Weekly Penguin
October 5th, 2025

Things that go boom

What am I working on right now?

Two sets of assets – from which only one each is needed for the scenes needed for the reveal. However, due to the similarity of the stuff in the sets, it just makes sense do the whole thing in one go.

The first set is a tad boring one; power-, fuse- and compressor cabinets. Can't have industrial vibes without tons of those, naturally. There isn't much to say about this set, it is what you'd expect and have seen in the previous games. I can't use any of the previous assets due to the rendering engine changes, mind, so it was time to revisit these once more.

The other set is far more interesting; the base set of exploding things. Gas bottles, fuel drums, jerrycans, the whole works.

Speaking of the explosives...

Optimising things that go boom

Since explosives became a focus of development, I decided now is as good a time it'll ever be to do the planned 'pyrotechnics optimisation sweep' of the new engine. I've already done similar things to almost everything else – projectiles, hitscan calculations, off-screen data culling, etc.

In the past there were three styles of pyrotechnics;

  1. splash damage (i.e. rocket explodes)
  2. projectile flames (i.e. gas burner fire cone)
  3. triggers (large damage areas)

They all could've needed some tinkering. They work well, but my motto for the new engine is to 'do more with less' to allow even more detail on screen – both in amount objects and graphical detail.

So let's run them down.

  1. Splash damage

    File under 'big changes with small tinkering'. It comes really down to culling out which objects to test the damage against. The new engine has several quick checks to see if objects are anywhere near each other before needing distance calculations or – in some cases – world geometry line-of-sight checks.

    Speed boost with this improvement is almost non-existent in a typical gameplay scenario, but using shared data culling checks streamlines the code a lot. Since I'm the numb-nuts who has to figure the code out, I'm sure it'll pay itself off down the line.

    On the visual side, there are less particles generated by explosions, meaning fewer rendering calls. That doesn't mean the explosions are less splashy, mind. For example; in TAGAP 4, the firey-smoke pushed away from the explosion consists of multiple particles, each processed and rendered separately. Now its a single particle that is procedurally rendered as dissipating fire-smoke ring. Shader magic!

  2. Projectile flames

    This is the one that needed the improvements the most. It's no secret that the way TAGAPs do flamethrowers and fire in general needlessly costly to process. On my older PCs the most sure-fire way to tank the frame rate was to take a flamethrower, get a weapon power-up and use the sticky-flame alt fire on the floor whilst running.

    This was because each and every fire is an entity that generates flames in rapid succession. And every flame is also an entity. So the sticky-flame-inferno scenario, the engine sweats the CPU to death producing new entities that in and out at a stupid rate. Obviously that's not good. It works, but can result in slow-down.

    The new engine has two ways to tackle this.

    • The first one has been in the game for two years already – projectiles can now be generated two different ways, as entities or simpletons. Simpletons are super-simple objects that use an entity for its rendering structure; you can define the object as an entity, but fire it from a gun as a 'simpleton bullet'.

      And the new system is a massive improvement. During the stress test I had 1500 projectiles on screen, continuosly generated and removed, complete with shader effects, without any frame drops or slowdown.

    • The second way is more a design fundamental change; do the flames need to be projectiles? For things like flamethrower cones and those spinning burners, sure, but most of the fire is s direct burst from a gas line or a fire raising up. That can be done even faster by having a simple collision test.

      In the past I ruled this out because it lacked the instant and clear feedback in comparison to the projectile-approach. This time, the engine now has a couple of new tricks that allow even the collision check flames to give you the necessary feedback – telling you to 'step out of the fire, dood'.

      And as a final performance boost, this can be done with one rendering call – as opposed to drawing multiple transparent sprites over each other like what happened in the past. Again, shader magic.

  3. Triggers

    For this one I didn't need to do much. The trigger collision checks have been tuned up in TAGAP 4 already. However, all the new visual damage feedback stuff added for the collision check fire can also be used with the triggers as well.

