feature request - noncumulative durations

[PowerDiver]

The discussion of prayer spells reminded me of something I do that feels borderline like abuse. Particularly if I am forced to buy out the alchemist a few times to get a restore or recharge, I sometimes find myself in town with huge stacks of buffing consumables such as potions of resist heat. I quaff the whole stack while waiting for recall to kick in, and often get an entire trip with the buff in effect without costing a slot.

Something is already done about speed boosts being noncumulative, but I don't like that either. It bothers me that if you cast speed and would get 15 turns out of it, it only lasts 6 turns if you have 1 turn left from a previous cast.

I suggest that any time you cast a spell with a duration effect, the time is set to the maximum of the new duration and the remaining duration. Perhaps it would be better for gameplay to just set to the new duration rather than taking the maximum even though it irks me.

This kind of change would make the difference in durations for bless vs prayer a little more significant.

[ewert]

A big +1 for this, yep. Just reset timer on renew to the new one, if the old one was longer, well tough break. Unless the coder wants to go the extra mile for a non-significant feature within a feature.

[Atarlost]

No reason not to use the longer. It really is a trivial bit of code in C.

[Nemesis]

Oh, thanks for telling me that speed was non-cumulative. In vain I tried to macro a pattern of "cast haste self", "cast haste self", "cast haste self", rest, without getting it to give me any longer duration. But this should mean that I really didn't do anything wrong, I should just have done it with the resistance spell instead.

[Derakon]

I've always thought that cumulative effects were screwy. Your suggestion sounds right and proper.

[fizzix]

taking the longer one is the right way to go. It should also be that way for negative status effects like confusion, blindness or fear.

[Derakon]

The game already has logic for one negative effect (paralyzation) to ensure that the player always gets at least one turn of normality before they can be hit by it again. It's possible we might want to use the same reasoning for other status ailments as well. Though, paralyzation is a different beast since the only way you can recover from it is by waiting out the duration.

Mostly I'm just saying "always take the longest duration" is not the obvious decision you imply it is.

[bron]

I suggest that any time you cast a spell with a duration effect, the time is set to the maximum of the new duration and the remaining duration.

This is a good suggestion, and has the advantage of simplicity. I admit that personally I like the idea of multiple casts giving some (diminishing) additional benifit. So I would prefer the following:

(1) Add half the remaining duration to the new duration.
(2) Set the duration to the result of (1) iff it is bigger than the current value.

With this, multiple casts (or quaffs) do add to duration (e.g. casting twice in a row gives 1.5x duration, thrice gives 1.75x), but the additional boost diminishes rapidly, and is capped at 2x the maximum duration of a single effect

[PowerDiver]

With this, multiple casts (or quaffs) do add to duration (e.g. casting twice in a row gives 1.5x duration, thrice gives 1.75x), but the additional boost diminishes rapidly, and is capped at 2x the maximum duration of a single effect

Is that really worth the effort in explaining it? The maximum of N IID uniform [0,1] variables averages N/N+1. If we ignore casting time, that gives something similar. E.g. cast twice, duration averages 4/3 * the average duration casting once, and limiting behavior of 2 * average duration casting once.

Hmm -- I guess I'm assuming durations of [-1 + 1dM] rather than the typical [M/2 + 1dM], but I think the point remains instructive even though it is not valid.

[ewert]

Tbh, I think keeping some form of duration stacking is keeping the same problem either with insignificant gameplay effects (read: pretty useless) or keeping the problem to some extent (read: still sort of broken but just not as badly). Oh I agree that if tuned just right one could argue it won't be "broken" but ... we could just as well then up the mana cost and duration to the "new default value of cast x times precombat" ...

MAX(New dur, Old dur), my vote.

[Pete Mack]

I disagree on this one. Cumulative buffing is just not that big a deal--more of a convenience than anything. I use it vs Morgoth to avoid carrying ?Chant and !Hero, but that's about it.

Making speed bonuses stack, even in the way you suggest, is a bigger deal:
it means even 1/2 casters can guarantee permanent haste self during a big fight, without ever having to reduce to normal speed. Just cast the spell every CL/2 + 1 normal speed turns.

[PowerDiver]

Making speed bonuses stack, even in the way you suggest, is a bigger deal:
it means even 1/2 casters can guarantee permanent haste self during a big fight, without ever having to reduce to normal speed. Just cast the spell every CL/2 + 1 normal speed turns.

Is that so bad?

A change that makes most of the game a little harder and makes M a little easier, allowing those who want to guarantee no double moves vs M not to have to scum for speed +30, is not necessarily bad IMO.

I know that in my own games this change will make things harder in total, so I don't mind if it makes some single aspect a little easier.