678
What's wrong with a one-man studio? I figured some indies do that.

halftrack_passenger_test3.gif
(7.7MB, 480x270)

>>9695

I don't think there's anything wrong with it, I'm just curious how I should actually set that up. I assume I'd need to actually register the studio as some kind of legal business, but I'm not totally sure how to go about doing that. Anyone have any suggestions?

2020-09-24_12-46-22.mp4
(6.8MB, 1920x1008, 00:21)

2020-09-24_17-01-16.mp4
(6.1MB, 1920x1008, 00:19)

2020-09-24_22-29-55.webm
(7.7MB, 1280x720, 00:18)

Posting some TECHNOLOGY (Chaos Theory wetness) and AI rewrite progress (no instant detection, suspicion stages, visual presentation of their suspicion)

>>9678
> I'd need to set up an actual studio and publisher page
nah bro
https://www.indiedb.com/games/beelzebox haven't bothered to add Peri yet but setting this up was totally painless

>>9742
No idea... But, contacting a lawyer is a good start. Not only will he give you guidance on the legal mumbo jumbo, he'll also come in handy in case some asshole claims your game infringes on copyright or some shit and tries taking it down, or worse taking you to court.

>>9773
Those are some sweet wet effects anon, video #2 is a bit too "noisy" though, maybe consider simplifying the ground a bit?

>>9773
This is fucking sick anon

fa5.jpg
(124.9KB, 1024x577)

How would you do interesting alchemy/potion crafting? Are there any games with interesting alchemy, something more complicated than just combining predefined items X and Y for some predefined effect Z.

TES puts several effects on each material, and then you figure out which 2 items that have the same effect to combine. That's a little better but hardly different than just predefined recipes.

>>9922
>video #2 
oh that's just to show off her coat. Think Fisher's suit in the lighthouse.

>>9926
thanks dude

>>10031
The question is, though, what you actually want from the alchemy? Do you want players to have to explore and experiment to get the results they want? In that case, you need to have some kind of mixing system with randomized components. If you aren't making a game that's supposed to be replayed a lot, though, that might not work out well. Are you looking to let the player build bizarre combinations of effects? There's a SNES RPG called Rudra no Hihou that has a magic system where you combine syllables to make words, and the words are the spells you use. Some of the effects are predetermined and logical (maybe WINDY gives you a wind spell, and maybe adding MORE to a spell makes it stronger). Some are obviously generated from hidden properties of the syllables and don't have anything to do with word meanings. If you look up a spell list for the game, I think you'll get what I mean better than I could explain it. Anyhow, you could go with a system like that: have a whole lot of degrees of freedom, take care of essential effects (healing, a few damage types, etc.) by hand-coding logical combinations, and then just tag effects to the components (maybe with a positive and negative effect weights system or something a little more elaborate than TES), and then let players sort out the useful from the useless effects. That could be elaborate or simple. For example, a component could restore health but also do fire damage, but there exists a second component that (perhaps among other effects) does negative fire damage (maybe "cold" damage or just reduces fire damage if it is in a mixture). So combining enough of component 2 into a potion with component 1 would negate the fire damage and leave a healing potion. 

If you were wanting to spend an awful lot of time on the system, you could try to incorporate some actual chemistry-esque features. For instance, you could let the player heat their potion and distill off the useful portion of the potion and leave negative-effect impurities behind--but the player might need to know to what temperature to heat a particular potion for how long, otherwise he'll either not do any distilling at all, or will drive a bunch of impurity into the product and undercut his work. SS13 actually has some features sorta like that for its chemistry, except that overheating things usually just results in an explosion, if anything. Then it becomes a question of how much work do you want to do making the system elaborate, and how much work does your player want to do to get the best results out of the alchemy. I guess the real question is "what of all that would be fun" and I don't know the answer to that.

>>10185
>experiment to get the results they want
Yes. I mostly just want ideas so I can come up with something interesting. How does it work, what do you actually do ingame, what kind of machines/processes could there be. This is mostly for a sandbox game like an MMO or elder scrolls where the player can choose whether or not to partake in it, for example either use normal recipes that you buy from a store or experiment to find really good potions, or just buy potions instead of making them.

What I'd like in particular is a system where there's no such thing as a "health potion", instead you combine some materials and various effects pop up, and you have to find the recipe that has the positive effect(s) you want with least negative effects. Ideally such that any 2 random items are likely to give some kind of effects, since that would make it much more fun to experiment. One idea I had was that each alchemy material has elements like "fire, blue, night, death" and then correct element combinations create potion effects. It just sounds very hard to balance the elements and effects with a system like that, especially for a multiplayer game where it's important not to have broken OP mechanics.

>negative fire damage
The idea that materials can reduce effects rather than just create them sounds useful, I didn't think of that.

Screenshot_2020-09-27_at_20.15.28.png
(113.9KB, 1024x768)

Weekly report
>Added energy bar
>Added energy
>>10185
I really like interesting magic systems.
Magica have simple, reasonable and flexible system.
I once saw prototype, where you have limited adjectives + element. Like "scattering, bidirectional lighning" to launch two projectiles in two directions. Could be interesting in some ARPG.
Not very related to alchemy, but I'd like to make one some day. Or play the shit out of someone else's.

GIF.gif
(52.6KB, 312x245)

3d technology™

I thought you'd require some special trickery for 3D but turns out you basically just need to add a perspective matrix.

>>10261
If it'll be for a multiplayer game, then you're more or less dooming yourself to extreme balancing. If enough people play it, it's guaranteed that there will be laser-like focus on effectiveness at the expense only of availability or cost. That may not necessarily be a bad thing, if you don't mind perhaps only a fraction of the player base doing "research" autism and eventually forming the alchemical equivalent of Bayer to make aspirin potions. Perhaps it's not a system that every player will need to or want to engage with to that depth, but I'm sure the ones that want to would appreciate it. One important way of balancing your components will probably be the cost of obtaining them, whether that's purchasing for in-game money, contracting players to obtain them, or finding them in the world at all. I had the idea a while back that an adventurer's guild might justify its existence partially by contracting with a college of magic to farm out contracts to procure difficult-to-find spell or potion components. That was when I was bullshitting for some other project.

29-09-20-1601413817_scrot.png
(354.5KB, 1680x977)

29-09-20-1601413825_scrot.png
(358.7KB, 1680x977)

29-09-20-1601413857_scrot.png
(298.4KB, 1680x977)

5930d445898286d51db0e4c70e9195d849f9504f09ffa6b93d4302c358e40d14.jpeg
(15.5KB, 289x225)

>Look up blueprints for the Panzerkampfwagen III Ausführung J
>Found 3 images of it
>One of the image has a different gun mantlet and a shorter barrel
>Use the second image instead and make progress with that
>just noticed now that the front and rear view has different proportions
God damn it this shit drives me up to the walls as it makes me waste hours having to try align the parts somehow so to make it look correct, how can those idiots fuck it up so bad where the misalignment is off by upto 1 millimeters? And yes off by one millimeter is significant.

>>8120
>it's too much get a grasp of when all the concepts about rendering and entities and scenes and sounds and everything come at you at once.
Yes, I wouldn't expect a beginner to fully understand. What they need to see is real, working game code that runs and actually does stuff. This hopefully gets them to start writing code rather than holding their dicks and talking about ECS and waterfalls and shit.

>>10618
>cyrillic
probably russian propaganda trying to suggest germans were shit at design

Any of you guys have a book on video game physics, especially for 2D platformers? Making my own game in c++ with and old school direct draw and have no idea where to even begin.

>>10618
Are the blueprints original or for museum recreations? Its quite possibly the latter as the originals would be sensitive military intelligence the nazis would have tried to destroy.

>>10699 (checked)
Sch-Scheiße(((
>>10708
Man I don't know, I just use duckduckgo and startpage for blueprint images, at the first sight I don't see any issues its when I start adding the image, subdividing the plane, cutting them and aligning it where I run into those issues. So the source is from various websites of whatever the search engine finds them.

>>10709
Then you should just use the blueprints as a guide and make adjustments as necessary, though you already know this. If you really want to be authentic then go learn armored vehicle engineering or something.

0ea4eb6fdceb741eca85594fa58e4828dec95991b6adffc5567249920dc27f54.jpg
(28.5KB, 600x532)

>wish C had function overloading and similar little conveniences
>finally try to get into C++ properly
>errors out of the ass
>turns out you need to manually cast void pointers
>instead of just saying "type x = y" you have to say "type x = static_cast<type>(y)" every time you use a void pointer
>apparently the ability to disable this behavior existed but was removed in some C++ version
>try Zig
>hello world gives me an error for using a tab to indent code
>well that compiler spent a good 5 minutes on my computer before going in the trash
>how about D in betterC mode
>check documentation
>void* cannot be implicitly cast to-
>well that compiler spent a good 0 minutes on my computer before going in the trash

It's a fucking void pointer. What's the point of a void pointer except to cast to other things, or for the type to not matter? Why do I have to manually type this retardedly long string of syntax spaghetti just to make it do what it was made to do? It makes me uncomfortable that C seems like the only language that doesn't wrap simple things in a load of bullshit, and that it's unironically half a century old.

>>10753
You should still be able to use C style casts in C++

>>10753
The last thing I'll do is defend the turd that is C++, but what are you doing if you have to constantly cast void pointers?

>>10753
I wish I could remember what a pointer is useful for in the first place. I haven't touched the language in fifteen years.

>>10754
That's a little better, but it's still annoying.

>>10792
>what are you doing if you have to constantly cast void pointers?
What? How about allocating memory? If you call malloc or any equivalent it's going to return a void pointer.

To be more specific I don't actually manually allocate memory often, instead I use my own dynamic data buffers/arrays of various kinds which do the allocating when necessary. C doesn't support any kind of template-like structures with a type attached to them, so I have to cast the array's data into something else every time I use it. In other words something like this:
possible language: c++, relevance: 9
struct Arr {
    int length;
    int itemSize;
    void* data;
};
Arr array = arr_new(int);
arr_add(&array, 2, ((int[]){800, 600}));And then to get the first item or loop through it I usually just do something like this:
int* integers = array.data;
for (int i=0; i<array.length; i++) {
     // integers[i] ...
}
Now C++ actually DOES have the template-typed-structure-whatevers which can(?) replace the void pointer in a structure with any typed pointer I tell it to, so maybe that would partly solve the problem after I rewrote everything. But I also use a temporary memory system which is used exactly like malloc, except you don't have to free the pointer since the data will be re-used after you tell the temp mem system to reset (e.g. after every tick).

>>10826
I don't know how you would even program without pointers. Whatever you program in now (you ARE programming right?) is guaranteed using them under the hood, and I prefer that the under the hood isn't obscured away from me.

>>10856
>If you call malloc
>in C++
You're probably not doing things in a smart way.

>>10866
I hope you have a good explanation for that statement.

>>10867
He probably means "new" and things like "make_unique"

>>10867
Malloc is more of a C thing, with C++ you must understand that 90% of the included features are there to troll you and you probably shouldn't touch them.
The New operator is usually the smart way to allocate memory for a specific class of objects, which seems to be the case for your dynamic buffers.
>but I can't extend the allocation
Be smart: create a larger allocations once the first one gets too full, copy the data over, when done swap to the new memory and free the old one.

eb3c0da73a987d982d190179b357a86fc0c14ffb9496ee4074c38ec6fdbb3517.jpg
(17.9KB, 257x320)

>no longer enjoy my go-to games
>hate my job
>no one IRL has anything interesting to say
>imageboards burn me out with their low quality or slowness
>pretty much everything in life feels like a huge let down
>the only thing that gives me satisfaction is working on my game
There is a chance I might actually be a yes-dev and wont have to compromise my vision either. Where most people would just kill themselves I have a way out.

>>10879
You can just as well use a vector for managing changing buffers, that's part of what they are there for.

a.png
(39.4KB, 508x560)

>>10870
>>10879
>Malloc is more of a C thing
Allocating data yourself is something you shouldn't do? It sounds to me like someone's been drinking too much C++ koolaid.

I don't know exactly what the mechanics of 'new' in C++ are so forgive if I'm wrong, but as far as I know it's just some OOP concept that doesn't actually do anything new or meaningful.
The example I gave is a simplified version of the actual array, it normally takes an argument for how many items you want to fit into it, for example arr_new(int, 100), and allocates a chunk of memory for 100 ints. The struct also has a member for the 100. It only re-allocates the memory if you add a 101st item, it doubles the space thus the next reallocation will happen at the 201st item, then at 401, 801...

Pic related is a more accurate representation of the arrays I use. How does 'new' improve anything? I'm not going to use a feature just because someone added it to C++ and said it's the new way of doing things, it has to actually be better in some dimension.

>>10896
Vectors have their own set of tradeoffs, for example they prioritize efficient reads over efficient writes, so I offered an overview of the general idea instead.
Hash maps are the "balanced" soloution afaik, but they have plenty of implementation gotchas than can absolutely destroy performance.

>>10907
>Allocating data yourself is something you shouldn't do?
In general, no: even malloc is in fact delegating the task of finding frree memory to the computer, because nobody wants to look for chunks of free memory by hand.
>It sounds to me like someone's been drinking too much C++ koolaid.
>as far as I know it's just some OOP concept that doesn't actually do anything new or meaningful.
You should try getting your head out of your ass and learn about things instead of parroting /g/ memes.
For example: are you sure every single time you called malloc and friends you also checked if the return is null to error out gracefully?

>>10950
>even malloc is in fact delegating the task of finding frree memory to the computer
Well duh. Any program that wants memory will have to do that, including whatever C++ feature you're proposing as an "alternative". At best it uses it's own memory management system of some kind which reduces the usage of malloc-type calls, if I cared about it that much then I'd use some memory library or make my own.

>malloc return is null
Firstly what are the chances this happens in practice, secondly if it happens then your program most likely can't continue running anyway, thirdly nothing's stopping me from error checking it either in the allocation function or outside of it if I wanted to, and fouthly using 'new' doesn't change any of this in any way at all. I'd care about that kind of error if I was making some security critical program, but I'm making a videogame.

>>10960
>but I'm making a videogame.
>in C++
>from scratch
You lack the intelligence required to pull that off.

Turns out C actually has a macro called _Generic that allows the same mechanic as function overloading.

#define vec3_add(v, second) \
	_Generic(second, \
		Vec3: vec3_add_vec3, \
		Vec2: vec3_add_vec2, \
		default: vec3_add_float \
	)(v, second)It takes the type of the first argument, and replaces the whole _Generic block with whatever matches it in that switch-like selection. The second pair or parentheses are just part of the normal macro, giving the arguments to the function.

>>10960
>Firstly what are the chances this happens in practice
Any time somebody is low on memory and doesn't have the insanity called overcommitting enabled. I have this problem with Firefox almost every day because it doesn't handle OOM properly.
>if it happens then your program most likely can't continue running anyway
This is complete bullshit. I'm writing this to you from a process that has probably run out of memory and gracefully recovered from that a hundred times since it was started months ago. I've also played games that pause on OOM to let you free up memory or properly save and exit.
>At best it uses it's own memory management system of some kind which reduces the usage of malloc-type calls, if I cared about it that much then I'd use some memory library or make my own.
Do you think malloc is a system call or something? I second >>10950, get your head out of your ass and start learning something. Though maybe not C++.

>>11087
I refuse to believe that every time you need more memory, you have some kind of elaborate error handling system that somehow works around your computer not having any more memory and your game just somehow keeps working.

>games that pause on OOM to let you free up memory or properly save and exit
I've literally never seen something like this before or heard of it second-hand. I'm sure it's neat to have it but again, what are the chances this actually happens when someone plays your game? And even then, if pausing the game is good enough for you then it's almost trivial to add a check for it in the allocation function, just pause execution right there. "Almost", because it might get more complicated if you're multithreading.

You're still not explaining how using 'new' is supposed to make any of this different. If you new something and it needs more memory, it's going to have the exact same fucking problem as malloc does.

>Do you think malloc is a system call or something
It's an alias for either HeapAlloc or VirtualAlloc on windows, but I don't know what point you're trying to make.

