Final Stand (Scjohnson Entry!)

Summary:

Scjohnson's Entry.

Final Stand

scjohnson243 21 Aug 2015 16:10

Idea!

The country is being invaded! After an initial first attack that has rendered all military response unavailable the fate of the entire country rests on your shoulders, just you and your experimental x47 fighter jet. Its up to you to hold off the next waves of the invasion while your countrymen build up a response! Can you hold out long enough to allow your base to get a defense in place before the final assault?

Goals

The goals I have set for my first 30 days are extremely simplistic; Create a "Ship" that the player controls that allows up and down movement (left/right maybe? depends on how it plays out) - player fires and destroys enemy ships that "spawn" at random places on the right side of the map and move to avoid players fire. Player can be destroyed, at which time the game ends. As the game goes longer, more and more enemies spawn. Score is awarded for each enemy killed, and game is over when player dies.

Stretch Goals

These would be ideas I would like to try to implement if time permits, as long as the "Goals" are met, I feel thats a good, playable game. These would be "Extra" that could possibly be added depending on how i'm doing. I am sure I will add more to these as I get more ideas and work on the game itself.

• Allow "base" to "Build up" as time goes on, perhaps offering different powerups, or bonuses of some sort….
• Powerups player can pick up
• Multiple Enemy types with different behaviors
• Break down into "waves?"
• instead of just a random screen the player is only allowed to move up and down on, actually make a "level", randomized terrain, give the player a "Base" on the left side and for each enemy that passes it will do damage to the "base". Base takes damage and game is over when the base is destroyed. Maybe allow the player to get $ for killing enemies and allow him to have a tech-tree like upgrade path? upgrade the base, add weapons, upgrade the ship, etc. This could spin out of control quickly, but could also make this something "new", instead of just a really basic side scrolling defender clone, I dont think I have heard of anything like this before.

Current progress/Problems

…

Re: Final Stand

scjohnson243 25 Aug 2015 13:21

Wow.. 3 days.

I'm glad I dumbed this down to be able to finish - I have probably spent about 5-6 hours in this so far, and managed to create a new background, and have a sprite ship go left and right and not go "out of bounds"… thats it… Thats my entire project so far!

I'm using Visual Studio + Monogame, I looked at Unity, but just didn't like the interface.. I felt like I was dragging and dropping and not programming, so I have a feeling if I keep with this I would be limited in the long run…

Tonight I need to work on bullets and collision detection, I guess..

(The biggest hurdle at this point is figuring out what to work on when!)

Oh well, slowly working on it!

Re: Final Stand

Brett208 25 Aug 2015 16:13

SC,

That sounds like a good start actually. Drawing graphics, handling input, and boundary conditions.

Brett

Re: Final Stand

scjohnson243 25 Aug 2015 19:41

Thanks Brett - I think the biggest problem is i'm not sure what order to work on things, like what to do next, then after that, then after that, etc.

Just flying by the seat of my pants so far. :)

Re: Final Stand

PiscesMike 25 Aug 2015 19:51

Ah, the fun of game programming!

Might I suggest Trello.com - RBWhitaker has got me hooked on it. It you're basic 'let's organize this' website which allows you to add and move 'cards' around. Each card being a task that needs completion. There are a ton of ways to set things up, but if you want to take a look at the one I'm currently using it might give you some inspiration!

https://trello.com/b/hrDuPmi4/pavement-pugilist

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Re: Final Stand

PiscesMike 25 Aug 2015 19:53

Also, here's the link for the one we used during the Xevious 8 hour challenge. Just a little different as far as the 'list' titles goes, but still interesting for a reference point:

https://trello.com/b/lnaJhdAy/xevious

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Re: Final Stand

scjohnson243 25 Aug 2015 22:00

This will be very helpful! thank you!

now to go spend an evening trying to get a bullet to fire (which you guys can finish in like 5 minutes)

lol

Re: Final Stand

rbwhitaker 26 Aug 2015 02:19

If we can do it any faster than you, it's only because we've been exactly where you are and slogged our way through it the first time. After that, you learn how to make things like that happen. Even if you don't reuse code, you can reuse the patterns. All things considered, if you get bullets firing, I'd say you're probably even well ahead of schedule in your original goals.

