If you're thinking about staking DUSK to become a provisioner, there's a risk conversation most people skip — not "will the price go up," but "what happens to my stake if I mess this up." Dusk actually splits that risk into two very different tiers, and the difference between them is bigger than most staking posts let on.

The paper separates faults into minor and major. A minor fault is something like failing to broadcast a candidate block when you were selected to — basically not showing up when it was your turn. That gets you suspension and soft slashing, which means part of your stake gets locked and your influence in future selection drops, but you don't actually lose the tokens. A major fault is a different category entirely — broadcasting invalid blocks, double voting, or putting out two different candidate blocks for the same iteration. That triggers hard slashing, and hard slashing burns part of your stake outright. Gone, not locked.

What I think gets missed in most "should I stake" content is that these two categories punish completely different kinds of mistakes. Soft slashing is basically the network being patient with you — it assumes you were offline, had a bad connection, or missed your window, and it penalizes you without permanently damaging your position. Hard slashing assumes intent, or at minimum, gross negligence — the kind of behavior that could actually threaten consensus integrity if it went unpunished, like trying to vote twice or push conflicting blocks.

My honest take: this two-tier system is actually reassuring if you're a serious provisioner running solid infrastructure, because it means a bad night of internet or a missed block isn't going to burn your stake. But it also means the tolerance for genuinely bad behavior is basically zero, and the penalty scales up with repeated offenses too. This isn't a "slap on the wrist either way" system — it's built to be forgiving for honest mistakes and brutal for anything that looks deliberate.

@Dusk
$DUSK
#dusk