Panzer_III_Ausf_J.jpeg
(11.8MB, 5227x3485)

PanzerIII_Ausfuerhung_J.jpg
(120.2KB, 1280x1024)

>>10618
Welp anudda model finished.

>>11101
Can't you simply rip a model from some other game, make a few adjustments and then claim it's your own? I truly can't see how anyone is supposed to know you stole it. These things don't have serial numbers.

cigarettes.jpg
(26.6KB, 406x345)

>>11103
> I truly can't see how anyone is supposed to know you stole it. These things don't have serial numbers.
They could probably tell that if I would suddenly upload models with much higher quality texture than mine that I obviously ripped off from somebody else, so its not a option.

>>11096
That doesn't surprise me because malloc without checking return codes seems to be your gold standard for memory management.
>I've literally never seen something like this before or heard of it second-hand.
I was surprised to see it as well to be honest, I fully expected a crash considering how terribly written most games are. I've already seen it a lot in other software though.
>You're still not explaining how using 'new' is supposed to make any of this different.
I'm not bcea6b and generally I would recommend a different language actually learning how C++ approaches memory management instead of trying to write C in C++ and then yelling that the language doesn't encourage that shit. Seriously, why are you even using it if you consider everything it adds stupid? Just use C. That said, https://en.cppreference.com/w/cpp/memory/new/bad_alloc exists and it plays well with RAII.
>I don't know what point you're trying to make.
My point is that you're dumb, gay and possibly also a nigger.

7b7bd5cacbdebd5eadd4426da69fcf3b8f9664d48668b09b942fe6f7b386f5bd.png
(128.6KB, 457x645)

>>11101
Good job, post more detailed pics for comparison

>>11103
>Can't you simply rip a model from some other game
Fantastic way to get fucked if you're ever found out.
>I truly can't see how anyone is supposed to know you stole it.
All it takes is one person noticing the similarity and asking on a forum about it.

Tiger1.jpg
(131.3KB, 1280x1024)

SU-100.jpg
(124.4KB, 1280x1024)

HT100_Flamer.jpg
(115.7KB, 1280x1024)

ConVeh.jpg
(108.7KB, 1024x1024)

VisaginaNPP.jpg
(123.5KB, 1280x1024)

>>11110
Thanks, you mean as in other models I have mode? Well here you go.

>>11107
>why are you even using it if you consider everything it adds stupid
I literally said in the original post that I wish C had some features like function overloading. I don't care about the 'new' object mechanic or 99% of the bloat C++ has, but the argument seemed to be that allocating memory yourself is stupid and using 'new' is somehow the """smart""" alternative, yet I'm failing to get any kind of explanation whatsoever about why either one of the points in that argument would be true.

I acknowledge the possibility that memory running out is a bigger problem than I've ever known before, but there's no C++ feature I know that stops that problem from existing and requiring you to deal with it. If there is something like that then explain that feature and give it as an argument instead of resorting to ad hominems like some leftist golem whose worldview is being challenged.

>>10950
>In general, no: even malloc is in fact delegating the task of finding frree memory to the computer, because nobody wants to look for chunks of free memory by hand.
Manual memory management by just allocating all the memory you need at startup and divvying it up at runtime is (or was?) incredibly common in large game projects, at least ones made using proprietary custom engines. It was virtually a reqiurement for console games, even.
Not sure how many of the nu-devs making their AAA gaymes in Unity or UE4 work. Many of those at least change the engine's source code, even the Unity ones since they get access to Unity's source code, so it wouldn't surprise me to much if the better ones still did it. But then again how many ones that use those engines in the first place could be considered part of the "better ones" when the majority of the industry is going down the shitter in terms of developer quality?

>>11115
>yet I'm failing to get any kind of explanation whatsoever about why either one of the points in that argument would be true.
That's because you're a complete retard who complains about "bloat" without even understanding what is required for function overloading to make sense.
Learn the fucking language, stop parroting buzzwords you heard from some random anon, and most importantly stop blaming your incompetence on everything and everyone but yourself.

>>11116
Coding your game with many fixed-size structures so that they might be allocated (and accessed) efficiently can be smart in some situations, but don't think for a moment than the epic spaghetti world of console code is a good example to follow: that's how you get parallel universes and crashes when the camera is angled just the right way and thus one texture too many ends up on screen.
>so it wouldn't surprise me to much if the better ones still did it.
They almost certainly don't.
See Factorio devblogs about the fluid pipe system for examples of how things get when you have extreme performance requirements AND run into memory access limits, a situation beyond what 99% of games will ever encounter.
The TL DR is that it's important to keep only the important parts of the object in the critical loops, and  that it's important to have some robust memory compaction system to keep those important parts in contiguous memory.

6193c1121b574f696e20d8f671f87bc34967b1e104951ade17ca3fe87f5191d2.jpg
(72.7KB, 650x564)

>>11117
>it's your job to prove my argument is true

1241989958-0.jpg
(151KB, 1334x1334)

>>11118
Learn2Code faggot.

>>11115
>If there is something like that then explain that feature and give it as an argument instead of resorting to ad hominems like some leftist golem whose worldview is being challenged.
How about you learn basic C++ instead, this shit is introductory material. Calling you a dumb gorilla nigger is not an ad hominem by the way, it's an insult. It's also a statement of fact, you dumb gorilla nigger.

>>9742
That's very, very different in every country and, in the US, every state.  
Look into your local regulations. You can probably enter a fantasy name for a Sole Proprietorship for very little money, with the problem being that you'll be personally responsible for any charges on the business since it's not incorporated. Chances are you wouldn't even need a lawyer for that.
 Incorporation is more expensive, but protects personal assets in case of a lawsuit/debts.

>>10031
Kingdom Come is generally recognized as having a good alchemy system. It's pretty fun, but only works if you value potions highly in your game. TES-style drinking of a hundred potions would be out with that system.

>>11124
>>11118

cam.webm
(2.3MB, 960x540, 00:22)

>>10281
Now with a camera®

I think I've seen at least 3 completely different ways to do cameras by now. I couldn't understand them, so I just used a normal transform matrix and it works exactly like you'd expect.

>>11199 (checked)
>>10281
Interesting, what plans do you have for your game engine?

>>11305
For now I'm just learning to properly understand OpenGL and rendering and all the related math. I'll probably switch to 2D and continue a "roguelite" I started before, I want to finish a smaller game before I try to make a 3D one. Who knows though.

1447195362638.jpg
(14KB, 249x202)

>look up my old microtank releases
>oldest version I have is MicroTank_v1d.pk3
>last edited: 21.04.2013
>it's been 7 years already
oh god, it has become the Duke Nukem forever of modding.

output.webm
(1.3MB, 1280x962, 00:31)

Yeerk's weekly
>AI now can react on Yeerk
I was hoping to also add attack, but unfortunately it stays for the next week.

UCTR_temp_titleback.png
(76KB, 545x415)

RT_Saki_Aubrey_SD_sketch.jpeg
(84.5KB, 969x719)

>>6917 (OP) 
Dropping in my first-time post for a project I've been posting on other parts of the webring:

 Uncommon Time Rewrite Project 
Current Codename : Retuned

'Features, both upcoming and implemented
-Complete reworking of the story, battles and gameplay of the RPGmaker game, Uncommon Time.
-New characters; both party members and side-characters.
-New areas, including a inner-mountain dungeon and a battle arena.
-Real villians, that do actually-unforgivable things.
-Tumblrshit is purged. All lgtbbq characters and references That don't apply to villains are reworked or replaced.
-Likewise the worm is now a straight man. The faggot is now a former-bi straight man He used to be a prostitute at one time in his life. The detail itself is ultimately insignificant and a throwaway line.
-Teagan is no longer the "bad guy" of the game.
-The developer's self-insert is no longer a self-insert. She doesn't bitch out Teagan in act 2, either. She and Teagan are friends through and through in the rewrite.
-Carbon-copy-stock-anime-personality characters have been rewritten to be more organic. Most notably, the earth-fairy is the most drastic change.
-Outlets for you glorious waifufags.
-Realistic depictions of narcissistic personality disorder.
-Act 2 retooling to replace boring-filler-dungeons with vision-quest-mini-dungeons for all the party members.
-A secret-true-end where you fix everything for good.
-Revenge subplots
-Maybe some tasteful lewds. OG-UT showed no tits despite swearing, blood, other nudity and suicide aftermaths.

Current progress
-Raw writing is about 70% done for all arcs, what is remaining is the uncommon time sequences, and most of the non-critical parts of act 3 and the secret-final-act.
-Gameplay Implementation is currently sitting around 30%. Progress is up to Bel Canto Act 1, Day 3 post-rehearsal.
-Art is literally just started. Barely. My plan is to really get into art around the Nuclear Teagan demo release, otherwise, I'll probably drop more dumb-sketch-concepts just to give you an idea what I plan for the characters to actually look like.
-Boss implementation is currently about 50%, most stock-sprite-bosses are not yet reworked or implemented, neither are most of the secret-final-act bosses.
-Arena implementation is at 30%, all that is currently there is solely part of the story. Later it will include more battle sequences at various progress levels, that will unlock at act 2.

Bugs, mistakes and other shit
-Arena progress and battles are broken as fuck. Players level up way too fast in the arena, and certain enemies are grossly overpowered.
-Healing that isn't percentile-based is broken as fuck, and full-heals the party. I just want to make non-static healing that somewhat-scales off the player's MHP.
-Most not-fully-implemented bossfights are balanced around my broken-as-fuck-healing-system, and will probably remain broken until I get my progress up to the point where it's meant to exist.
-Typos and broken nametags galore. It's hard enough being an editor when you're trying to gamedev. So any further releases I post, if you find any I miss, let me know so I can crush them, please.

Second pic related is a super-deformed-sketch of how I imagined how the ex-faggot and ex-worm look like. Left is Saki, Right is Aubrey. I didn't want to deal with my anatomy perfectionism, so I made it SD on purpose.

>>11579
You're still working on it motherfucker? You've been at it since /vr/ was made goddamn. Does it still support Zandumb or have you moved to graf zahl's shithole of a sourceport?

495aca9d5854b552951bb24ac5b95c196d85828a00bb3826149c8b38a2f037be.jpg
(49.8KB, 800x900)

>>11877 (checked)
Yes, the past few days I mostly cleaned up the code to make it expand a bit more easily.
>Does it still support Zandumb
Not anymore because the feature parity distance is so far apart its not funny anymore, I mean it's like about 3 years behind in features and it lacks basic things such as letting me use A_Jumpif(Inventory(Actor) == n, State) since A_JumpIfInventory has some retarded boolean operation that is assumed by default which is >= and when the number is 0 it has a different behavior. Secondly I need the A_Warp which is a bit cleaner implementation then spamming A_SpawnItemEx() on every damn state as of how Sgt. Mark IV does with his tank code.  Currently I cannot do shit with ACS since Slade keeps mysteriously crashing on me despite it worked before just fine and I have absolutely no idea why, the error box doesn't give me any useful hints or information just gibberish. 
>have you moved to graf zahl's shithole of a sourceport?
Yes, it fucking sucks I know but there is no better alternatives. Maybe when k8vavoom supports more ACS opcodes/decorate features I could try adding support for that, I tried running the mod and well so far its seems everything works apart that the sbarinfo/custom HUD is badly butchered.

>>11873
>>11879
I truly do respect your autism/commitment to this project and I can see you've come a long way and made some great progress.  I must ask, however, why you can't just remake it in some variant of RPGMaker instead of whatever Zandumb and the other thing is.

I've never actually watched/played UT all the way through; every time I saw it, the experience would have been 20+ hours of watching livestream archives.  Why keep going with this project for so long?

>>11879
Oh well, it just shows how fucking old every damn thing is. I'm somewhat even more surprised vavoom is still in development and somehow catching up faster than others, though half of that i'm assuming is not having to worry about netplay code desynchs.
At the very least you're still working on your mod unlike some other individuals.

>>11881
He's modding in the doom engine, or at least what would be called the modern interpretation of the ZDoom engine. MicroTank dev, not uncommon time.

>>11881
>I truly do respect your autism/commitment to this project and I can see you've come a long way and made some great progress.  
Thanks.
> I must ask, however, why you can't just remake it in some variant of RPGMaker instead of whatever Zandumb and the other thing is.
hmmm I did have some idea to use RPGMaker to make some gaems with it, though I'm not sure which variant is the best ones, I'd prefer the version which lets me change how the combat loop works by scripting. But yeah I really should find a babbies first video game maker program or something before I use something more serious like Godot just to give me some damn basic idea how games programming work. Well I have another friend which I sadly have very rare contact with him and I'd told told to him that I am going to continue working on this working instead of letting it collect dust, which he appreciated it. 
>RPGMaker
I suppose I could try make a list of features I'd would like to implement and see how far it goes, however I want at the very least finally release V1U 2.0 of my mod which should be much more polished than the last RC version which is around 3 years ago. Also RPGMaker as it says is mostly a RPG oriented game engine, not a FPS/TPS. But I do have some ideas which should hopefully be fairly "realistic" with that game engine. 

>>9742
>Visible troop comparment
huh interesting that OpenRA supports that, since when did they do that change? I wonder if its also possible to make it tow AT guns like in Sudden Strike 1.

>>11883
>>11884
>modding in the doom engine
Oh, fuck, I misclicked.  My post in >>11881 was meant to be directed at >>11873 alone.

>>11877
Are you aware that Zandronum is at this point basically a fork of GZDoom?

>>11883
>Oh well, it just shows how fucking old every damn thing is. 
Yeah man and nothing gets any younger, for me personally it's sad to see that many modding communities such as the one for UT99 and UT2k4 are gone and among other gaems. With modern games there is no such thing such as a community with dedicated serbs, own culture and modding groups that do custom shit for fun and giggles. A era long past by. 
> I'm somewhat even more surprised vavoom is still in development and somehow catching up faster than others
it's actually K8vavoom not vavoom which hasn't been in development in ages, the developer of vavoom acknowledged K8vavoom on his website, the developer of K8vavoom is a different guy who is posting on the doomworld forums. It's interesting his engine has shadow mapping and stencil mapping something which Graf up to this day still fails at horrible but wants to have always the latest rendering technology without putting them actually in use. 
>>11887
No big deal.

Also some minor news, I finally managed to improve the ACS script that is responsible for tank movement/rotation which actually now does the rotation changes according to the player input, it still however for technical reason gives out 4 movement items so to make animation/sound play to work. The bad thing however is that the tank moves even faster than it should be because the stupid function ThrustThing is a fucking integer value and not decimal/float, there is another call that lets me manipulate velocity but it requires me fucking around with the 3 vector types (X, Y, Z) which I do not know how and I'm not even sure if it lets me a value less than 1. It's so fucking annoying.  Maybe I should really suck it up and move that part and only that one to ZScript instead and be finally done with the stupid tank code.

>>11881
>>11887
No biggie, and just for the record, I've been at this for one year, not seven. And I'm using the same engine the original was made with; RPGmaker VXAce.

>Why keep going with this project for so long?
Three reasons:

1. Because this is my personal journey to get over the emotional hurdle of "doing something big".
2. I've found my calling in teaching and/or inspiring people to mentally and emotionally help themselves, instead of just relying on others for either. Uncommon Time is partially about introspection and self-analysis, so expanding upon that, while using what I've learned through my own experiences, and the stories I've read sounded fun as hell.
3. Why not?

>Zandumb
Never heard of it. In fact, searching for that just gives me either polish shit, or some streamer with an idiocy gimmick. Do you guys mean 'zandronum' when you say 'zandumb'?

>>12076
>No biggie, and just for the record, I've been at this for one year, not seven. And I'm using the same engine the original was made with; RPGmaker VXAce.
Yeah, that makes much more sense; I was initially shocked that anyone would devote so much time to rewriting UT because I thought you'd been working on it for as long as the guy making the RTS game in a fucking Doom engine fork.  I wish you the best in your work.

>>12083
thanks man, I'll try my best. At the very least I'll make something fun and memorable.

