Here's the beginning of that typing bullet hell game I mentioned before, feat. one of my favorite tracks on NG Newgrounds: Previewing Text Shooter

I've come up with a LOT of ideas for this one.

posting here to help crash the servers

He could shut down the forums to help out the resources needed to keep the site chugging!

But if that happens, where will go to sustain my nostalgia fueled lurking habits? Without it, I might grow up or something. Nobody wants that.

Wasn't there a "highest ever" users online thing? And isn't this way higher than that number used to be?

Yep, I remember a "most active online users" record as well. Dunno when it got taken out, or whether it was some sorta mandela effect.

Probably - we had 30,000 sign-ups yesterday!

Holy crap. The numbers that come to mind were either 1749 or around 3200 users at once.

Good god. Those poor servers.

Has anyone managed to get SWF assets running for Haxe HTML targets? OpenFL says it can be done, but am I imagining it right - that it somehow decodes the animations to its base SVG (or other form) and uses that in its non-Flash targets?

UGH. I had to switch away from SWF and just use PNG because it looked like crap for my Pico Day game. Some more complex/nested anims are bugged, components jumping around. A 300 KB SWF turned into 12 MB of library assets, and scaling doesn't look good.

<library path="assets/swfs/swf1.swf" id="swf1" preload="true" generate="true" />

Preload if you want it preloaded, generate if you want your AS linkage names to show up as classes you can instantiate. Without generate:

var clip = Assets.getMovieClip("swf1:MySymbol");

I wasted a lot of time doing this, so I am still peeved off rn 🙃

Ah. Well, that's the nail in the coffin for that lol

Regarding scaling - do you get a similar issue with SWFs running in Ruffle?

Imagine this, but with shitty upscaled resolution instead of infinite. That's what SWF import was giving me, especially since they're usually rotated.

Edit: ok maybe this one without NG scaling stuff:

I haven't played with ruffle much since I don't have any AS2 content.

I'm happy with everything I learned with this game (shaders!), but I'm definitely not making a physics-based game again.

Newgrounds Snowboarding

I've been working on a game for a couple of months now, where you avoid enemies and use them to destroy each other.

I recently uploaded a WIP to my itch.io, would appreciate some feedback: https://kurt-c0caine.itch.io/avoider

When it's done, I'll upload it to NG.

This is a really cool mechanic! I love that it takes something that happens so rarely in games, like getting enemies to take each other out, but it feels so cool when it does happen, and making that into the main mechanic.

I think one thing that'll help a lot is if there was a physical indicator on the ship itself when the blink is still charging. I kept thinking I was ready to blink, but I wasn't. But then when I started watching the bar in the top left I kept getting hit because I wasn't paying attention.

At one point I stopped using the blink and just relied on moving out of the way which felt more fun because it was more risky. Would be kinda cool if the game rewarded that, like rewarded "near misses".

Would also be interesting to experiment with like ricochet, maybe like a special block where the enemies bounce off of and you can try to target some chain reaction off of that.

Hey, thanks for the feedback!

Yeah, I'm thinking of adding a sound effect for when blinking is ready.

And rewarding near misses is a good idea, but I don't know how to implement it yet.

I guess one way to do it is to calculate the trajectory of the player at the next X timesteps and award a near miss if the predicted trajectory results in a collision, but there's none in practice, i.e.

near_miss = distance(predicted_player_location, actual_player_location) > epsilon

The smaller X is, the more finesse it would require - the player has X seconds / frames / timesteps to swerve away.

Sounds pretty complicated.

Pretty fun, although I agree it was more fun when I stopped trying to use blink. It may be too incentivized, unless I'm just using it wrong? The timing between "showing red" and "dashing" is really tight, although after a while I got more used to it.

The laser dissipation effect is still really cool.

Probably, but it doesn't need to be too complex - a simple linear regression will do. You'd have to store the previous coordinate of all enemies/players, and determine the next position by moving forward the same distance along the direction of travel of the player. An illustration might help:

The colour matrix is as follows: (where T=-1 is the previous timestep, T=0 is the current and T=1 is the next)

Name\Timestep |     T=-1    |  T=0  |  T=1  |
--------------+-------------+-------+-------|
Player        | Light green | Green | Blue  |
Enemy         |    Orange   |  Red  | Brown |

The black line shows the calculated trajectory of the player, and the grey line shows the calculated trajectory of the enemy.

Then, you determine if the distance between the player and enemy at T=1 is smaller than some tolerance E. If it is, then you got yourself a near miss. If not, then you don't.

