hold opcodes

[fuzzie]

Has anyone worked out the differences between Hold opcodes 109 and 175, other than the obvious icon one?

 

In my tired state I see a different foot circle colour and also different behaviour, but thought it better to ask here and see if anyone knows better (maybe I missed some existing discussion?) before looking into it.

[devSin]

The paralysis effect (109) just holds the target (maybe with the state icon); the hold effect applies corresponding visuals along with the hold state (because the effect allows IDS targeting, this is how you get the icon and color glow and lighting effect and maybe the string if they run that too only on a character that can actually get hit with the hold effect; this was back before they could call external EFFs); and the hold 2 (185) effect should be duplicate but with a different value (so you could bypass the immunity to effect that was already given to creatures, e.g., for hold undead).

 

There shouldn't be any functional difference (the hold state is the hold state).

[fuzzie]

There shouldn't be any functional difference (the hold state is the hold state).

 

You're right; I got confused.

 

Just to have this on the forum somewhere, although perhaps it's already known: Since instant actions are executed immediately when they're pushed onto an empty action queue, ActionOverriding an instant action onto an idle held actor will still execute that action immediately, allowing you to (for example) teleport held actors around.

 

Obviously, if the instant action *isn't* instantly executed, then it'll just stay stuck on the queue until the actor is un-held and the queue gets run again.

 

I looked into this because it's important for Spellhold (you get HOLD_PARTY cast onto all your other party members to make sure they sit still during the cutscene/dream), and stupidly forgot all about the fact that they were also made unselectable and of course that's why I saw a difference in foot circle colour. Oops.

 

While I am derailing my own thread, something vaguely related: The "doesn't decode objects" interpretation of 'nodecode.ids' makes sense because this allows actions to do crazy broken things (ChangeAIScript) or re-evaluate even their non-dynamic parameters occasionally (AttackReevaluate).