>>12083
> thought you'd been working on it for as long as the guy making the RTS game in a fucking Doom engine fork.
>RTS game in a Doom engine fork
U wot?  No anon where working on that, well somebody made a C&C styled mod for it if you look around at ZOGtube.

>>11873
>Raw writing is about 70% done
I remember you pitching your idea way back on pre-kunt /vg/. It's great to see you've gotten this far.

>>10031
In a terribly fucking complicated fashion
materials have properties, similar to TES, but in addition to just "it does x" have potency based on health of plant, soil quality or rarity or health of whatever creature you pulled the reagent from, have the possibility for reactions only possible in certain environments i.e. ignition points, chemical reactions (think ss13 style boom/smoke/foam as the most basic)
I was autistically dreaming of an idea for a game just mostly centered on alchemy and its applications in an RPG or shopkeeper kinda' setting

i built a system for saving and loading controller profiles since my game involves a bunch of joystick/throttle/footpedal/headtracking nonsense. 

I often wonder if anyone will ever actually use this stuff though.

>>12390
There're only two ways, I think, to handle that. Either find your audience and get actual responses about whether they'll use it or not, or don't worry about it at all and put in what you yourself would want in the game. If you start second guessing yourself on every feature you're could end up in that dark place where nothing seems worth doing. It's hard to do that because you do have to maintain perspective and not spend six months on something not really critical, but that's part of the balancing act that comes from experience, I guess.

>>10031
A potion is a combination of ingredients.  An ingredient is one type of object within a larger class of objects.  That simple fact is what bad crafting systems forget, and you encounter problems like Fallout cooking where you have a recipe for Mole Rat Soup and you can ONLY use Mole Rat Meat in that soup, and there are no other types of soup for Bighorner Meat, Deathclaw Meat, etc.  The way I would create a good crafting system is to make each potion require one of the following:
>container
>1 or 2 base ingredients
>1 modifier ingredient (optional)

Say you wanted to make a health potion.  The recipe for that might look like:
>glass vial
>red herb
Which would produce your archetypal red liquid inside a small glass vial, assuming that a red herb has a simple effect like "restores HP" and the glass vial has a simple effect like "does not affect potion."  If you wanted health and mana in a single potion, that might look like:
>glass vial
>red herb
>blue herb
And you get a potion that restores health and mana.  However, if you wanted to change things and experiment, then you could do:
>magic seed shell
>red herb
The red herb makes it restore health, but the magic seed shell (I'm imagining a Deku seed from the Zelda games; something like a magical coconut) makes any potion within it slightly stronger at the cost of being applied over time instead of instantly.

Then you could do something spicier, like:
>enchanted vial (potion is doubled in potency, but temporarily reduces max health)
>red herb (restores health)
>black herb (increases attack and poisons user)
You get a potion that restores double the health of a potion with 1 red herb in it, increases attack compared to a potion with 1 black herb in it, and poisons the user twice as badly as a normal black herb.  It's a strong but risky potion, but there should be a way for crafty users to get around bad effects.  Maybe there's a magical beehive in your world where the bees are giant and produce extremely high quality honey - if a player fights his way through the hive and collects some honey, then he could use that as a rare and powerful crafting ingredient.  Final example:
>enchanted vial (potion is doubled in potency, but temporarily reduces max health)
>red herb (restores health)
>black herb (increases attack and poisons user)
>magical honey (removes any negative attributes from potion)
And the result would be a potion that restores double the health of a red herb, increases attack by double what a black herb normally does, and doesn't reduce max health or poison the user.

I hope all that kind of made sense.  The basic gist is that a crafting system should be a set of logical rules that the user can follow and which should encourage the user to experiment with new ingredients.  Bug Fables had a very simple crafting system (food A + food B = food C), but there were consistent rules in place and lots of attention to detail from the developers, so I had a lot of fun exploring it and finding new ingredients.
>Egg + [nothing] = Fried Egg
>Leaf + Egg = Egg Wrap
>Leaf + Fried Egg = Egg Wrap
That last line is also extremely important: your crafting system should handle small deviations well, which in this case means the recipe for an Egg Wrap doesn't care if the egg is raw or fried.  In a more linear system, Leaf + Fried Egg might equal failure, because the Egg Wrap recipe specifically only cares about a raw egg; such an outcome would be programmatically consistent, but unsatisfying for the player.  The example I mentioned earlier with Fallout's cooking system has exactly this problem.

>>12540
>Egg + [nothing] = Fried Egg
>Leaf + Egg = Egg Wrap
>Leaf + Fried Egg = Egg Wrap
How bout instead of following a formula you enumerate the ingredients first and then decide what it should be logically? Consider Minecraft shapeless recipes.

>>12545
>enumerate the ingredients first and then decide what it should be logically
There's no reason you can't do that in the system I described.

how do meke ?gem

>>12596
just like press the game make button

>>10702
You don't need a book. Plenty of howto guides online. Do you know enough math and coding to build a line-line and box-box intersection test? If not, go work on your math first.

>>10826
Do you know what an address register is on a CPU? I'm guessing you don't.
>>10856
I do similar things for custom temp memory "scrathpad space", as well as a few other custom memory routines. In C++ you could try using templates but if what you have works, just use it. C++ just punishes you for it, as they believe you should be using templates, auto keyword, and the like.

mkgm.jpg
(12.4KB, 285x305)

>>12596
paypal me $900. I'll show you the 100% fool-proof method to make game.

>>12514
I've had people tell me they'd use a feature only to never actually use an option once it's there. 

luckily i have a decent rhythm going so the profile system only took probably a few hours over a day of mixing in doing other stuff. Hopefully it will save people more time down the road than what i put in.

>>12893
>I've had people tell me they'd use a feature only to never actually use an option once it's there. 
Asking people what they want and expecting a useful/honest answer is genuinely one of the most naive things I can think of.  I cannot possibly think of a worse way to get true feedback about what people like.

>>12893
That shouldn't matter, not really.  If you think the feature is useful and improves your game, you should put it in.  Feedback matters more when people tell you they don't like a thing.  It's not very often they'll tell you they do like it.

Also, controller profiles is a good idea in general, keep at it.

So, I took a programming course back in highschool, making fibronacci calculators and console menu trees.
If I wanted to make a small 2D game from scratch, what should I focus on learning?

>>13061
>from scratch
if by that you mean "make your own engine", then learn more about whatever language you used in highschool and figure out how to make OpenGL work with it.  Or just learn C if it was some meme language.

If "from scratch" you mean absolutely everything from the ground up, assembly.

output.webm
(212.4KB, 1280x962, 00:01)

Yeerk's weekly
>Still working on attack
Animation is in place, but no hitbox yet and no mechanism to exit the state.

>>13061
If from scratch means that you want to write the engine yourself, basic geometry and trigonometry is really all you need. Disregard the idiot recommending C and OpenGL, you can do this in basically every language.

bfoh3xnd.png
(71.4KB, 291x274)

>>13114
>you can do this in basically every language
>but the person who's recommending this particular commonly used language is an idiot and you shouldn't listen to him

>>13074
>Or just learn C
C is a terrible language to start with: it's extremely niche, obtuse, and in practice it has been mostly replaced by C++ (which is even harder to get into).

>>13122
>>13114
C/C++ aren't difficult at all. It was the first language I learned and I started shitcoding small games at 14 with SDL. Sure any language can maek gaem but I still recommend C++ to start because it'll teach technical details that Python or whatever soyfusion language won't.

>>13119
Yes, because OpenGL is an awful start into graphics programming, C drowns you in technicalities and tends to create bad habits, and both are complete overkill for what he wants to do. The point is that almost every language has something to programmatically MSPaint on the screen, so he should choose one that he's either already familiar with or is easy to pick up. C is likely neither.

>>13122
>in practice it has been mostly replaced by C++
wat

>>13124
>C/C++ aren't difficult at all.
They're so difficult that even teams of professionals end up writing code with tons of memory safety issues.
>I still recommend C++ to start because it'll teach technical details that Python or whatever soyfusion language won't.
This is ass-backwards, you start with the easy concepts and then learn the technical details (that don't even faithfully map to modern hardware that well, but still) and how to take advantage of them.

>>13134
>teams of professionals
Imported pajeets and liberal college trustfund kiddies, you mean.

>>13127
Oh so basically you're saying he should use something that lets you call DrawEpicSpriteOnScreen(x, y) instead of learning actual graphics programming. It's definitely easier to make a game that way, I started that way too, though it's less "from scratch" than using something like OpenGL directly.

>>13074
I haven't programmed in a while. Care to give me a challenge to get back into the mindset? A mid-range project, as in "If you can code this, you'll probably will be able to make games if you keep learning" 

A stepping stone if you will.

>>13157
MMORPG

>>13217
can't make a worse one than TORtanic amirite?

>>13157
As others have pointed out, unless you go with a full engine already you likely want to go with OpenGL since it abstracts more than Vulkan but isn't some useless bullshit like pygame. Get a basic window with gemoetry going and a shader or two, getting that going should be if nothing else good motivation.
If you want something even more goal oriented, try implementing a graph or tree search and visualize it.

>>13061
If you've never made games before then by all means learn an existing engine first. It help you grasp the high level concepts (e.g. colliders, instances, camera control) before moving on to the low level ones (e.g. rendering, buffers, implementing the high level concepts yourself) which are required to make your own engine. When that time comes you can use any language provided you optimize your code, and as you gain experience you'll gravitate towards the lower level languages like C for more control and fine-grained optimization.
For starters I recommend making a small game or two in Unity/Godot, then remaking those games in C/C++ with the SDL2 library + OpenGL.

>>13157
There was this list of games that aspiring programmers should go through to ease themselves into gamedev, but I can't find it right now, so maybe someone else can post it. The first entry on it was Pong, followed by games like Breakout, Tetris, Space Invaders, Pac-Man, Super Mario...etc.
The idea here is that you start small; a game with simple mechanics, minimal assets, and one level. Then you gradually move on to more complex games, learning concepts and practices along the way. That way you'll stay motivated and not overwhelm yourself. Remember that resources like tutorials and books are a great help, whether using or designing an engine.

>>13157
Make Tetris.  I recommend reading the book "NAND to Tetris" or whatever it's called if you really want to go from scratch.  If all you're looking for is making an engine though, just make tetris with some basic libraries.  Ignore all the people on here saying "hurr durr it's hard", no shit it's hard.  None of that other crap they say is "from scratch", I imagine you want to do this because you want to learn, and if so, go with a language that is as close to machine language as you can stomach, you'll learn a lot more than some joke of a language that isn't even statically typed or has garbage collection or other nonsense that just makes you a shittier coder if you don't know what you're doing.

game-programmer-compressed.png
(1.3MB, 800x18907)

Thought you guys might like this.
Sourced from https://miloyip.github.io/game-programmer/

>>13294
Just what I needed to have an excuse not to actually do work.

>>13127
>wat
C++ is dominant in basically the entire "perfomance desktop application" space (amongst other areas), and games are one of the areas where performance is in the highest demand since there's constantly a push for them to do more and more - better graphics, lower requirements, higher framerates, anywhere you can squeeze a bit of extra peformance out counts: so C++ is the natural winner.

>>13300
You're right, except for a lot of performance reasons they then proceed to not use most of the features of C++ and instead mostly write C. Do we really need to link to the Mike Acton talk every thread?

>>13300
It's hard to pinpoint what exactly "perfomance desktop application" means, but even with a liberal interpretation this is a tiny space. Most of the shit I use is written in C. Game companies use C++ because it's easier to find developers for it, this is well-documented. And like the other anon said, they largely use C++ as a weird C.

>>13302
Yeah but the times you do want the C++ features, they're there, and you don't have to awkwardly try to recreate them by hand. Moreover, C++ generally has better standard libraries than C because C++ caters to library builders. Most of the weird bloated features in C++ are because of that.

>>13307
>It's hard to pinpoint what exactly "perfomance desktop application" means, but even with a liberal interpretation this is a tiny space. 
A tiny space that happens to include all game engines. Since we're talking about game development, and programming languages (as opposed to scripting) it's a safe bet that we're talking about making game engines. Come on, anon. Use basic logic.

>>13307
>they largely use C++ as a weird C
Isn't that what the name "C++" implies? It wasn't supposed to be something totally different than C. It was C but one better. Early C++ came about before many of the outre programming concepts that have evolved since then and been back-ported into the language. Programming languages are like that, globbing on new features as their devout users try to evolve with the times. I think you can find object-oriented versions of FORTRAN if you want these days.

>>13311
In '80-'90 maybe. Modern C++ has little to do with C, a lot of C-style stuff has been discouraged for over a decade, and that's the direction the committee is taking the language too.
>Early C++ came about before many of the outre programming concepts that have evolved since then and been back-ported into the language.
Not really, no. A lot of these concepts are older than you'd think.

Clicked submit too early god dammit
>>13309
Only the big commercial game engines are, which are not relevant if the man can barely program and just wants to make a 2D game from scratch.
>programming languages (as opposed to scripting)
This is a nonsense distinction.

>>13313
Nope. You can still write what amounts to C98 and all modern C++ compilers will compile it fine. You can use C++ just to avoid passing a pointer to every single function call you make (as you end up doing with vanilla C, without classes). If that's all you use C++ for, then you're already ahead. Might as well use the auto keyword here and there while you're at it.

>>13315
What's your point? Of course they're not going to break backwards compatibility, look at the Python 3 disaster for what happens when you do that. That doesn't change that it's not how you're supposed to use the language and it's not how the language is mostly used.

Weekly Progress Report

I would've had this up yesterday but snooze was acting up for some reason.

-Changed up character class names; only the Cantabiles will have musician-themed class names.
-Continued progress on day 3 post-rehearsal cutscenes. Pic related.
-Added a reset button for Metronome Tower's pre-boss area to prevent a potential post-cutscene soft-lock for underleveled players.
-Certain Act 2 writing sequences put in limbo for now. Not technically cut, but close to the chopping block and I'm hesitant on that. Elaboration below in spoilers, because well, this discusses serious spoilers despite how little I've actually elaborated on.

SPOILERS ON CUT WRITING

I've been jumping back-and-forth on whether Soprano deserves an Uncommon Time, based on the deplorable actions she'll take in the endgame of act 1, and in the first third of act 2. I've really wanted to give her an Uncommon Time, as to artistically portray the "mental brokenness" of people with Narcissistic Personality Disorder. On the plus side, I feel like it would create some solidly-portrayed cutscenes and maybe even create a few tasty reaction images. On the minus side, it would just create more busywork for me when the Uncommon Time sequences of the party members are already gonna be as large of a time sink as Bel Canto has been for me right now. Ultimately it wouldn't be too detrimental to the overall story if I did consider it cut, but for some reason I'm really hesitant to cut this section.

>>13370

the_pain_of_vomiting.png
(108KB, 640x414)

>>13370
>>13372
I not only fucked up, I forgot to include my credentials and screenshot of the week.

>>13316
>That doesn't change that it's not how you're supposed to use the language 
Who cares? It doesn't matter that you're not "supposed" to use the language that way.

>>13464
>Who cares?
He does: >>13311
>It wasn't supposed to be something totally different than C.
And he's right to do that because fighting the language at every turn is not fun.

14-10-20-1602698027_scrot.png
(2.1MB, 1680x1050)

HeavyTankA1_R3_0.png
(844.2KB, 1920x1080)

HeavyTankA1_R3_1.png
(719.7KB, 1920x1080)

Welp I made a new tonk model after I didn't managed to get the stupid AMX-30 turret to work. Based on my older heavy tank a1/2 model, as well a chemical and a bee hive shell which are still not done yet. The chemical shell has some issues when the shell hits the wall where the chemical clusters are getting clumped instead of spreading around which is going take a while to fix it as well having to polish the particle effects some more.

>zzzchan finally has an agdg thread
>click
>it's one month old
And I've been watching the dead smugloli thread all the time. Whatever, I haven't done anything noteworthy in the meantime that's related to agdg, I'm still searching for my enthusiasm or at least a minimal volition to do anything useful, which I've been missing for at least 10 years now.

confusion.png
(367.6KB, 486x463)

>>13552
If you wanted one earlier, why didn't you make it yourself?