Re: Final Stand

scjohnson243 26 Aug 2015 14:41

Ship up, moving, all that fun stuff… I never got to bullets last night, I had everything in my main Game1.cs file, decided it would be "better" and cleaner to move things to different classes, saw a tutorial on monogame that i'm going to use as a blueprint…

Added a "Player" class, and I'll just use an extenstion of this for my enemies….

I could be completely wrong, but in my head if I seperate everything I may be able to re-use it later, or when I'm adding "more" at a later date maybe I can re-use the code earlier……

I added another method(could be wrong syntax here) for "update player" instead of keeping everything in the main "Update" statement - so I just call updateplayer from update…. in my Updateplayer I'll check for keyboard input, and try to limit the output to the boundary of the screen (this is giving me an issue currently, still working though it)

Is this a bad line of thought? Will having the update call another update slow things down? its not currently, but I only have one ship online.. if I get to the point of tracking 20-30 ships at once, not to mention bullets and all the locations it may…..

Re: Final Stand

Brett208 26 Aug 2015 17:50

SC,

Method is the right term (means the same thing as function in some other languages). Typically separating the code into smaller parts is an excellent idea.

Polling for keyboard/game pad input in the update method is probably appropriate. I will often make an InputManager class in my projects that interprets inputs from different devices and then feeds the rest of the game. This way only the InputManager cares if you are pushing the left joystick, pressing the left arrow key, or pressing the left mouse button. The rest of the game just sees the InputManager saying go left. However, this may be too complex if you are only worried about using one input device or have very simple commands.

In general you are correct that calling more methods is slower than inlining the code. This might be noticeable in a very large loop that references multiple other methods. However, running through update methods for 30 items should be insignificant. If the code isn't running slowly, I would recommend optimizing readability and re-usability by separating out into methods as you see best. Later, if you continue optimizing and find your code running slowly or jerky, then at that point it is worth investigating why. In the meantime, you will just make your code harder to read and update by chasing possible future performance issues that may not materialize.

-Brett

Re: Final Stand

PiscesMike 26 Aug 2015 20:08

I completely agree with Brett208! I've been the kind of programmer that tried to stuff as much as possible into classes, and am coming to the realization that separation and encapsulation is much better. Me and RBWhitaker even had a talk about this as recently as last night.

Seems like you're still doing a great job figuring out where things should be… with a project of this scope, Update should be more that adequate for checking user input. In later and more advanced projects, splitting the input manager out into its own class (and interface) is a better option, but shouldn't really be an issue here. The main reason for doing it later is the ability to add additional support. You would define an interface and supporting class for input, then if you wanted to add, say, joystick controls as well, you could just add a 'joystick' class based on the input interface, and you would have both! Anyways, going a little off topic here.

As far as optimization goes, that's also something that was bored into my brain when I first started programming. (Of course that was machine code on a z80 processor). Seems these days with multi-Ghz processors and on video card video processing, it isn't as much a worry. RBWhitaker has even told me that optimization should come LAST. That is, do everything you need/want to do, and if you see an issue, then address it. I was a little skeptical at first, but to be completely honest, I've had TONS of stuff flying around and still haven't seen any problems. I even ran a test once with multiple instances of a textured cube model - the program let me add and remove instances as I wanted - and I didn't see noticeable lag until the cubes numbered in the 10,000's+. Bear in mind, that was 3D with rotation matrix in play as well, 2D should run a little better. So, 30,100, or even 300 'sprites' should probably be fine. (I believe that after the initial texture is loaded, multiple instances using that same texture just reference the original memory. So, when you add another draw method with a new position thats all the computer has to keep up with, the new position, not an entirely new texture).

Anyways, sorry for another wall of text! Take care, and good luck!

(As a PS: I don't think I could get a bullet firing in 5 minutes. Maybe RBWhitaker)

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets