Curses feature branch

[Huqhox]

Some quick ideas:

"curse of the vampire": increased strength and/or speed, but take damage from all light sources (stronger the source the higher the damage) and when stepping on to lit squares (gives staff of darkness a use!)

diamondskin: higher AC but amplified damage from sound/cold

insane metabolism: increased speed but food consumption exponentially higher (no time even to rest fully without getting hungry/weak, must eat constantly).

shadow curse: increased stealth but reduced visibility.

Liking these (especially) the last two a whole lot

[PowerWyrm]

Currently, items that only differ by curses will stack. So don't pick up boots of speed [2, +10] (+10) (??) when you have boots of speed [2, +10] (+10) in your inventory

[PowerWyrm]

They are probably a little rare at the moment.

There's 1/20 chance at considering adding a curse, but you get only one shot vs tval. If the pick is incorrect, no curse is added. This should give 1/(20*curse_max) chance of getting a sticky curse... and probably close to zero chance of getting something else. Maybe lower the initial chance, but make the pick guaranteed vs tval.

[PowerWyrm]

The "stone" curse has no "time" value, so it never triggers.

[PowerWyrm]

Yeah, looks like I've broken random teleport - d_m has noticed it too.

Just adding this at the end of do_curse_effect() should fix the problem for all curses with timed effects:

curse->obj->timeout = randcalc(curse->obj->time, 0, RANDOMISE);

[PowerWyrm]

Just adding this at the end of do_curse_effect() should fix the problem for all curses with timed effects:

curse->obj->timeout = randcalc(curse->obj->time, 0, RANDOMISE);

Well that's still not working. It appears that the original timeout is set when a curse is appended (normal object gets cursed), but not when an already cursed item is generated. In this case, the timeout needs to be set when the object is created. The easiest way to fix that is to add the same code in copy_curse() inside the "randomise" part, which is called by ego_apply_magic() for ego curses and object_prep() for kind curses.

[Ingwe Ingweron]

Not sure which branch the nightlies are now; I assume it is the curses branch.

So, a couple bugs still around:

1. kill a pass through walls monster, and sometimes their drop appears inside the wall. @ has to dig or stone-to-mud to get at the drop, and has to guess as to whether a drop occurred.

2. Very odd to kill Sauron, get the message "... a magical staircase appears", BUT no staircase! @ had to deal with Angalacon and a Greater Balrog, summoned by Sauron before his death. Having done so, @ looks around for the staircase, but it is nowhere to be seen.

Sorry, I forgot to get a savefile for either of these events.

[Ingwe Ingweron]

Another bug in the nightlies...

The One Ring, retains the "??" even though identify scrolls reveal no new runes. "You do not know the full extent of this item's powers." But, everything seems known, including that it activates for bizarre things. Is it that @ is required to risk the Bizarre activation before knowing the item fully? Or even multiple activation until each possible bizarre activation is known?

[PowerWyrm]

Another bug in the nightlies...

The One Ring, retains the "??" even though identify scrolls reveal no new runes. "You do not know the full extent of this item's powers." But, everything seems known, including that it activates for bizarre things. Is it that @ is required to risk the Bizarre activation before knowing the item fully? Or even multiple activation until each possible bizarre activation is known?

I don't know if it's the same issue, but in my variant (which is roughly based on the latest code) jewelry doesn't get fully IDed until worn. Of course, that would mean equipping the One Ring to check.

[MattB]

Can one ever know everything about The One Ring?

[Ingwe Ingweron]

I don't know if it's the same issue, but in my variant (which is roughly based on the latest code) jewelry doesn't get fully IDed until worn. Of course, that would mean equipping the One Ring to check.

As I recall, the events occurred as follows:

1. Found The One Ring and picked it up.
2. Since @ was running with the birth option of all runes known, The One Ring when examined in the pack showed everything as known.
3. @ Wielded The One Ring.
4. Immediate Game CRASH.
5. Reopened game and wielded ring again. This time it shows with "??".
6. ?Identify show nothing needs to be learned, but The One remains "??".

[MattB]

3. @ Wielded The One Ring.
4. Immediate Game CRASH.

Well...naturally.

[AnonymousHero]

Well...naturally.

On the upside OP is going to live forever!

EDIT: Or... wait. Is that an upside?

[Nick]

Not sure which branch the nightlies are now; I assume it is the curses branch.

The feature branches have all been rolled into the nightlies now. This thread is still fine to report bugs on, or there are several other options

So, a couple bugs still around:

1. kill a pass through walls monster, and sometimes their drop appears inside the wall. @ has to dig or stone-to-mud to get at the drop, and has to guess as to whether a drop occurred.

2. Very odd to kill Sauron, get the message "... a magical staircase appears", BUT no staircase! @ had to deal with Angalacon and a Greater Balrog, summoned by Sauron before his death. Having done so, @ looks around for the staircase, but it is nowhere to be seen.

Sorry, I forgot to get a savefile for either of these events.

If you do get a savefile for either of these, that would be very helpful.

[Nick]

Well that's still not working. It appears that the original timeout is set when a curse is appended (normal object gets cursed), but not when an already cursed item is generated. In this case, the timeout needs to be set when the object is created. The easiest way to fix that is to add the same code in copy_curse() inside the "randomise" part, which is called by ego_apply_magic() for ego curses and object_prep() for kind curses.

Thanks for this analysis. There's a further subtlety - artifacts don't get other curse values randomised. So what I have done is make copy_curse work differently for new and copied objects.