>>13572
Why bother making a thread when I don't have anything significant to contribute to? To have another dead thread?

18-10-20-1603047804_scrot.png
(2.7MB, 1920x1080)

UKM-2000_0.png
(351.7KB, 1920x1080)

UKM-2000_1.png
(327.4KB, 1920x1080)

>>13514
I fixed the issue with the clumped goos which is caused because their collision boxes were bigger than their carrier projectile so I just set the height and radius to 4 but I since I made also several tweaks to it it gets stuck in the ceiling instead of spreading chemicals at the floor at least its more usable now. I retextured the cannon to black-ish/grey colour and made a new MG model based on the polish UMK-2000 which is just a modified PKM that is chambered for 7.62x51mm NATO rounds. I'm wondering, is it a bad idea when I just copy this text to other /agdg/ thredas? As it feels to me it is a waste of time rewriting the sentence structure which has basically the same meaning.

lots of game-related progress and polish done this week. Hopefully by the time I've ironed more mistakes out and rounded out the game portion things will start coming together more smoothly and quickly during implementation. This process should be even smoother once art is done.

Weekly Progress Report
-In my efforts to unfuck my system so it actually can play .mid files through the machine itself, I've messed around LMMS, importing .mid files and replacing instruments. Not sure if anything substantial will come from this, but I might replace a track or too depending on how it sounds.instruments with ZynAddSubFx presets.
-Raw writing progress continued for act 2.
-Added a magic circle just after the Hellbird fight in Arpeggio Pass to give new players some breathing room after the fight. 
-Zero-resist state, and monster-skill curse of the exposed added.
-Shuffled the names of some of Mezzo's skills. Volante is now called Stentato, Forza is now Volante.
-Changed Alto's Vibrato to a new skill:  Zusammen. Afflicts the enemy with Zero-resist, Alto's sole non-damaging affliction spell.
-Changed Teagan's Bioluminescence to a new skill: Mirror-Shattering Icicle. Removes magic reflect and causes freeze/cripple/concuss
-Changed Aubrey's Blending to a new skill: Amorevole Revives with 75% HP and adds affliction shield
-Mezzo's skill Loco Attaca no longer removes magic reflect.
-Mezzo's skills are now considered magic and thus can now be reflected.
-New items added: Iron brew, Mirror Curtain, and Shield-Eating Potion. Former two are rarer, grant ironbody, and magic reflect. The latter removes both effects and can be bought in Stave, Bel-Canto, and Tenuto Spring.
-The secret Libretto container that contains pine rosin, now contains a tuning fork, for those having trouble with the boss fight of that area.
-Added an npc to give vague hints about said tuning fork.
-Sprinkled the new items in various parts of the act 1 dungeons.
-In my efforts to trim and streamline music and reduce bloat, I have replaced Dance of the Sugar Plum Fairies with the Ruins track in Euphony Temple. FP has too many music tracks that are used exactly once, that shit kinda annoys me when it goes overboard like that.
-Reworked the old ability functions of Meirin's Strawberry Barrage, now it will function as intended and use the player's equipped weapon as the attack.
-Crushed the unformatted name tags in Stave Inn's cutscenes, Euphony Temple's hint rooms, and Nuclear Teagan.
-Fixed a broken new-line tag in Teagan's win quotes.
-Ran into an RPGmaker bug that involves accidently deleting, and undoing said deletion of the entire animation database. It fucking unbound every single animation from items and weapons and I had to stitch everything back up by hand. Well, at least it didn't undo skills, thank fuck.
-Made the decision for the writing I had in limbo: it is going to be cut for the completed mod. The decision is made based on Soprano losing story-involvement for much of acts 2 and 3 despite being part of the main cast. So if by some whim I get back to UT-RW some time after I've completed it, I'll probably include the scenes I'm cutting. The boss fight I had planned for this will be moved to Glissando Arena's optional second tier.

No screenshot this week because I had a .ogg file of blue forest with edited instruments for it instead. Snooze didn't like the file format, and I don't feel like trying to make a video for a shitpost-tier track.

>>14519
Jschan should support ogg tho, if you rename it to oga. Maybe.

Decided to get some actual work done today.  In order to implement some kind of modding system within Unity, I've been using .NET features like file.load and xml document streams and crap like that in order to load from a file instead of getting my assets from within Unity's package system.  This works fine, but I have a question about it - since C# and .NET are Windows proprietary nonsense, will this break Linux compatibility if I do a Linux build package?  I have my Unity setup so that it can do Linux builds, but as I don't have a Linux distro on my current computer and don't feel like setting up a dual boot, I have no way to tell if it does, especially since I'm a layman when it comes to how Linux actually works.

I'd upload an image of my progress or talk about my game if it was anything other than just a barebones prototype that isn't even a game right now, so hopefully I can get some actual progress to show in this thread at some point as well.

>>14520
>>14519
Thanks for the tip anon, here's the track in proper file format

>>14631
>since C# and .NET are Windows proprietary nonsense
I'm pretty sure .NET Core is licensed under the MIT license which means Microsoft can't really jew everyone with it otherwise the opensource community would just fork the latest version and continue without them.

>>14631
AFAIK Unity uses Mono even on Windows, so as long as you don't try to load native code, you shouldn't have to worry too much about compatibility. File I/O or XML parsing is definitely supported under mono, unless you try to do something really fancy, it'll likely work under mono too. (WPF is the biggest missing piece, but there's no point in using it in a Unity game...)
Of course, this doesn't mean you don't have to test under Linux, because even with the most cross platform libs, you'll eventually find one little detail that works differently on different platforms...

>>14669
Unity doesn't use/support .NET Core.

TL;DR: If you're worried about proprietary nonsense, you should be worried about Unity, and not the .NET/C# parts.

>>14688
Maybe I'm wrong, but I thought Microsoft was moving towards Core as the main implementation of .NET. So Unity will be forced to support it eventually and probably not too long into the future. But if these things are a concern then who knows, Godot 4 is supposedly right around the corner and we can all tell Unity to fuck itself.

>>14696
Not necessarily, while it's based on mono, based on the amount of shit they changed in it to optimize that shit for games, I don't think they'd have any problems if they'd have to maintain their fork in the long run. Or create some frankenstein implementation where they pull the new class library and whatever, but keep their old runtime.
I'm not saying they'll go this route, but they can. Like how many people/engine/game still use lua 5.1, when 5.2 was released in Dec 2011.

>>14688
>no point in using it in a Unity game
big rip.  I've been using it to make a dialog window appear so that people can choose files to import.  Is there a good way to do this without using System.Windows (which is the WPF namespace)?  Might just have to get rid of that in a public release, though I was only using it for the map editor to upload specific files, so I guess it's not that big of a deal.

>>14773
Windows Forms should work under Mono (mostly), and I'm sure it has a file chooser dialog, so you could just use that instead. I used this tool in the past on Linux under mono without problems, other than being ugly.
https://github.com/jaquadro/NBTExplorer

output.webm
(1.1MB, 2048x1202, 00:09)

After two weeks instead of one
>Pineeres's attack is finally working
>Yeerk is dying
I've had to implement additional animation system that is unrelated to movement for that. Pretty happy with result.
Core mechanics are almost done. Hope to finish them next week.

suffering.jpg
(27.9KB, 430x430)

>think about for several days if I should bother using ZScript and foregoe the possibility of K8Vavoom support or not
>settle on ZScript instead 
After some fiddling around with ZScript I managed to convert a few actors to ZScript mainly the base class and the weapon class. There isn't any noticeable changes though I had to fix a few bugs that happened after the conversion. It fucking sucks because now it makes my mod now entirely dependent on Graf Doom, as I would like to see if after tons of bug fixes if K8Vavoom would run overall better or not. Still I have a older version of my mod that sort of runs under K8Vavoom so I can still check it out later when more engine progress has been made. Now it is going to take half a million year to figure out how ZScript works and taking advance of it so that I can have more complex features. Preferabbly I want to have only 1 selectable tank class, the player should be able to switch to a different tank model when he is certain check points (yet to decide to how its gonna work), so for example when the player made enough score he can switch to Leopard 1 HA (Heavy Armor) model. 
I hope learning a little bit of python will give me a better understanding how ZScript works.

>>15531
Zscript predates the jump to 3D so if you have any particular issues you can do a rollback and stick to that version exclusively. You'll have to prepackage that version's files but that isn't a big deal.

>>15654
My current issue at the moment is figuring out how to adapt the brutal doom tonk code which has also parts for pitch changes which my previous tank code lacked which I didn't have it yet because the version I took was before he added in the pitch changes. Right now its so fucking broken it doesn't work right, the chassis does not correctly pitch according to the sloped surface and I had even bugs where it went absolutely bonkers where it kept spinning indefinitely also the pitch values are sometimes xbox huge which are above 360 which I'm pretty sure that's way above the rotation transform. Here is my shitcode: 
https://pastebin.com/1Vphx7Jx
https://pastebin.com/raE8p151

>>14519
A day late this week but whatever, Last night was more trouble than worth posting that day. No screenshot either because nothing I've done this week is really worth showing.

Weekly Progress Report
-Finally got around to downloading inkscape and testing shading on it. My aim for this week is to draw a character and get a completed art piece out to you guys so I can show "yes, I can do art, and at the barest of minimums have it not be shit".
-Crushed the rest of the unformatted nametags in Nuclear Teagan.
-Crushed unformatted nametags in Glissando Arena.

>>15655
Fugg, I managed to make it a bit less buggy so the tank pitch does change when its in sloped surface and moving over staircase, however I have absolutely no idea how to devise a formula that calculates how strong the pitch difference is from 0 to whatever height difference the collision box actor is to the chassis actor (which when this cbox actor is "active" it gives the player a item that triggers pitch changes), so in brutal doom the tank spawns on every frame 3 actors at the front and at the rear and 2 actors are using different spawning method that snaps the chassis back to pitch 0 (in brutal doom he just used different chassis frame to change pitch). By default the pitch is at 0. 
>I had even bugs where it went absolutely bonkers where it kept spinning indefinitely
It was mostly caused because I didn't managed to write a condition that snaps the pitch back to 0 which also had a calculation bug that I fixed.

output.webm
(313.1KB, 960x642, 00:09)

Screenshot_2020-10-31_at_21.48.31.png
(13.4KB, 292x236)

Yeekr's weekly
>Implemented reaction on dead bodies
>It's now possible to eat pioneeres
Art guy was busy so visuals aren there yet. Shouldn't be a problem though.
Main mechanics are ready. It makes me happy.
I plan on making 1st level by the demo day.

02-11-20-1604342376_scrot.png
(1.2MB, 1920x1080)

02-11-20-1604342378_scrot.png
(490.6KB, 1920x1080)

1f5c47c01e571f91427c230bb5a3c15626a1682ac54aca476a8cb9341aaae3ef.jpg
(36.1KB, 275x369)

Great, this is going to take ages debugging this shit and then modifying all the necessary lines so that it doesn't crash and adding my previous code back to it. That's a lot of work just for the stupid pitch shit.  Fuck OO programming seriously, fucking Graf Zahl enabling this shit.

>>16786
Oh don't worry, you aren't the only one suffering horrible bullshit from GZD latest. I've been trying to run the lastest version of HD in it and it manages to always crash on the second map of any run, on any mappack including doom 1 and 2, every single time. Not to mention the absolutely awful performance the engine has in general.

11-10-20-1602445831_scrot.png
(1.7MB, 1680x1050)

18-10-20-1603011314_scrot.png
(2.9MB, 1920x1080)

18-10-20-1603052162_scrot.png
(2.6MB, 1920x1080)

19-10-20-1603109442_scrot.png
(2.4MB, 1920x1080)

>>16796
I don't know if you are doing any modding, but despite me being a pleb programmer I can already tell how horrible organized the ZScript structure is, the BoA tank code is heavily wired together and I failed completely trying to separate "SetPitchRoll" function since it seems to rely on pretty much all of the other tank functions/objects. If ZScript were sane such as multi-paradigm/procedural programming then it could have been rewritten in a Component/"Has-A" relationship model instead but nope everything is single hierarchy "Is-A" relationship model instead, it is also impossible to extend the gzdoom.pk3 Actor class to add generic functions so that it can be called on any other actors. So in other words I cannot separate BoA tank functions into 2 nice pieces one for the player and one for the monster objects, which is just dumb. It's no wonder according to the ZDoom wiki ZScript literal got its inspiration from Java and Unrealscript both which have horrible programming paradigm and it shows. With Unrealscript it is even worse as you can define only one object per script text. Good lord of all the programming language it exist it has to be fucking Java/Unreal Script, if this is not the definition of idiotism I don't know what else is.  

For the crash thing luckily I managed to fix it by adding a integer cvar at cvarinfo lump which fixed it, but holy shit man this is the worst tank control I have ever seen, who the hell is programming it like this that the chassis rotates automatically depending on the player camera angle as well heavily gimping the player camera, this is just pants on heads retarded. Panzer Elite for example is a tonk sim for autist and it did not gimped the camera, using left and right arrow rotates the tank accordingly, using the mouse is a bit different but it does camera only changes, clicking the "mini tank" buttons, selecting objects and adding markers. 
<but muh diamond problem - Graf Zahl
Dumb idiot, fucking python managed to fix this problem ages ago and it supports multi inheritance just fine. 

> I've been trying to run the lastest version of HD in it and it manages to always crash on the second map of any run, on any mappack including doom 1 and 2, every single time.
Yeah when a mod runs and it crashes then you'll be damned trying to debug this whole shit of what caused in the first place because most of the time it doesn't even tell you which functions did this. Some while ago I played the tranny UT99 mod (which uses OG Unreal 1 model format for some stupid reason) and the Brutal XVME monster pack and almost everytime I try to kill one of the monsters the GZDoom would crash on me because the headshot somehow crashes the ZScript part of UT99 weapons or whatever it was. I also had to make changes to complex doom legendary items so that they don't crash either because the PowerTargeter class causes crashes as how complex doom makes usage of it and on the ZDoom wiki I couldn't find any more info of what changes has been made since Zandronum 3.0. 

>Not to mention the absolutely awful performance the engine has in general.
Yeah I agree 120% here, I played on the Lost Civilization map pack with Complex Doom and my tank mod 
and on several sections mostly on areas where there are a lot of open spaces and large vistas I am getting terrible performance, one of the Lost Civilization map I even had to skip because the framerate was terrible low. This is not funny anymore this is just fucking ridiculous and sad that GZDoom cannot handle basic bitch Boom maps despite I own a fucking RTX 560 Graphics Card which supports Vulkan. On one of Deus Vult II map it lagged badly when several dozens of monsters where shooting projectiles at me and all of them had fancy particles. And guess why it is because of that? Because in 2020 Graf Zahl is literally too dumb to add support for a hardware accelerated particle system something that TA: Spring and UT2004 ages ago have and doesn't causes performance issue, you have to make particle effects with Decorate/ZScript which is the worst way to do it but there is no other options if you want more than single sprites being shown. GZDoom is such a shit show, I wonder if I ever get to see the day where K8vavoom has feature parity with GZDoom or it is going to suffer the same fate that Zandronum does.

>>16806
I haven't done any modding mostly due to personal issues with time schedules, but it is rather insane how obviously flawed GZ is yet how many people programming for it still refuse to fix glaring issues with the structuring of the code. Wasn't Zscript supposed to be C++ or QuakeC but for ZDoom? How did they manage to somehow make an unholy abomination combining the worst aspects of both Java and U.Script? How fucking retarded do you have to be to accept that kind of shit into your sourceport-- oh wait we're talking about the guy who cuts support for older versions of OpenGL but doesn't even use 1% of the capabilities of OGL to lessen performance drops from things like particles, dynamic lights WHICH ARE STILL CLIPPING THROUGH WALLS EVEN TO THIS DAY or object draw distance.

>Debugging
Oh god, don't even get me into that shit. I haven't even tried to run random enemy mods like complex and hard doom lately exactly due to this bullshit, DECORATE was made exactly to be super simplistic, very easy to use and visualize yet somehow they managed to make it into spaguetti hell with every update ZDoom had up to the point nobody knows what the fuck kind of shit they're even doing with it.
It's even worse if you try to run old mods like DRLA, it just straight up breaks entire things.

