Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Footkerchief]

LOGIC HATH SPOKEN.

The best thing about real true logic is that it announces its presence.

Still holding out for the morning of January 1st.  Debugging can suck it.

[Aqizzar]

I'm still holding out for 11PM Eastern Time December 31, just so Toady can say he released it in 2009.

Debugging can suck it harder.

[Cavalcadeofcats]

Terribly sorry 'bout this inquiry; I've been trying to read through the entire thread from the beginning, but 500 some pages is quite a lot.

Does anyone know if, anywhere in the thread, it's been mentioned if wearable/wieldable items will be able to use some form of tag to define what part is allowed to hold them?  Say, if I modded in a race with a [GRASP] tag on its mouth as well as its hands, that uses spears and daggers as weapons, would I be able to let them wield daggers in their mouth but spears could only be wielded in its hands?

I don't believe so, no, but Footkerchief would know for certain.

[Sean Mirrsen]

Not for this version, I think. Weapons use any GRASP they can get a hold on, pun intended. Armor uses whatever part type is assigned to it - gloves go on GRASP, boots on STANCE, helms on HEAD, etc. I didn't notice anything in the item raws that could specify the target part. This will have to wait until the items get the major overhaul and are merged into one class, if such an overhaul does happen in any foreseeable future.

[Footkerchief]

Terribly sorry 'bout this inquiry; I've been trying to read through the entire thread from the beginning, but 500 some pages is quite a lot.

Does anyone know if, anywhere in the thread, it's been mentioned if wearable/wieldable items will be able to use some form of tag to define what part is allowed to hold them?  Say, if I modded in a race with a [GRASP] tag on its mouth as well as its hands, that uses spears and daggers as weapons, would I be able to let them wield daggers in their mouth but spears could only be wielded in its hands?

I don't believe so, no, but Footkerchief would know for certain.

I didn't have anything to add beyond what CobaltKobold said above:

Hasn't been addressed afaik, rather like, say, 3 or 4-handed weapons.

[smjjames]

Man, there is a good possibility for a very late december release, but a release within the first week of January seems reasonable too. Heck, the remaining stuff (the section he is working on) might get finished or almost finished over the weekend.

Still, Toady does deserve to take as long a Christmas holiday break as he wants as he has worked so hard on this.

[CobaltKobold]

LOGIC HATH SPOKEN.

The best thing about real true logic is that it announces its presence.

Still holding out for the morning of January 1st. Debugging can suck it.

Funny how this turns out to a negative expression...

But actually I might like to see some horrific bug for this generation a la "civil war", or "The Forever Sleep"

[zchris13]

"Civil War" sounds good. I obviously missed that one. What happened?

[CobaltKobold]

Forget how it starts- hammerer punishes someone?

Anywy, someone of fort gets marked as civ-enemy of yours for a fistfight or something.

As an enemy, soldiers/guards target them and attck them, but-because the "enemies" are still part of your civ-the gme goes "Oh No! These guards are attacking us!" and flags THEM as enemies of your civ (while still being members) and they get attacked- it explodes from there.

[TKTom]

Forget how it starts- hammerer punishes someone?

Anywy, someone of fort gets marked as civ-enemy of yours for a fistfight or something.

As an enemy, soldiers/guards target them and attck them, but-because the "enemies" are still part of your civ-the gme goes "Oh No! These guards are attacking us!" and flags THEM as enemies of your civ (while still being members) and they get attacked- it explodes from there.

 Funniest thing is that it is a very dwarfy civil war, with there being as many different factions as there are dwarfs.

[smjjames]

Forget how it starts- hammerer punishes someone?

Anywy, someone of fort gets marked as civ-enemy of yours for a fistfight or something.

As an enemy, soldiers/guards target them and attck them, but-because the "enemies" are still part of your civ-the gme goes "Oh No! These guards are attacking us!" and flags THEM as enemies of your civ (while still being members) and they get attacked- it explodes from there.

 Funniest thing is that it is a very dwarfy civil war, with there being as many different factions as there are dwarfs.

Sounds alot like, if not exactly the same, as the loyalty cascade I've heard about that can happen in adventure mode.

[madman]

It just occurred to me that this might be an opportune time to fix a longstanding (and hugely annoying) bug.  The function that calculates straight-line distance for job items ignores z-levels.  Here are a couple reports that include simple ways to reproduce the problem: one, two, three.  It should be a quintessential one-minute fix, but I know how that goes.

I don't see it as being a trivial fix, because there is a tradeoff between accuracy and speed. The obvious options would be to use full pathfinding (slow), or assign a penalty for each z-level (breaks vertically designed fortresses depending on the penalty).

I'm not about to post on very old threads, but if I was trying for a quick fix, here's two options I'd consider:

(1) (quick and dirty) A fixed penalty for each z-level difference. No more than 2-3 per z level (otherwise vertical forts get broken - I always have many z-level spanning stairs right next to my workshops) or (best option) have a Z_LEVEL_MATERIAL_SEARCH_PENALTY value in the init.txt file (default it to 2 or 3).

(2) (nicer, but longer to implement). Instead of using the full pathing algorithm, do a depth first flood fill on the fortress which should usually terminate very quickly if just looking for 1-3 items (which is when this gets used), and possibly have a cutoff (say a distance of 10 or 15 or settable in init.txt, or have 'filled' a certain number of points) where it goes back to a simpler calculation cases, in which case you already have a distance d you have checked to, so any item is at least d+1 away. As a nice bonus, the flood fill could store enough information (each square stores which direction it got filled from) that it provides a path for the dwarf (if the dwarf is already at the workshop - a common case), and a tentative path back, and full pathing doesn't need to get used at all (unless the geometry changes).

Some sort of full pathing already gets used when picking an item to construct something (say a new workshop), but I suppose that the hit there is hidden by the fact that it is a user interface driven action anyway.

[Footkerchief]

Your floodfill idea is a good one, but the problem is that it's trying to do a straight-line distance check, a simple Euclidean distance, and is ignoring the (z₁ - z₂)² term.  As Draco18s mentioned in those linked threads, the game literally thinks that the distance from the trade depot to an item 20 z-levels directly below is ZERO.

[CobaltKobold]

madman:I'd guess 1.

Sounds alot like, if not exactly the same, as the loyalty cascade I've heard about that can happen in adventure mode.

Whzat?

Footkerchief: I have a counterexample save to that. But it's presently tied up in competition. Conclusion: Ramps seem to not invoke this mistake.

[smjjames]

madman:I'd guess 1.

Sounds alot like, if not exactly the same, as the loyalty cascade I've heard about that can happen in adventure mode.

Whzat?

Its been mentioned and talked about over on the adventure mode section. The description of how it starts is very similar.