Here's some python / pseudocode:

getPos = lambda x: (0, 0) #returns a tuple of the object's (x, y) position
diffPos = lambda x, y: (0, 0) #returns a tuple containing the result of x - y
distance = lambda x, y: 0 #returns the distance between x and y (scalar)
updatePos = lambda x, y: {} #returns a dictionary of all x + y

def calc_near_miss(threshold, all_projectiles, current, previous):
  """
  Calculates whether there's a near miss or not.

  * lookahead - how many timesteps of warning does the player have to move out of the way?
  the smaller the lookahead, the less time they have (e.g. if lookahead is 1, they have
  to wait until the last second to move away for it to count, or else their trajectory 
  will change)
  
  Let's assume the player character is called "main_player" and "all_projectiles" is 
  a list of all characters, projectiles, etc. Since we're using a simple linear 
  regression, we can assume that the position at T=N is just N * distance(T=0, T=-1)
  """
  
  if (currPos is None or prevPos is None):
    return False #no trajectory => no collision

  difference = {
    player: lookahead * diffPos(
      current[player], previous[player]
    ) for player in all_projectiles}
  }
  
  #updatePos() returns a new dict where each 
  #element = current[player] + difference[player]
  newPos = updatePos(current, difference)
  for player in newPos:
      if(player != main_player and distance(newPos[main_player], newPos[player]) < threshold):
        return True
  return False

timestep = 0
currPos, prevPos = {}, {}
#game loop
while(True):
  ###
  # <rest of game code>
  ###

  #calculate the distance the player has moved since the last timestep
  #basically, X_0 - X_{-1}, Y_0 - Y_{-1}
  currPos = {player: getPos(player) for player in all_projectiles}
  if(timestep > 0 and calc_near_miss(threshold, projectile_list, currPos, prevPos)):
    #custom code here - display a message, etc.
  timestep += 1
  prevPos = copy_dict(currPos)

The choice of distance function is up to you - be it euclidean or taxicab. Taxicab distance makes it much easier to reason about.

There's probably something I might have missed here, but the principle isn't too different from Continuous Collision.

Thanks for the detailed post, I'll consider it, but later.

Anyone considering joining the GMTK game jam (https://itch.io/jam/gmtk-2021) on June 11 ?

I haven't published any games in a long time. Last thing I made was Hikari's Station which was porting a friend's game maker game to the web which brought me so much joy to make and share:

I think I want to try making something small and really polished. I usually go for overly experimental. Just gotta find an artist to colab with.

I haven't made any games in forever (and no, clock day one-off BS doesn't count imo) either. The last thing I was working on this year was a fighting game, but life got in the way big time and progress in that ground to almost a halt.

I want to get more familiar with heaps before jamming. It really makes you think from the ground up it seems, like the base API is more of a demo for how to deal with the innards yourself when the time comes. I'm definitely not writing my own renderer in a jam.

Also, my snowboarding game got frontpaged, and there have been over a hundred notifications in a few days. I've had several things there before, but it really does feel like there are a lot more ppl visiting the site now. I know there actually are, but this hits it home for me.

It only needs 0.02 more score to be my #1. Vote 5.

Oh that's awesome! I also noticed today the view counts for all the Pico Day winners are super high. Seeing that definitely is a good motivation to spend some time making something and put it up. And hopefully it has a similar effect on others and more people make awesome things and it just snowballs from there!

I don't know anyone who would write a renderer in a jam, especially if it's one of the shorter ones lol

Whoa, that's neat af.

I actually finished a game for the GMTK game jam! You play as a pair of snakedragon spirits and do little rain dances to bring rain to the village:

https://itch.io/jam/gmtk-2021/rate/1084264

People seem to like it. Balancing it was really hard. It's definitely really confusing at first right now, there's kind of a lot of hidden mechanics.

On a side note, it's kinda weird joining a game jam run by one Youtuber. I feel like every other jam I've ever done has been organized by like a community, as opposed to "Mark". But this is the biggest online game jam now (is that right?) with over 5000 entries, which is more than the last Ludum Dare (https://ldjam.com/events/ludum-dare/48/stats).

I see a few of them made it to NG: https://www.newgrounds.com/collection/gmtkjam2021

RIP Audacity: New privacy policy is completely unacceptable! · Issue #1213 · audacity/audacity (github.com)

My old version didn't try to auto update, and I couldn't find a setting to do enable it, so it's hopefully still OK.

So who's planning on joining the upcoming NG jam (https://www.newgrounds.com/bbs/topic/1474833) ??

I'm super excited. I'm planning on it, pending finding an artist to collab with.