1447168759717.jpg
(177.5KB, 619x457)

"Great".png
(403.5KB, 1528x1084)

Graf_Dynamic_Light_improvement*_outdoor_maps_not_supported.png
(1MB, 1152x864)

8f28f81697b3670813b8cf048bd6a32a9c6b8961d4868fe70beda76821b9c216.png
(66.3KB, 689x440)

the_graf_community_3.png
(3.4MB, 2447x3683)

>>16811 (checked)
>but it is rather insane how obviously flawed GZ is yet how many people programming for it still refuse to fix glaring issues with the structuring of the code.
People on ZDoom forum are sucking him off any day of the time, any form of criticism against the source port gets you ding dong bannu by Wild Grease yet that same stick in his ass moderator allows himself to criticize other people and project for any flaws he seems fit. There was one random guy who criticized a project for being poorly made and got warning from him yet that same moderator made a yuge list of flaws of other projects and he didn't withheld himself of "strong words", what a hypocrite.  

>Wasn't Zscript supposed to be C++ or QuakeC but for ZDoom? How did they manage to somehow make an unholy abomination combining the worst aspects of both Java and U.Script? How fucking retarded do you have to be to accept that kind of shit into your sourceport
I am asking myself every time I have to deal with ZScript, It also comes with limitations for variables which you are only allowed to have 256 for various kinds of variables and other types and the rest is hard capped to 65535, it doesn't explain if how this limit is shared i.e global, inheritance based or both. Which is low for my taste. 
>worst aspects of both Java and U.Script? 
They clearly haven't seen the worst programming aspects of U.Scripts then, such as actually looking up a average UT2003/2004 mods and looked themselves if its a *really* good idea to copy U.Scripts paradigm in the first place, admittedly my grasp of U.Script is awfully poor too but when I think about that a UT2004 rocket launcher need several thousands line of code for all the functions which all have various task and goals clumped together into one script file and wired together looks pretty niggerlicious too me.  For example why the hell would a weapon have to handle bot related code? That shit belongs to its own function. It's no wonder then that ZScript performs even worse than U.Script
>Wasn't Zscript supposed to be C++ or QuakeC but for ZDoom?
I don't know the whole story behind ZScript, I'm pretty late to get aware that GZDoom I think what was it 1-2 years ago or so? That it has added support for ZScript, but the documentions does refer several times it is also inspired by C/C++ like programming. ACS definitely is more close to C programming which is something I like as it has none of the stupid OO crap, apart from some certain limitations it suffers though, such as save backwards compability which will break on every changes. 

>oh wait we're talking about the guy who cuts support for older versions of OpenGL but doesn't even use 1% of the capabilities of OGL to lessen performance drops from things like particles, dynamic lights
Yep that's Graf Zahl for you, he talks all the damn time that Nvidia is the best shit ever, has the best OpenGL performance, best drivers et cetera and says that AMD is shit for everything and on other posts he mentions he doesn't know shit about OpenGL yet also wants to have all the advances of Vulkan and doing a very bad job at it, good lord this is just terrible. Speaking of deprecated OpenGL support, the guy DrFrag also cuts supports for OpenGL 2.0 on his forked version I couldn't find the reason why but it got stated on the ZDoom news page which makes me wonder why in the first place because he did made several forks for the sole version so that it still works on older computer. Even the K8Vavoom developer later got infected with "Grafism" first he argued with Graf on the Doomworld forum that support for older computer is still worth it and several post later he is on Graf's side now! Toaster users be damned if they want all the shiny Decorate features. 

> I haven't even tried to run random enemy mods like complex and hard doom lately exactly due to this bullshit, 
I'm in the process of extracting the monsters and items from complex doom + all of its addon into its own package so that the weapons doesn't clutter with my weapons set. The modders of it does a bad job of taking advance of inheritance properly which would help them cutting out a lot of duplicate states, their way of implementing monsters is poorly done as well. It is going to be a bitch going to through all the items and monsters and making it slightly less cluttered because of how much I have to wade through it but most of the basics are working already. I implemented complex doom and LCA addon so far, other addons are coming soon then I will try to declutter as much garbage as possible and adding cvar options for more customization, the other motive of mine was also because I would like to deal with only 1 health item instead of having to juggle around 5 different health items it has. Dust and Clusterfuck will also follow suit soon. 

For example my weapon works a bit different, I have 2 weapon base actor one as a total generic class and one as a generic weapon for the tank class, the Multi Purpose Device and the Cannon weapon both inherits the base weapon class and have 0 codes related to the machine gun and very little about periscopes states, so I have to edit only 1 class if I want changes to the coax gun instead of going through all the weapons. I am going to do soon something similar with the particle effects and the projectiles in order to allow for greater projectile flexibility. It sucks though for the APHE/U-APFSDS rounds because I need FastProjectile inheritance in order to increase its speed above 60 because the stupid legendary monsters keep evading my AP shells. 

<LCA Readme
<Note: You CAN NOT use this mod or any previous versions as a base for modification. Any additional addons containing content from LCA must be authorized by me first. 
Heh, I'm going to print this text and throw it to the trash because that's how much value those sort of "license" have when its out in the open. Which is fucking funny considering it has several ID software sprites all heavily edited or lazy color editing which is totally not a infringement to ID Software rights (Or Bethesda/Microsoft for this matter.). 

>DECORATE was made exactly to be super simplistic, very easy to use and visualize yet somehow they managed to make it into spaguetti hell with every update ZDoom had up to the point nobody knows what the fuck kind of shit they're even doing with it.
Exactly, the main intention as it was for its name sake was to allow modders for easier definition of custom decoration object without having to resort to using Dehacked which causes incompatibilities (but still has sprite frames incompatibilities, jej). So instead of making a proper scripting language in the first place which takes advance of DECORATE they just extended it and extended it for years till it is to the state it is today, now it can become even more complicated with the features of anonymous functions but at the same time not allowing regular custom functions which is just stupid, as it would clutter the Actors even more and it is the worst and most terrible form of coding practice which any programmers should know that code re usability is the main priority for a good looking and well maintainable code base. Obviously the word "code reusability" seems to be foreign for them. 

With that all being said it is not a wonder that GZDoom is such a shit fest as it stands today, there is at least 2 incompetent programmers in charge of it that keeps making retarded decision one after another, while picking also the worst programming language for "inspiration" and having even a inferior performance to Boom based source ports. Ages ago on the Doomworld forum a more competent programmer tried to lend his hands on GZDoom so that its less shit but Graf had a sperg fit with him over some issues whose detail sadly escaped me and so this one programmer decided to not work on GZDoom at all. What a sad state of affairs the modern doom modding community is nowadays. There is so many people willing to make assets for it, music, maps, mods, monsters et cetera yet it is held back by several idiotic moderators and retarded programmers that keep making dumb decisions one after another. It is so bad that they are even defending Grafs retarded decision and are going with the flow instead of supporting other source port instead which also allows modding. 
This whole crap is making me depressed, I'm hard pressed to continue making progress with my mod knowing well the source port is not steering in good direction held back by idiotic captains.

>>16814
>This whole crap is making me depressed, I'm hard pressed to continue making progress with my mod knowing well the source port is not steering in good direction held back by idiotic captains.
Obviously the choice is to just move onto Vavoom, but it is not an easy choice due to the modding features that you lack in sourceports that just aren't modern ZDoom. It's a sad state of affairs mostly due to the cult of personality such figures have adquired and I believe, had ZDoom's main developer not become a tranny and subsequently quit developing the main fork of the sourceport leaving only the german hack autist in charge, things would be much better nowadays with the possibility of alternate rendering sourceports from the main ZDoom base not being subject to this faggotry.

Please do not quit working on you mod, the work you've put into it has been humongous and it'd just bring dissapointment not only for others but to yourself if you'd just quit. Graf is a piece of shit and so is his sourceport but that doesn't mean you can't make good games out of it, much like how other modders have done before and will do after. Find resolution into making a good game that you enjoy playing and vent your frustration by finishing up the features you've always wanted to see and interact with ingame. Porting or remaking into a better engine can come latter if that is the issue, just don't give up.

>>16816
>Obviously the choice is to just move onto Vavoom, but it is not an easy choice due to the modding features that you lack in sourceports that just aren't modern ZDoom.
Yeah that sucks that Vavoom is lacking tons of features now and that the K8Vavoom dev started out way too late which would take him years to catch up, that is if there isn't other major changes being made to Decorate/ZScript which requires additional changes. I think I'm going to adapt and fix the BoA tank script first and hopefully get a playable demo on next demoday which is the most realistic goal at the moment, also for the steering stuff I forget to set the cvar to 0 which disable this retarded behavior and it seems to drive like a proper tank, now I just need to fix the model rendering somehow as I can't make quite clear which part is responsible for rendering the model, then I also need to fix the camera stuff, adding back the weapons and I think the majority of the features that I want should be there then. Code cleanup comes later but it is going to be quite difficult as I can barely understand how any of that stuff works, Brutal Doom code is a walk in the park in comparison. I also hope the "Morph to Marine" script will still work as it is a essential feature of my mod because tons of doom maps aren't designed for micro tank and it is a gargantuan task doing so.
Actually I think I could redo my mod from scratch and slowly build pieces by pieces back into and see which one requires VavoomC scripting and which one doesn't, but also I need to familiarize with VavoomC first before I can do any major changes. Hopefully it does perform better than ZScript. 

> It's a sad state of affairs mostly due to the cult of personality such figures have adquired and I believe, had ZDoom's main developer not become a tranny and subsequently quit developing the main fork of the sourceport leaving only the german hack autist in charge, things would be much better nowadays with the possibility of alternate rendering sourceports from the main ZDoom base not being subject to this faggotry.
I agree, something like a resurrection of GLZDoom would have been great, maybe it would have turned out several times better than GZDoom and people would flock into that source port instead, leaving Graf Zahl nothing but the most idiotic lemming on his side, then maybe he will fuck off from Doom community for good. But fate is not on our side and with their safespace fortress makes it impossible to penetrate and persuade their opinion and shining light into the darkness, enlightening any people willing to see into it and realize what a false messiah that Graf is, instead they are following the dark path of eternal hardware hamster wheel, preferring to give money to the Nvidia CEO so that he can afford himself another set of leather jacket, at this rate Graf might as well advocate for the Nvidia ray tracing GPU because it has better BSP performance or whatever nonsense he is going to come up with. 

>Please do not quit working on you mod, the work you've put into it has been humongous and it'd just bring dissapointment not only for others but to yourself if you'd just quit. Graf is a piece of shit and so is his sourceport but that doesn't mean you can't make good games out of it, much like how other modders have done before and will do after. Find resolution into making a good game that you enjoy playing and vent your frustration by finishing up the features you've always wanted to see and interact with ingame. 
Out of all the projects I have worked on this is the only project that I have at the moment that shows any promise of it might actually fulfill its goal. A few months ago it was on hiatus because I was busy learning python on the meanwhile so that I understand more of programming concept which it did paid off, because I finally managed to delegate the "ammo display" part from SBARINFO to ACS part which allows me greater flexibility, so that I need to change only a array variable in order to add support for other weapons instead of adding another tons of block just to display a damn icon and the current ammo count. There is still plenty of features I want to add in order to make the core mod more complete, for example I would like to program a different weapon slot system so that its less cumbersome when I have for example 30 weapons added in which is a big issues in any FPS game with weapons mods, so some kind of Deus Ex / Stalker inventory system. Though the main priority is that I want to have complex weapon features, not weapons quantity so more akin to UT2004 ballistic weapons or more closer Gunman Chronicles which has some cool weapon concepts. 
For example If I would program in a Howitzer cannon it should be possible to have different amount of propellant charges being loaded it, and on certain ammo types even when it should detonate. It is still a concept heavily in development because I need to take care of other things first, at the moment what is taking most of my time is cleaning up the code base in such a way that it allows for the most extensive expansion possible to add new and complex features in a relative sane way. I don't want my mod end up becoming a "complex doom or brutal doom but in tanks form" edition which is too shallow for me. 

>Porting or remaking into a better engine can come latter if that is the issue, just don't give up.
Yeah in a way you are right, I think like what 8 or 10 months ago I lost count I tried making simple 2D game in Godot and even fiddled around with Pygame but I quickly lost the motivation for it because the documentation in Godot case is terrible compared to ZDoom documentation and with Pygame I would have to add in so many boilerplate codes that it is not enjoyable for me anymore. Godots documentation page even looks like it is made for mobile phone users for fuck sake! Why the hell would one design it like that instead of copying Wikipedia software which is superior than whatever the fuck "read the docs" one uses. There is no other good open source game/engine either which I can use at the moment, so GZDoom/K8Vavoom is the only realistic option if I want to finish any project at all, it sucks for me but thats my reality I guess. I should have learned programming instead of fucking around with Wings 3D back then during TA: Spring heydays but then I didn't lurked in any imageboard at that time either so what do I know.

03-11-20-1604419763_scrot.png
(959.9KB, 1920x1080)

03-11-20-1604419788_scrot.png
(992.1KB, 1920x1080)

Okay I managed to do some minor progress with the damn code, first I managed to isolate the code that was responsible for fucking up for both perspective views so it is restored to normal camera behavior. Secondly I also need to find out why the turret height offset is immensely large as there isn't a obvious property something like called as Tankbase.TurretOffsetHeight; . The code really is complicated written which makes it for me quite hard to "study" how it works, I'm not certain if I will be able to manage to push a demo version on the upcoming demo day next week. I won't be using their camera related changes at all because it is just stupidly made, however for some reason the actor "SecurityCamera" is tied with the turret and the chassis and when I omit that part then the chassis and the turret doesn't move to the players position at all. What is also interesting to note is that it doesn't uses A_Warp() function at all. I figured also out that I had to add "TankTreadsBase.Length" to the US_Sherman object which seems to fix the pitch calculation, so the tank angles correctly most of the time when the surface is sloped, not great not terrible.  
Hopefully I can manage to get rid of all the unnecessary crap and isolate the code more so that it works more on generic base instead of being tightly coupled together. Also I need obviously a method to hide the turret model when the player is in first person view, because of the HUD weapon models.

leejy5.mp4
(13.2MB, 1920x1008, 00:41)

snake making the level real pretty

>>15839
>"My aim for this week is to draw a character and get a completed art piece out to you guys so I can show 'yes, I can do art, and at the barest of minimums have it not be shit'."
Yeah, that didn't happen at all. So I'm going to hang around and contribute to drawthreads in other boards for a while. I will put my main gamedev focus onto things I can do right now. (music/writing/implementation)

So here's something I actually did work on: a demo of Dischord, the main villain's theme.

Weekly Progress Report
-Played around with LMMS, worked on main and battle theme for Dischord. File related is the current demo of his main theme.
-Did some raw writing at various points in act 2/3, mostly for one of the two bad ends of the game.

>>17135
Also, can someone please tell Prolikewoah and and Anon.cafe to fix their fucking captchas? I can't even post because their broken-ass shit is stuck in a "reloading" loop.

>>17141
just post on the superior website then, aka here

behold_the_power_of_my_master.png
(1008.9KB, 960x640)

Due to a lack of ideas I have ascended to enginedevhood.
See you all in 60 years.

>>17186
I have more ideas than I could make use of in a lifetime but I still ended up enginedeving.

05-11-20-1604581571_scrot.png
(1.6MB, 1920x1080)

05-11-20-1604581626_scrot.png
(1.5MB, 1920x1080)

>>16932
I'm starting to get desperate, I have absolutely no idea why this shit happens in the first place, apparently there is something wrong with how "SetPitchRoll" function calculates the damn number and I don't know what lines does this. Fucking hell this is not good. Trying to adapt the brutal dood tonk code failed hard and blades of agony tonk code fails miserably as well, when I'm close to a edge the chassis spazzes hard. I tried looking if there is other code I could jank instead and there isn't any so I'm pretty much stuck with those 2 code types I can use.

>>17135
Very nice music anon! Reminds me of tracker MODs from the 90s, did you happen to have experience with that?