So, in a nutshell; everything takes less time for CPU to process and for GPU to draw – which leaves more room to make things look even better better and still run faster.

A simple example related to the aforementioned set of assets. Like before, a fuel drum exploding starts a fire. However now I can make the fire spew out smoke and embers without any fear of having ten of these on screen at once causing slow-down of any kind.

I'm sorry I didn't provide screenshots with this one. Again, my aim is to introduce all this when the game is revealed proper. Thank you for your patience – and also impatience if that's the thing, it means there are folks hyped for The Next Game.

I don't have 'escape from a burning building' on the level scenario list, but I'm really tempted to shoe-horn it in just to see what happens.

Playlist

Playlist is a regular feature in our Penguin DT blog; A chance to highlight cool games both old and new that I've been playing. As always, I believe that in order to make games, you need to play them, preferably with a broad scope when it comes to genres, so each day I dedicate at least an hour to actually playing games. The rest of the free time? There is no such thing, it all belongs to TAGAP!

I haven't updated this topic in a while, huh? Sorry about that. Obviously if you've been following the Weekly Penguin feature, I've been doing the Disgaea sequels playthrough. Disgaea 4: Alliance of Vengeance and Disgaea 4 Complete+ have been featured already and I'm currently going through Disgaea 6 Complete. Once we get to the D6 DLCs and onwards, I'm into new Netherworld territory.

In between the Disgaea titles I've been playing the two of this year's most anticipated titles (that are not called The Dark Ages).

First up was Shadow Labyrinth by BandaiNamco. If you've missed it, it's the dark reimagining of... Pac-man. And its a metroidvania. Yes, it is real. And its so good that it might even cause problems for me when it comes to deciding the GOTY. Coming from me that's saying quite a lot.

I don't want to explain the game's story, as it is one of those things that reveals in bits and pieces along the way as you explore the world. I went in almost completely blind when it comes to the story and the setting and there are some real surprises there.

Mechanically it is a metroidvania with all the tropes you'd expect, with one exception; the map is humongous, I'd guess at least three Symphony of the Nights big. Each locale is different enough, with their own enemies and unlockables. I've seen some mention that the boss fights are Souls-like stuff, but they're really not. If you get your ass kicked, you just likely need some upgrade or two, as is the 'vania way. They do give you a challenge, but not that kind.

The other great 2025 title is Teyon's RoboCop: Rogue City – Unfinished Business. I already hyped the first game to high heavens earlier and 'Unfinished Business' is a direct sequel to it. There's about week or two between the games, but the story continues straight to this one.

The sequel is more linear that the original, I guess partly because if I understood correctly, this was originally supposed to be a DLC campaign but the first game's success gave the publisher enough faith to fund it into a full title. However, that's the only part where the 'DLC past' shows; it's a full game alright without any bloat or unnecessarily stretched segments.

What the game looses in full linearity it wins back in having sequences played from perspectives other than RoboCop's, including Alex Murphy before his fateful district transfer. These scenes are short but effective – and smartly you're very vulnerable in them, making you respect the chromed titanium armour even more.

TLDR; both Shadow Labyrinth and RoboCop: Rogue City – Unfinished Business are must-play titles of 2025 (just play the first RoboCop before the new one, naturally).

What's Next?

The Next Game; back to asset grind. Next entry on the list seems to be textures.

The Playlist; I have no idea how much DLC Disgaea 6 Complete adds to the game – if that's anything like D4 Complete+, I might be in the Netherworld quite a while yet.

Until next time,

Jouni Lahtinen, the head penguin

Blog history

24-07-31

Seventeen!

24-04-19

April update

24-03-11

A speed bump

23-10-13

Dev update!

22-03-02

The End

11-06-23

Almost there

08-08-26

What's up?

08-07-02

Wrapping up

08-03-27

Gone surfing

08-03-14

Coincidences