Increase the amount of particles that breaking a node generates

[ShadowNinja]

When a node breaks the particles don't seem to indicate that. This is especially noticeable with dig_immediate = 3 where it simply seems to disappear rather than breaking.

[VanessaE]

Nice tweak. +1

[Jordach]

I agree as well; just for the fact of little to no particles suck.

👍

[PilzAdam]

Have you tested this on slower hardware?
When playing around with this a while ago I noticed that many particles can cause a significant FPS drop.

[ShadowNinja]

@PilzAdam: No, particles are very slow. They would probably be much faster if we used Irrlicht's particle systems instead of entities.

[paramat]

Good grief 256 particles? Nope ;) I was actually considering making a request for halving the number of particles on node dig, but 32 is a good balance. Remember the particles are not representing the entire volume of the block, just a few bits breaking off when the player picks up the entire volume of the block.
Also, particles are slow even without collision detection, even more so with. 32 makes a very satisfying complex explosion of paticles, how can anyone not see 32 bouncing particles as a large amount? Please remember the central philosophy of Minetest: K.I.S.S. Even with irrlicht particles we should exercise some visual restraint and keep it at 32.

[Calinou]

Please remember the central philosophy of Minetest: K.I.S.S.

I'm… doubtful. You are only one of the core developers.

[paramat]

Well, K.I.S.S. is Celeron55's phrase, i agree with him and am repeating it here =)

[ShadowNinja]

Particles themselves are actually very performant; You can have something like 10,000 Irrlicht particles on screen before it starts to become a performance problem.
32 particles just looks like you brushed off the node a bit, it should look like the node broke apart. This isn't enough to make it look like it's completely solid, but it's much better. It looks particularly good with oddly_breakable_by_hand = 3 nodes.

EDIT: Note that this can noticeably affect performance on slower hardware, due to how slow Minetest's particle system is, so it probably isn't ready for merging.

[kwolekr]

The controversy in merging this PR is due to the slowness of particles, which shouldn't be a concern after a new, hardware particle implementation is added.
What this PR actually does is a rather trivial tweak. It modifies some constant value that should really be a setting (perhaps a per-node setting, even). We feel as if there simply is not enough value here to merge: closing.