>>17186
Nothing wrong with that. Make something cool and share it with us!

>>17190
Why not implement your ideas now using an existing engine, then write your own engine(s) later? Seems like a waste of ideas if you're not doing anything with them.

>>17195
I had a brief (less than an hour) fling with famitracker, that's it. But I grew up loving chiptune thanks to playing a shitton of NES games. I've developed a liking for MOD tracks later in life, and one of my backburner goals is to try and learn an OpenMPT-style program.

>>6917 (OP) 

>>17195
>Why not implement your ideas now using an existing engine
Because then I'll end up as an eternal unity dev and never know the joy of developing games on my own engine.

bored_aliens_hicks_sleep.jpg
(33.2KB, 718x458)

>>17194
Fuck it, I guess there is no other options left than making a damn account on the fucking forum and asking there for help. Hopefully I don't get ding dong bannu when one of those stalkers found out that I worked on the moonman mod or any other non kosher shit I have done previously, since my post got approved I need to now wait 6 million years for a reply. Thanks for joining my ted talk. 
>>17186
I have at least 3-4 game ideas I'd be willing to do but can't really even execute it to the nearest due to lack my of programming experience, welp maybe when I'm done with this damn mod I should just learn fucking Unity till I git gud enough with gaems programming and then jumping ship to Godot for seriously! gaems development, since all the cool kids are using Unity nowadays, it still pisses me off though that out of all the engines to exist it has to be one of those proprietary ones that gets popular and has tons of tutorials and probably gud documentation.  Godot documentation is such a mess to look at, while ZDoom documentation is much more organized. Just alone the thought of having to touch C# is making me anxious. 
>>17135
>muzyka
At the beginning I almost thought it was a Tiberian Sun/Twisted Insurrection muzyk.

>>17375
I have developed a few engines myself and made games with them, then switched to unity to make more games, now I'm switching back to making my own engine... Don't be afraid to try different things, it's always a learning experience. Besides, it's better to implement your ideas with a premade engine then revisit them later with your own engine, than not implement them at all.
Mind you this is coming from someone who absolutely hates unity, but can't deny its power in rapid prototyping.

>>17385
If I was making anything other than a roguelike I know for certain doesn't need to be optimized all that well, I wouldn't be using Unity either.  I'm halfway using Unity as a learning experience too, since I wouldn't call myself an amazing coder.

what are some methods I could use for checking if a player is touching the ground in a 2D platformer?

>>17393
What are you using?

>>17416
stencyl

ClipboardImage.png
(460.2KB, 1017x574)

How do you punish guessing in a puzzle game? Since there is a finite combination of commands for each level a player could in theory just randomly guess until they get it right on every level, which makes them feel really, really bad. Any suggestions?

Have too many possible wrong answers for the player to brute force it.

>>17450
>>17446

>>17446
Completely randomize the levels, either the order they appear in or the design of the levels themselves, and if they make one mistake flip to an entirely different one. Alternatively maybe give them something if they DON'T make mistakes the first time they do a stage and then take it away once they do, but odds are they may reset their progress entirely just to get those rewards even after fucking up repeatedly. Ultimately, i'd personally say you should just let people brute force your puzzles if they want to since smacking hands away just because they're playing your game wrong is pretentious as fuck.

ClipboardImage.png
(311.1KB, 720x404)

>>17450
Yeah this is the "easiest" one but you pay the price of having very large and complex levels generally. I think it might be fine though.

>>17452
>smacking hands away just because they're playing your game wrong is pretentious as fuck
It's not really about smacking hands or being pretentious, it's more that in the playtests I've done so far when players managed to just guess the solution they get zero satisfaction out of the puzzle, just feel cheated and get angry with the game, even when it was their own fault for just randoming out commands.

>>17446
Have a timer or a score with multiplier and penalize the player on either of these respective scales?
You could also go the other way and make "combos" very satisfying with audio or visual feedback.
I think tetris effect did this.

>>17457
It would actually be pretty easy to keep score by just counting the number of ticks a player goes through in the map. I guess you could then compare them to the "par" time for a rating?

>>17460
Yeah pretty much, most games that do it this way provide like 3 dev made times - think bronze, silver and gold time and maybe a chime and a little "victory" screen.
In my opinion it would be most engaging if you reward the player with a multiplier on their score for successive correct actions but that multiplier goes back to zero over time giving a cost-benefit angle to the player. But that of course depends on the game speed you intend to have.

>>17446
I remember one puzzle game that sorta disguised itself as an RPG. It didn't change the solution to a puzzle much because the puzzles would be randomly generated, but the RPG aspects such as armor and HP allowed more mistakes. It was some F2P shit on mobile so it was most likely to sell energy or some shit.

>>17460
If you do that, what's to stop players from just going back and doing the solution again to get under the par time?

>>17446
The point of a puzzle game is to figure out the puzzle, not to get to the end of the game by any means. If you have playtesters who are just brute forcing puzzles, isn't it possible you simply have playtesters who don't actually like puzzle games? Now, if it's much easier to brute force a puzzle than to figure it out, and if many different players opt to force them rather than puzzle them out, then maybe you have an actual design problem. I mean, look at any basic puzzle like one of those number slider puzzles. There are people who figure them out, and there are people who get frustrated and just slide at random until they win. You may not be able to force the latter type of player to enjoy a slider puzzle by any means.

>>17453
>you pay the price of having very large and complex levels generally
Not necessarily. If you withhold feedback until the player input the entire solution, you make bruteforcing brutal with very little effort. Consider guessing an N-bit sequence where you're told that the solution is wrong (a) as soon as you input a wrong bit (b) only at the end. The first one takes at most 2N guesses, the latter takes up to 2^N.

>>17446
What kind of game is it? What are the puzzle rules? Maybe it's too hard or uninteresting to do it the right way, or maybe you're not teaching the mechanics slowly enough. >>17453 seems way too hard for the first level. You should start much simpler, with the absolute bare minimum, just one piece that needs to be placed down. Even if the level takes 10 seconds to complete, it's fine, the game is better from those 10 seconds because it hammers home the mechanics and what you're here to do, you're turning the mechanic into intuition rather than something you're suddenly juggling along with several other things. Make the next level be an alteration of the previous, so the amount of things the player has to figure out is less, and they can guess what you're after by the fact that it's in the same "set" of puzzles.

I don't think you should put a penalty for failing a level, you're just going to make people hate your game. I see the issue of bruteforcing as your game not being good. The players are bruteforcing because they don't know what they're supposed to do or have interest in thinking about your puzzle.

Say what you want about Jonathan Blow or his games, but I think a lot of his ideas on puzzle game design are really good. He thinks that making puzzles "hard" doesn't make the puzzles or the game better. Instead each puzzle should have some new idea or obstacle in it, for example you want to go to the exit, but the obvious way is blocked and you must figure out how to go around it, or there's a piece that you must put into the right place first to kick off of, or there's a thin gap that's hard to cross correctly, or you must move multiple pieces at the same time without messing it up. There can be many levels in the same set, with similar obstacle used in different ways so the same solution doesn't work exactly. It's not just some random puzzle, there's clearly a certain goal or something you must figure out. If the puzzle is just hard and you're repeating the same shit you've already done before without any new ideas or obstacles, then it's much less likely to entice people to actually think about the puzzle.

Also as >>17549 said, make sure your testers aren't some normalfag call of doody gamers, if they're not the kind of people who would play a puzzle game on their own, they're not going to play yours properly either.

>>17460
That's what Zachtronics games do, trying to optimize your puzzles after completing them is part of the game. If you see that people have done it better, then you know there's a better way to do it than the way you did it.

ClipboardImage.png
(441.4KB, 1017x577)

ClipboardImage.png
(445.9KB, 1018x571)

ClipboardImage.png
(444.8KB, 1014x567)

>>17549
>playtesters who don't actually like puzzle games
I think this is at least one part of the problem.

>>17565
>withhold feedback until the player input the entire solution
Don't you lose a big part of the visual element then? In our game when you input a sequence of commands the robot will start playing them out and the idea is that you can incrementally add commands when you encounter a problem.

>>17570
>What kind of game is it? What are the puzzle rules?
It's a programming game where you have a few available commands per level and a max number of commands that you can put in your queue. The queue loops until you reach the goal. The level in that screenshot  is pretty late in the game, early maps and tutorials look more like pic related.

It's interesting that you mention Blowfish because The Witness actually has failure in his puzzles, if you try to brute force some of the puzzles it will send you back to the previous one. Incidentially, it's considered one of the worst parts of the game and make me decide fully to not implement failure in our game.

>>17542
Nothing, I think it's perfectly fine that they do it. It wouldn't be interesting to use to compete with others.

Sorry if I missed any feedback, thanks everyone for your ideas.

>>17576
>It's a programming game where you have a few available commands per level and a max number of commands that you can put in your queue
That doesn't sound like a game that people would start bruteforcing. It looks kind of interesting to me, makes me think your testers are the main problem.

I've heard about how the visuals of the game can make people care more or less. If the game looks good, people will have more motivation towards it and are willing to take it more seriously. It's probably not the most important issues, but I'm noticing your levels look very monotonous and similar to one another, and the orange levels are just painful to look at because it's so bright and saturated and the objects blend into the floor. Of course that's just an issue with the game being unfinished, but it could have an effect on the testers, like they go to the next level and it looks almost exactly like the previous and it just feels tiresome. Are you planning to add different tile sets and backgrounds?

>if you try to brute force some of the puzzles it will send you back to the previous one
I hated them too. Though I only remember that existing in small sequences and tutorials, for example the audio puzzle tutorial, which seemed more about making sure you understand the point of the puzzle.

ClipboardImage.png
(453.1KB, 1027x578)

ClipboardImage.png
(452.4KB, 1017x573)

ClipboardImage.png
(453.3KB, 1015x574)

>>17582
Yeah we have 4 different tilesets but I think at least 2 of them need some tweaking in terms of colour. Only one background though, not sure if we could make some more, it's slightly randomized but it doesn't change much.

The pink set is supposde to be turned into a nighttime city tileset so we're making it more purple and adding props that are street lights and apartment buildings. The brown/grey one is desert so we're going to add cacti and similar.

ClipboardImage.png
(226.9KB, 1183x552)

>>17584
Forgot to mention that it's split into worlds, so each world has its own tileset and adds more mechanics. This is the "meta" level select where you unlock more worlds and select which one you want to play.

>>17584
I meant more in the sense of texture. Like a metal tile, a decorated temple tile, wavy snow/sand patterns. Also variations of tiles in a single level, for example stone tiles mixed with cracked/mossy stone tiles. If you don't want so much art effort you could add slightly different hues among the tiles so they're not all the exact same color.

varying.jpg
(90KB, 1017x577)

>add slightly different hues among the tiles so they're not all the exact same color
Forgot my image regarding this.

>>17587
>>17588
That's not a bad idea, I'll bring it up with the artist.

ClipboardImage.png
(131.5KB, 826x511)

ClipboardImage.png
(56.8KB, 343x776)

Double post but I think we found the primary culprit to why guessing is a problem. We made a tool that simulates all possible solutions and many maps have way to many which makes guessing so easy.

>>17446
Do you have a demo for your game somewhere?

>>17685
Yeah you can go play it here: https://jaoel.itch.io/asar-post-jam-version

>>17689
so we'll have at least one demo day entry, lol

>>17689
I tried it and I already saw the problem on the second level.

The problem is that it very quickly becomes very hard to think through the puzzle in a logical way if the path is anything but straightforward. When you have to make some repeating pattern that runs into walls along a long path, it's very hard to plan out a movement pattern through that path.

It might just be a problem fundamentally with the game idea itself because I can't think of a way to fix it without some kind of huge change like making the map 3 dimensional or something. I think making the path shorter but more complicated/specific could be good, for example requiring the robot to step on a button or push a block along the way.

You could also make the program queue be very long or even infinite, and then do the zachtronics thing where after the level you tell people that their solution was shit. Add bonus levels that can only be accessed if you get a good enough rank on X number of levels. That way it's easier to come up with a solution and more people can complete it, but you'd have to try harder to 100% the game.

>>17912
Yeah, that issue persists in the early levels mostly because people seemed to struggle with the concept that you can't drop off tiles. Once you get a bit further into the game it should get better, especially once  you hit world 2.

stalker.jpg
(258KB, 1920x1080)

>>17384
>I made the threda 5 days ago and it's at 51 views now
>not a single god damn reply
>demoday will come at 2 days
>started working on the Leopard 1 (heavy armor) model yesterday
>didn't finished the turret yet or applied applique armor to it
Man, sector slopes haven't been invented yesterday (meaning it existed ages ago while ZDoom was maintained) and there is still no practical solution to this fucking pitch problem. This is just getting tiresome.

>>18015
>not a single god damn reply
You making the tankmod? I've been watching your progress from afar and your struggles with GZDoom being both one of the more advanced doom engine ports with one of the most assbackwards, self-absorbed inflated ego main developer. I like what you're attempting to do, to turn Doom into a vehicle combat game featuring tanks with turrets that work and like many other silent supporters I would like to see you succeed in your work but aren't sure on how to best vocalize this sentiment. Incidentally are you the same anon that worked on DoomSS13?

the_green_OpenGL.png
(7.9KB, 641x562)

>Graf_Zahl_post.jpeg
(40KB, 800x539)

>>18033 (checked) 
>I've been watching your progress from afar and your struggles with GZDoom being both one of the more advanced doom engine ports with one of the most assbackwards, self-absorbed inflated ego main developer.
Yes yes, it is absolutely terrible. I don't even bother reading GZDoom changelogs because I instinctively know it will bring nothing good to the table wherever they introduce new features. The difference between GZDoom 4.2.0 and 4.4.2 is negligible and even more so from Zandronum 3.0 apart from newer Decorate calls and ZScript support. For over 7 god damn years I haven't seen any fucking mod for GZDoom that takes advance of its piss poor shader support or anything with the graphics that Graf Zahl constantly like a parrot is harping on of what GZDoom is supposedly makes it so "modern".

In fact you are not even allowed to bring up the topic for toaster compability in their forum anymore, that's how much freedom of speech they value in their stupid forum which has even stricter rules than TA: Spring does. This is how a fucking SJW Communist like Wild Greasel, Rachael and Graf is infiltrating a community and actively sabotaging it. Instead of making decent progress they are in truth regressing. How is a mandatory support for OpenGL 3.0 acceptable while having piss poor performance on Lost Civilization wad? This is absolute a embarrassing, shameful disgrace and a big insult for any user that owns a toaster. 
Doomsday engine for example handles graphics better which has support for several texturing methods such as bump mapping and that with OpenGL 2.0. It pisses me off to no end that absolutely nobody in the ZDoom/Doomworld forum challenges Graf monumental incredible display of sheer incompetency that would have you kicked your ass in any damn company that takes it matter seriously. 

I will try later to bump my ZDoom thread again which will mention the Blades of Agony adapted code which I cannot fully use it yet because there 2 cases which totally fucks up the chassis orientation such as when the player is on a edge or near a pit which makes the tank look like it is about to fall down, I previously didn't because I was hoping for a solution without having to use BoA tank code. 

>I like what you're attempting to do, to turn Doom into a vehicle combat game featuring tanks with turrets that work and like many other silent supporters I would like to see you succeed in your work but aren't sure on how to best vocalize this sentiment. 
Thanks I appreciate it. It is mainly my desire for a lack of good arcade tank action games out there that is not heavy on PC specs, does not require internet connection and is relatively open source. Currently 3 tank arcade games dominate the market and well just ask the other anons next thread of what they think about it. 
>with turrets that work
Ironically it was with the help of Brutal Doom code that finally let me manage to have it and I improved it further with A_Warp() call to make the code much easier to work with. Though on the flipside I would have to add support for a casemate tank which needs different code if it were to have a tank like driving characteristic. 

>Incidentally are you the same anon that worked on DoomSS13?
No, I never played Space Station 13. But I do remember him that he posted a few times at 8chins back then, its nice to see that he managed to create a black hole that destroys sectors (?). After the fall of 8chins I haven't heard of him since. My other projects are these:
Panzerstahl, Spring RTS (abandoned, reason: shit pathfinder code)
ARWM mod, Factorio (abandoned, expansion of combat items is not worth the effort with no RTS like element)
Estonia Reich mod, OpenRA(abandoned, OpenRA dev said it is impossible to have larger resolution at the same scale)

>>18042
>Black hole that destroys sectors
IIRC, it should be relatively simple to make an ACS check that, when X object decides to move to Y sector, said sector changes properties to damaging, texture change, gravity, color and other things. SS13 works the way it does in BYOND because every square can also be handled almost the same as every sector can in Doom, just make the map into a bunch of squares and it becomes fairly simple to emulate SS13 behaviour, minus actual atmospherics mechanics that is.
The only issue that props up is how taxing it is for ZDoom to actually render each and every sector visible, and in GZDoom every sector in your FoV is technically visible and rendering for some bizarre reason, regardless if it's blocked by another sector that masquerades as a wall, unless said wall leads into void space.

multi_projectile_hit-fullpage.png
(326.9KB, 1821x1955)

>>18042
One more thing I forget to mention about Graf Zahl incompetency, he literally doesn't know that its possible with pure Decorate(!) code that works even on Zandronum 3.0 for a multi target hit capabilities, the same thing can also be applied for hitscan bullets which I did for the Ben Garrison rifle in moonman mod, so I hinted that out on the other thread of a user that asked for this feature. Can you believe this? A sourceport developer that worked much longer on it then I do doesn't know about this simple trick, no ZScript fuckery is needed. 

>>18049
>The only issue that props up is how taxing it is for ZDoom to actually render each and every sector visible, and in GZDoom every sector in your FoV is technically visible and rendering for some bizarre reason, regardless if it's blocked by another sector that masquerades as a wall, unless said wall leads into void space.
Yeah its not a wonder that GZDoom way of rendering things is absolutely backwards and idiotic. If I recall it correctly on the Joy of Painting wad the one which has a complex winter themed house hub map the performance regressed as he changed the way BSP rendering works, on older version the performance penalty was slightly lower. His Vulkan rendering code barely improved it. Speaking of sectors, it makes me wonder ZDoom/GZDoom did not added support for Quake 1+ BSP maps instead of tackling on features of its current sector system in order to allow for more 3D like mapping with things like portals, stacked sectors and 3D floors, which all of these 3 and especially the portal method is very performance demanding, instead of wasting time tackling on features that makes 2.5D mapping even more tedious whose solution feels more like a shoddily done hack instead of much more pleasent mapping experience that new features should provide, such as Cube 2 Sauerbraten or Trenchbroom. Will it take longer to add support for 3D maps? Of course it will but the effort will be worth it, since then users could port any quake derived games maps back to GZDoom  to have both incredible amount of modding and mapping features set.

>>18015
>demoday will come at 2 days
Well, there's always next time I think?

Anyway, after so much rant, can I ask you why are you sticking to GZDoom? Because it feels like even restarting from scratch on a saner engine would be easier than to do anything in this crap. I'm just asking as someone who has no idea about your project, sorry if I'm my usual utterly ignorant self

do_it_like_a_white_man.mp4
(409.3KB, 640x368, 00:09)

Marcus.jpeg
(50.9KB, 850x400)

>>18056
>Well, there's always next time I think?
Well I rarely participated in demo days previously, I only did it once and for this time I would like not to miss it again due to my depressed state.

>Anyway, after so much rant, can I ask you why are you sticking to GZDoom? Because it feels like even restarting from scratch on a saner engine would be easier than to do anything in this crap.
I addressed this issue here: >>16823 . The main roadblock is my lack of game programming experience and almost no knowledge of mathematics in order to do stuff like projectile spawning at correct velocity/vector. I tried the 2D tonk example for godot and well its methodology is absolutely different than what I am familiar with in GZDoom. If anything I need a game engine whose scripting language is very easy to grasp and requires for example only one or two lines of code to spawn a projectile and among other things, so basically a python version of a game engine in terms of learning curve. 
The next issue is there is so many wads for the Boom source port which also many of these are clocking in 32 maps such as Epic I+II, Memento Mori I+II, Eviternity, Hellbound et cetera that it would be a impossible inhumane task for me to offer the same maps variety, as building up a loyal indie fanbase is extremely difficult that does the additional maps for my future game, take a look here https://onemandoom.blogspot.com/p/index-of-reviews.html of how many maps there is for Doom!. 

So in other words I need to first familiarize myself how game programming works and I need a suitable engine for that, good FOSS game engines with active community are very rare (if there is even one), nowadays when somebody wants to make a game they get recommended either Unity or Unreal Engine 4 (later 5?) which makes them tremendously popular and thus has tons of tutorials that FOSS engines desperately needs. Then the next issue is the language, there is almost no "gold standard" of what programming language one need to know in order to use the majority of the game engines as each of these utilize programming languages differently, one supports C#, the other uses instead Javascript, then there is that one that has support for a obscure programming language the other one has its own programming language with its own syntax et cetera it is one gigantic "Not Invented Here" issue. Since I familiarized myself with Python and if I were lets say want to use Unity then this Python knowledge is pretty much useless for me as then I would have to learn how C# itself works, its syntax, how it handles the variables, how functions are called and so on, before I get to use Unity itself. Then when I decided that Unity is absolutely shit for my purpose I would have to look for another engine again and this whole drama starts from again. 

Lets assume this case:
I know C# and I am familiar with Unity.
If I were to port Microtank for all its intent and original purpose then I would have to write some kind of a parser function that understands Decorate and ZScript code, then effectively I would have to copy tons of GZDoom features set and replicate them with  1:1 or 1:2 accuracy as the main intention of this mod is also that is compatible with any Doom wads. So that means I would have to spend a lot of time to write just for the parser alone and it will be much more complex if ZScript support is desired. 

Then I would have to familiarize myself how the BSP code works and how it translates to Unity map format of whatever the fuck it is, then I would also have to write thousands upon thousands of code to make the maps work 1:1 as they were in PrBoom+ or in GZDoom because it has a lot of features that allows the mapper to control the map flow and its design. Writing the portal and 3D floor support code is going to be a massive pain in the ass and that is just the tip of the iceberg. I would have also to deal with a lot of edge cases that certain maps previously relied upon in order to make them work which ZDoom/GZDoom already does with its map compability flags.

One of the famous cases would be the "Dead Simple" map it relies on a behavior where when 4 mancubuses are killed 4 sector doors are opening up, which is one of the many source of bugs that most mod author failed to fix it because they did not used randomspawner actor to replace the monster. 

So for this part alone I would have spend for half of my valuable life time just working on features to essentially port GZDoom into Unity and I would have to fix both Unity and GZDoom expected behavior and bugs to make all of that thing working. In short, porting Microtank to Unity will be a herculean task. For the maps, then since the average players are idiots I cannot except them to use a command line utility to convert the doom maps to Unity/Microtank/GZDoom hybrid kind of a game, so I also have to spend effort and time to get that part working in a user friendly manner. So as a one man developer that where the average player is excepting from to do fucking everything from writing code, fixing bugs, adding features, adding models, sounds, textures and all other damn content preferably all 100% from scratch, 0% stolen content or copyright infringement it is truly a awfully gargantuan and immense task that is absolutely not humanely possible. And why as a one man developer you may ask? Simply because a one man developer is the only reliable developer as developers compromised with other people and skill sets doing a mod and/or game from scratch without any monetary compensation are extremely rare and if they do show up there is a high risk that this project gets abandoned and never to be heard of again and it obviously it did happened several before and it will happen now and in the future.  

That is the very fundamental core issue of porting Microtank to any other engine. Time is one of the most underestimated resource to ever exist, and many players simply do not know the difficulties of creating mods and games in a manner to make it enjoyable for a certain audience. As a one man developer compromises between code, engine and asset creations have to be made, there is absolutely no way around this. If I were to use a different engine then I might as well make a different game with different concept which will be no more relatable to Microtank spare that it features tank combat. 

Of course a average anon such as this place or any other image boards might have different worldview when it comes to mods and games, this is the least concern and of mine. I frankly do not care much either what the average player thinks of my mod, so far on my moddb page I haven't received any feedback of what changes of what I should do to my mod or regards to this content. But for the average people that have absolutely no trouble regarding with freedom of speech that are browsing TA: Spring and the ZDoom forum things are different here. If I were to announce my mod here then they will rip me apart like masses of crows tears away a corpses to delight themselves upon its flesh just for the sole fact that I rarely tend my credits.txt file, rarely care about content attribution and that I do use assets being used from other mods and games. Those kind of people except ME to do everything 100% from scratch, nothing "stolen" (a digital content cannot be stolen.)  in order to build a mod and investing tremendously amount of time just to become accepted in their particular brand of circle jerking. 
Those same people except me to become the mythological egg laying wool milk and meat providing pig creature (eierlegende Wollmilchsau, German term for a jack of all trades) in order to not get banned from their community. While at the same time those people are hypocrites themselves and are using assets from other games in order to build their mod, just look how many countless dukenukem, unreal, even half life, wolfenstein based mods there are and the mods give them all leeway because they are part of the same circlejerkers, those same people that shitted on Sgt. Mark IV all of the day because he didn't tended his credits file properly, what a bunch of hypocrites and liars they are!

>>17689
I like your game so far.  I don't think the brute force method is really all to effective anyways, and I got further by just thinking through the problems, but having just one or two true solutions would likely help you with that anyways.  I like how you use the same control method to pick the levels as you do to play them, really forces the player to get used to the controls and I dig that, especially since it removes a menu selection.

As far as criticisms are concerned, I think you have a mild issue with conveyance.  It's difficult to tell what exactly you can or can't do sometimes (like what you mentioned about how you can't fall off tiles, though I did figure that out myself).  Often times, I got stuck because I couldn't actually see part of the level was traversable.  In addition, I'd say that your game is lacking some "juice", as they say.  It's certainly fun to play, but it's lacking feedback, imo, and the generic star you get when you win is pretty meh.  I'd recommend trying to pump that up a bit as your game moves forward.  Overall, though, as I said, I dig it.

ClipboardImage.png
(375.1KB, 720x622)

>>18082
Thanks for the feedback anon, it's really useful and I appreciate it! It's interesting that you bring up the issue with conveyance because I haven't really thought about it much but you're absolutely right.

The game is extremely stiff right now because we have no animations. I don't know how we will solve it because neither of us in the team can really animate but I'm sure we can work something out, pretty sure we can juice it up with tweens/particles/more props. Currently reworking all maps to minimize the number of possible solutions, all but world 4 is done and I hope to finish that as well before demo day.

>>18066
>languages
I think you're overmystifying things. Unless you're looking at functional languages or esoteric languages, they're pretty similar, if you know one, you can learn to do basic stuff in the other in a few hours.
>FOSS 3D game engines
Yeah, they're a bit lacking. But at least unreal has an advantage over unity that you get access to the full source. I'm not sure about unreal, but unity's asset store is full of "code" assets, you can wire together a full game with minimal coding knowledge. Not that I would recommend going this way, but it can be useful if you don't want/can't write everything from scratch.

So, the TL;DR version of your post is that you want to be able to use doom wads, and it's a bit difficult to do it without a doom engine? Yeah, that sounds right. What am I saying is that it might be easier to just drop wad compatibility, and create a game with some levels (these can be ports, creating an importer that works 90-95% of the time then fixing up the remaining by hand is much easier that creating a 100% one), just to avoid dealing with those shitheads, but it's just my $0.02.

>>18130
>I think you're overmystifying things. Unless you're looking at functional languages or esoteric languages, they're pretty similar, if you know one, you can learn to do basic stuff in the other in a few hours.
It is going to be a problem when I'm the phase of having to test engines how well they can handle ported wads such as okuplok which has 18,532 monsters and that requires me to write quite a large amount of code for testing the performance. PrBoom handles any slaughterwads that are made just fine, admittedly GZDoom wouldn't be to handle the extremities of a slaughterwad, it is still possible to play slaughterwads relatively fine but not to the scale of what I mentioned previously. 

>Yeah, they're a bit lacking. But at least unreal has an advantage over unity that you get access to the full source. I'm not sure about unreal, but unity's asset store is full of "code" assets, you can wire together a full game with minimal coding knowledge. Not that I would recommend going this way, but it can be useful if you don't want/can't write everything from scratch.
Then I guess it would be wise to pick a engine that has a very low learning curve to git gud at gaems programming, instead of having concerns of the open source nature of a engine as long the free version isn't too restricted it should be sufficient for most purposes. Is there any other engines that has a low learning curve to learn, or is it only Unity and Gamemaker left? 

>What am I saying is that it might be easier to just drop wad compatibility, and create a game with some levels (these can be ports, creating an importer that works 90-95% of the time then fixing up the remaining by hand is much easier that creating a 100% one), just to avoid dealing with those shitheads
I rather prefer a parser/translator that doesn't require additional manual work to make the converted wads playable. Though when I would go to this route might as well start trying to use Quake 1/Sven Coop maps instead I think. UT2004 onslaught maps would be also preferable but that requires having to reverse engineer the map format and extracting as much info as possible. 

If anything I rather stick with GZDoom for this project, I made too many failed projects in the past and doing the switch again would just be a developmental suicide at this point. I will try with my best of my abilities to make my mod not as shit and burdened with all the flaws this source port has and when the damn version 1U is ready for release then I'm going to switch to a different engine and bury GZDoom modding for good.

1343498257386.png
(113.5KB, 219x330)

So are we doing demo day today?

>>18310
yes

1585557246281.jpg
(160.7KB, 1920x1080)

>>18310
Every demo day I start rushing on my game hopelessly, but obviously I can't throw anything meaningful together in a day or two, then I feel bad for not having worked on my game before. If only it was demo day every day, maybe I would feel bad enough about it to make real progress.

DEMO DAY IS HERE

>>18328
>>18328
>>18328
>>18328
>>18328

dubs.jpg
(5.8KB, 236x285)

>>18333
Checked and posted

>>18333
kike is hosting demo day too
https://8chan.moe/v/res/149845.html

Got called out for not posting in the AGDG threads, so I'm back to doing that.  Language wars are gay, progress bashing is gay (different from criticism), excessive shitposting is gay.  Crossposting is also gay but we're a bunch of rule breakers around here anyway, so I'll join the party.

> ANUBITEK
Squad control has been updated and is much better, UI has been updated but still experimental, Tactics control has been updated and squad members/NPCS can take turns.  Work on NPC AI today.
https://youtu.be/NuRU4qo4tjE

>PLUTONIA: BLEEDING EDGE
Map01-Map06 has been modified, it is designed with my fork of Brutal Doom Sperglord Edition in mind but also works with vanilla (pistol start past Map02 not tested yet).  Trying to make new videos but I forgot how much I suck at vanilla Doom, I've been spoiled by fun to use and OP weapons.
https://youtu.be/-ppCxS7QULs

Leopard_1.jpg
(134.6KB, 1280x1024)

Leopard_1.jpeg
(248.7KB, 1280x800)

Welp I almost finally finished the damn Leopard 1 model which I started working on saturday, the damn gun mantlet gave me some trouble to get it right as it was for me hard to tell how the hell it needs to be shaped, the turret was quite bitchy to get it right as well. Now I just need to find ways so that it looks like it has thick armour and applying crappy textures to it. 
It does worry me that it has 5,5k~ triangles now compared to my Tiger 1 model which has 3,9k triangles.

>>18701
5.5k tris isn't bad unless you're targeting 15 year old computers.

We made a new icon for the scanner command, hopefully this one makes it more clear what it does. We want to avoid text in the game completely because it's such a pain in the ass to work with.

Also what the actual fuck is this fucking captcha? Kill yourself pasta nigger.

>>18784
I managed to get rid of 600 triangles by deleting some polys from the treads which I forget to do, Blender keeps adding unnecessary polygons whenever I use the nurbs path but its easy to delete it at least. My main concern is mostly that GZDoom doesn't handle models that well, I tried a stupid anime girls monster replacer to see what the fug it is and at the entryway map from doom 2 it lagged like hell for me, I inspected the md3 models of it and those models clocked at whooping 50-60k triangles and they barely looked detailed. 

>15 year old computers.
pfft not with the attitude that Graf Zahl has, you need at least a graphic card that is capable of OpenGL 3.0 minimum so that you can enjoy playing wads made in the 90's with no gud graphical enhancements to it. 

>Also what the actual fuck is this fucking captcha? Kill yourself pasta nigger.
Eh for me it is easier clicking on those boxes then having to squint my eyes looking at those warped text. Check the meta thread, BO mentioned it is because this place got hit by a spammer again.  

>>18510
>it is designed with my fork of Brutal Doom Sperglord Edition 
I forget to ask, what changes did you do to this mod? And on what Brutal Doom version is it based on?

15-11-20-1605445439_scrot.png
(1.1MB, 1920x1080)

15-11-20-1605445478_scrot.png
(915.5KB, 1920x1080)

Leopard_0.png
(452.6KB, 1344x756)

Welp the Leo 1 model is ingame now, I need to improve the color a bit more so that the turret stands out, texturing the treads and animating the wheels which I'm going to do it later, took me long enough to model it. Then I can work on other aspect of the mod, I need to improve the cannon code a bit more adding upgrades for the bee and chemical shell, improve the chemical shells effect more. 

Then I think it might be time to also work on a new buy menu which is going to take a while as I'm not quite sure where to begin if I can do it with ZScript only, I need to work on a mock up GUI first before, one of the challenging aspect would be having to account the script so that it displays only eligible items depending on the player class. The player class should allow for customization but I haven't thought up of any good idea at the moment, the new game menu should only display one class that's for sure. What I have in mind is that the player should be able to switch to a new tank model when he has enough credits for that for example when he selected the Leopard 1 model he will be able to upgrade to Leopard 2 or Leclerc tank and once purchased he can switch back and fourth at certain checkpoints, when the player switches back to Leopard 1 then his weapons will be swapped with lower tiers/lower caliber as it cannot support a 105mm+ weaponry.

Orson.jpg
(13.7KB, 480x360)

ideas for implementing walljumping in a 2D platformer? my method isn't working no matter how I rework it
>when the player jumps, check if they're touching a wall on the right or left side
>if they are, push them towards the opposite direction
it seems like when the player jumps, it stops accepting inputs/making checks related to jumping until they touch the ground again

>>19171
>it seems like when the player jumps, it stops accepting inputs/making checks related to jumping
It shouldn't, unless you're using an existing platformer engine.
Some things that could go wrong:
When checking if the player is next to a wall, you need to give some extra margin. E.g. it should work if the nearest wall is nearer than 5 px. Depending on the physics engine, it can be simply impossible to be just next to the wall (esp. if positions are float), and even if it's possible it can be a frustrating experience for the player.
You should not just push them in the other direction, you have to give some upward speed too, otherwise the player will just bounce off the wall, not jump.
And yeah, double check you don't have something like if (!on_ground) return; in your jump code

>>19185
giving some extra buffer space between the wall is a good idea for landing the jump, but I do know through debugging that the player is indeed touching the wall. it's not doing it reliably though, almost seems like an engine bug. right now I just have it set to push the player away for debugging since I wasn't sure that actually increasing their x and y speed was doing anything.

>>19213
Well, if nothing else helps, do a printf or something in your jump input handler and print out everything that seems relevant.

2020-11-17_13-55-36.mp4
(7.7MB, 1920x1088, 00:10)

Working on particles and animations, added fireflies to trees.

>>18701
>>19822
Why doesn't the leopard 1A1 have the side spaced armor on the turret?

>>19828
Didn't meant to reply to>>19822 meant to>>19144

Leopard_1.png
(125.6KB, 1611x2096)

>>19828
>>19829
Because I based the model on this blueprint.

>>19830
Oh. But the mantlet has some more angular features. But eh, I wish you luck.

>>19831
It was annoying for me to get that part right so I decided to say "fuck it" and looked up other pictures of the Leopard 1 model which seemed to have a slightly easier gun mantlet to model off from. 
>But eh, I wish you luck.
Thanks.

what's the best IDE for ganoo/linoox that has support for haxe?

18-11-20-1605733470_scrot.png
(2.1MB, 1920x1080)

19-10-20-1603081274_scrot.png
(2.5MB, 1920x1080)

19-11-20-1605771545_scrot.png
(2.3MB, 1920x1080)

15-11-20-1605457018_scrot.png
(1.3MB, 1920x1080)

15-11-20-1605464267_scrot.png
(2.4MB, 1920x1080)

Hmm, damn I played though a few different monster mods and so far only my modified complex doom monster pack (to make it work with GZDoom) are the only ones that are quite challenging for my current player class and its weapon. I played also the touhou* doom mod, hard doom monsters, and the madhard monster replacer mod and they barely make a dent to my tank class with composite alloy armor. The legendary monsters from complex doom LCA are the only ones that can absolutely wreck my shit if I don't make careful Sabot shot placements. 

*Some Touhou monsters does a lot of damage to my player class but its limited to beam attacks and charging attacks. 

I think the cannon weapon is indeed a little bit way too versatile than it should be, but to be honest I don't really plan on nerfing it either, I could probably get rid of the Sabot and the Chemical shell and delegate them to other weapon types instead. Being able to use a fast firing highly accurate machinegun seems to do really make a difference when having to deal with all the monsters that are less than 100 HP heavy that makes up the bulk of the map monster roster. I can still recall that the autocannon mode on my much older version barely made a dent against hell knights. Oh well. Actually now when I think of it I think I should check out if there is any monster mod made for guncaster/trailbailer and whatever there is that makes the player ridiculous OP and see how well they are against my classes, this is something I totally overlooked. 

A way I could balance it is I could make it so that the supply crate only gives shekels and make the ammunition prices much more expensive to force the player to be more thoughtful of its ammo type usage, so for example it is not wasting APHE against a bunch of weak blob targets as a HE substitute, then of course when I get to add different tank classes they will have also different ammo classes, so smaller caliber are cheaper and the reverse is true for higher caliber weapons. 

I might also see later how well the Panzer III Ausf J will fare with its 50mm cannon which wouldn't be able to 2-4 shot a cyberdemon with a fully upgraded cannon.

For now this week I'm going to focus on smaller things, so tweaking particle effects, improving the weapons and other boring crap. So around monday when I'm motivated enough I'm going to take a look how calling functions work in ZScript, I know I can call ACS functions anywhere on decorate and even as keybinds but I'm not quite sure hows that's the cases for ZScript. It would be a bit stupid when I have to rely on ACS in order to call a ZScript function via keybinding. I think I should also work on adding a items replacer addon because it annoys me looting vanilla items that are absolutely useless for my class, the items will appear around 20-30% less because there are certain mappers that thought its funny scattering around bonus items in a shotgun pattern. 

>>19949
I'm not sure about Haxe but you could give geany a shot which is lightweight and lets you customize a few things, if by support you mean things like highlighting then I think you would have to add them on your own.

>>18333
I had planned to post this demo of Dischord's battle theme a few weeks earlier, but never got around to it.

Weekly Progress Report
-Continued raw writing for act 2 and act 3.
-Officially finished something - the raw writing for Ending 2 Rewrite of the bad end where you lose the normal-end final-boss-fight at Metronome Tower.
-Continuing my art training in /monster/'s art bootcamp.

In my efforts to limit feature/scope creep, I will warn that lewds are gonna much more sparse than I had wanted them to be. At the barest minimum, you'll get full-frontal Bravura and Scherza's Tits.

I want to make a FPS with multiplayer type of game using godot, but im an artfag that doesnt know how to code. anyone have any tips on how I can learn how to code?
Am I aiming too high for a first project if so where should I start instead?

Anyone have any updates on that deus ex like game set in poland? that looked pretty cool

>>20565
Ignoring >>20565, if you're serious... Yeah, aim lower, your first project should probably not even a game or something very simple. But if you're going with a game, at most 2D, 3D is just too complicated to start with. Things like "contacts lists" might sound boring, but it helps you get used to how to store and manipulate things in memory.
Sorry if it didn't help much, I suck at giving advice for complete beginners

>>20565

Here is a channel which covers how to make FPS/Multiplayer game on Godot.
https://www.youtube.com/c/Garbaj/videos

>>20565
Is there any reason you're not considering a mod for some existing FPS game?

ClipboardImage.png
(808.9KB, 1181x646)

Our game has a new look.

>>20565
>artfag
>godot
>multiplayer fps
Just find something to mod.

>>20565
You can get the skeleton for an FPS in 5 minutes with ue4 then art it up while picking up programming concepts along the way. BP will help you with concepts, but you should be careful to move bp prototyped classes to c++ before moving on from them if you really want to learn programming.

MP fps in godot for a first programming project sounds more painful than educational.You will be wrangling syntax you dont understand, trying to implement programming concepts you don't understand, while dealing with netcode you don't understand. One at a time is better.

25-11-20-1606326125_scrot.png
(1.2MB, 1920x1080)

25-11-20-1606313852_scrot.png
(769.9KB, 1920x1080)

das_fühl.png
(121.8KB, 680x497)

Well about god damn time I managed to get some progress going, after several days of screaming inside my head how the hell displaying graphics works in ZScript I finally found some solution including keyboard input. Right now I just used a double for loop to display grids of cube but this isn't the layout I'm going for, its still going to be grid based but I want also have text inbetween so at the left side the item icon is shown then when it is a weapon also its ammunition type and it should display the quantity and the cost of it. Of course I also need to figure out that this whole thing needs to be categorized also.

I think at the very least I need 1 week to get the barebones stuff working, hopefully I get this whole system finished before the next demoday as the HPIc (Hyper Production of Items cube) is a fundamental feature of my mod.

I'm going to make an MMORPG.

>>21847
Do you buy ammo or craft it? I don't really get what the grid is supposed to do.

>>21947
>Do you buy ammo or craft it? 
Crafting, the HPI cube inputs materials and outputs selected items.
>I don't really get what the grid is supposed to do.
I'm figuring out how to render those images and what items got selected. Then I need to find out how to handle structs and array variables because I want the weapons also display its ammo type on the same grid.

2020-11-18_060034.png
(66.1KB, 1288x747)

>>21946
Go for it!  I did!

I started from scratch and had to learn network programming.  The first commit was a client sending a message and a server echoing it.  5.5 years later and development is still going strong.

If you use something like UE4 instead, there's a lot that you won't need to do yourself.  Depends on why you want to do it.  For me it started out as something fun to make, not necessarily because I planned it to be popular or make any money.

26-11-20-1606420993_scrot.png
(978.5KB, 1920x1080)

>>21947
I got the array thing working and found out I need to initialize the variables at  Worldloaded() for EventHandler, so now I just need to figure out that a weapon item contains also the ammo item at the same grid column. Though for some reason I fucked up the position and I have no idea how to fix it at me moment.  Once I got this damn thing sorted out I also need to figure out which item got selected.

What's a realistic road map for making a cube world clone that doesn't work? Like starting by making a platformer, then make a smash style game to develop a combat system, then make it 3D but linear, then make it open world sort of thing. Making a series of games which build off each other until all the components merge into the final idea.

>>22092
Or you could just make the game?

I'd start with making the voxel map system since everything more or less depends on the way it works. Then add a player character that can jump on it and then it's mostly up to adding features until it's done or until you abandon it and don't refund anyone for scamming them.

eeeeh.jpg
(42.8KB, 960x945)

>>22092
If you want to make Cube World the first step would obviously be making basic world generation. A working voxel map that is either finite or infinite, whatever your goal is. Then you make it possible to move around it, then you make the rest. Good luck, don't kill yourself like Wollay probably tried to multiple times.

>>22120
>Hi I'm new Dev
>I'm making a multi year project open world action RPG as my first game
>This is my suicide note.

>>22138
The maps the most important part of it. Making that when you have no experience is a bad idea. That's why I think setting small goals is a better one.

>>22140
>The maps the most important part of it. Making that when you have no experience is a bad idea.
Exactly, it's your core feature so if you can't make it don't bother even making that kind of game. Start with the hard shit and fail fast, otherwise you're just wasting your time and should make an entirely different game.

>>22143
That means doing it twice. Once when you suck and once when you have a clue what you're doing. And building a map when you haven't laid out the core combat means a bad map for your gameplay. Your combat is going to evolve as you implement it after all.

>>22145
You're not supposed to implement actual proper map generation, it could just be a flat voxel plane because the hard part is getting your architecture with chunks, streaming etc working. You're never going to git gud at that by implementing a combat system or a platformer and you're definitely not getting it perfectly right on the first try regardless of skill level.

Do any of you have any experience with SDL2's software rendering? If so, how bad is it? I'm aware that this only benefits people running on weird SBCs, toasters so ancient they don't have a GPU, and people running niche OSes without OpenGL or Vulkan drivers, but I still find the idea appealing.

Game_programming_guide_for_dummies.jpg
(2.5MB, 1624x9454)

>>7067
I feel a little sorry for you. Making the very basics of 2d collision game engine with entity movement should take about 2 hours, even if you are mentally retarded. I'm (not) joking.

Anyway, you should try deepnights 2d tutorial. Just google it. He uses haxe, but the tutorial is very universal. It's a really good place to start if you are having trouble figuring out the basics. In literally two hours you will have primitives colliding with primitives, with minimal code and many ways to continue from there. In a week you should have a working test game with basic features.

I think you might have started from the wrong place to make a game. Start simple and gradually get better. Your games will grow side by side with your skills. It's a great journey, but remember to have a real education and job also.

I started wrong years ago by making my own game engine from nothing, and burning out by making it. I only made a few "demo tier" games with it. It was still a very good learning experience, but a terrible way to start making games. Just don't do it.
my engine was a basic 2d blitter with some advanced collison/deformation routines, full opengl scripting language shader support and some fansy audio stuff. It was shit - and way too much time wasted to have "full control" of your game engine. If I would start today, I would use  torgue 2d or haxe and just learn to script well - literal years of life saved and maybe you'll end up with an actual game in the end.

Stop worrying and just play around with torgue or haxe. You will learn more by doing more and not focusing on coding shit with bad tips from an arabian nose waxing board.

28-11-20-1606551125_scrot.png
(418.3KB, 1024x600)

28-11-20-1606551039_scrot.png
(616.1KB, 1024x600)

>>22369
>I feel a little sorry for you. Making the very basics of 2d collision game engine with entity movement should take about 2 hours, even if you are mentally retarded. I'm (not) joking.
Well I guess either Godot is not that newbie friendly or maybe I'm indeed too retarded to figure out what fucking functions I have to call to get a shooting projectiles to work, all that I managed to do with my tonk game in godot is a movement system that I janked from the examples and that it shows different sprites depending which movement button got pressed, the firing button was absolutely broken as the projectile moved around with the player.

>I think you might have started from the wrong place to make a game. Start simple and gradually get better. Your games will grow side by side with your skills. It's a great journey, but remember to have a real education and job also.
Mhm yeah I hoped when I try to do a very basic tonk game I would be able to get something close similar to battlecity or something but that didn't worked out well. Also I made a shitty coin collector game with pygame but seeing that it needs so much boilerplate code to get anything to work was not very encouraging for me.

>Anyway, you should try deepnights 2d tutorial. Just google it. He uses haxe, but the tutorial is very universal. It's a really good place to start if you are having trouble figuring out the basics. In literally two hours you will have primitives colliding with primitives, with minimal code and many ways to continue from there. In a week you should have a working test game with basic features.
Alright I'll take notes then when I'm the phase of making a game instead of modding it. 

>I started wrong years ago by making my own game engine from nothing, and burning out by making it. I only made a few "demo tier" games with it. It was still a very good learning experience, but a terrible way to start making games. Just don't do it.
I don't have any plans to write my own game engine because that shit just takes tons of time, investing on time that could have spend on refining the game instead. 

>Stop worrying and just play around with torgue or haxe. You will learn more by doing more and not focusing on coding shit with bad tips from an arabian nose waxing board.
How good is the torgue engine documentation? I have heard of this engine before and the only games I played on it was the orange block (continuation of the better mod/death to blocks) and blockland mod.