0001068: Dwarves don't choose closest garbage dump (new in 0.31.03!)

(0003205)
Greyhawk (reporter)
2010-04-12 22:36
edited on: 2010-04-12 22:57

I thought for a moment that it just gives higher priority to dumps near pits than those not near pits, but I was mistaken. Still chooses farther dumps even after I got rid of all garbage dumps near pits.

(0003346)
udontknowme (reporter)
2010-04-13 17:57

Afair that isn't new, I realized that in 40d and only use one garbage dump at a time since.

(0003347)
Cryten (reporter)
2010-04-13 18:00

This did happen in 40d when the dumps where differnt levels but closer in range ignoring z levels. The simplified pathfinding most likely is to put less strain on the cpu for mundane tasks.
Is this the case here?

(0007917)
philomory (reporter)
2010-06-08 02:04

This bug really puts a damper on my 'garbage chute' idea, with a vertical tunnel that has a dwarven atom smasher at the bottom. All the dwarves run down the stairs to bottom-most pit (1 z-level above the DAS) and toss their stones into that, rather than the one that the stone was next to, 11 z-levels above :(

(0027293)
lightman (reporter)
2014-07-23 00:47

I'm seeing this in 0.40.03:

I created a garbage dump zone by some stones that I wanted removed, inside the fortress. I marked the stones for dumping. The dwarves proceeded to take the stones outside and dump them in a hole (not declared as a pit).

(0028803)
Footkerchief (manager)
2014-08-11 09:30

Awaiting a save from 0.40.06+.

(0029080)
Kirkegaard (reporter)
2014-08-15 01:29

Save from 0.40.8: http://dffd.wimbli.com/file.php?id=9399 [^]

(0030175)
Kirkegaard (reporter)
2014-09-13 13:54

@Footkerchief.
See save above.

(0030177)
Footkerchief (manager)
2014-09-13 14:51

Thanks!

(0033748)
Toady One (administrator)
2015-12-10 15:34
edited on: 2015-12-10 15:37

I understand the frustration with Z levels, but it's hard to calculate those distances quickly -- if I deal with that, it'll be part of a large pathing rewrite that has to account for central stairwells in some special way. Are there any saves where this occurs with the horizontal distances? The original report indicates there were, unless they meant 30 tiles over and 30 tiles back.

(0033750)
Talvieno (reporter)
2015-12-10 16:21
edited on: 2015-12-10 16:21

Would it not be possible to simply say, for each garbage dump, get distX + distY + distZ, and return the closest? It would ignore walls, and it would be somewhat messy, but it [should] work in a way that's a lot cleaner than how it works now (ignoring z-levels entirely).

(0033792)
Toady One (administrator)
2015-12-11 21:34

It doesn't ignore Z levels entirely. That's why I'm asking.

(0033793)
Talvieno (reporter)
2015-12-11 23:16

Ah, misunderstood then, apologies.

(0033794)
Toady One (administrator)
2015-12-11 23:48

I went ahead and messed around with the job z-calcs a bit, enough to address this problem (I think). Hopefully I didn't break anything else.

Issue History
Date Modified	Username	Field	Change
2010-04-12 21:52	Greyhawk	New Issue
2010-04-12 22:07	Greyhawk	Issue Monitored: Greyhawk
2010-04-12 22:08	Greyhawk	Tag Attached: dumping
2010-04-12 22:36	Greyhawk	Note Added: 0003205
2010-04-12 22:57	Greyhawk	Note Edited: 0003205	View Revisions
2010-04-13 00:19	Khym Chanur	Issue Monitored: Khym Chanur
2010-04-13 17:57	udontknowme	Note Added: 0003346
2010-04-13 18:00	Cryten	Note Added: 0003347
2010-04-13 18:01	Cryten	Tag Attached: Z axis
2010-04-28 18:16	Logical2u	Relationship added	has duplicate 0001621
2010-04-28 18:16	Logical2u	Relationship deleted	has duplicate 0001621
2010-06-08 02:04	philomory	Note Added: 0007917
2010-06-08 02:06	philomory	Issue Monitored: philomory
2010-06-08 05:47	Footkerchief	Relationship added	related to 0001554
2014-07-23 00:43	lightman	Issue Monitored: lightman
2014-07-23 00:47	lightman	Note Added: 0027293
2014-08-11 09:27	Footkerchief	Assigned To	=> Footkerchief
2014-08-11 09:27	Footkerchief	Status	new => confirmed
2014-08-11 09:30	Footkerchief	Note Added: 0028803
2014-08-11 09:30	Footkerchief	Status	confirmed => needs feedback
2014-08-15 01:29	Kirkegaard	Note Added: 0029080
2014-08-15 01:29	Kirkegaard	Issue Monitored: Kirkegaard
2014-08-18 16:07	Kirkegaard	Tag Attached: pathfinding
2014-09-13 13:54	Kirkegaard	Note Added: 0030175
2014-09-13 14:51	Footkerchief	Note Added: 0030177
2014-09-13 14:51	Footkerchief	Status	needs feedback => assigned
2015-12-10 15:34	Toady One	Note Added: 0033748
2015-12-10 15:34	Toady One	Assigned To	Footkerchief => Toady One
2015-12-10 15:34	Toady One	Status	assigned => needs feedback
2015-12-10 15:37	Toady One	Note Edited: 0033748	View Revisions
2015-12-10 16:21	Talvieno	Note Added: 0033750
2015-12-10 16:21	Talvieno	Note Edited: 0033750	View Revisions
2015-12-11 21:34	Toady One	Note Added: 0033792
2015-12-11 23:16	Talvieno	Note Added: 0033793
2015-12-11 23:48	Toady One	Note Added: 0033794
2015-12-11 23:48	Toady One	Status	needs feedback => resolved
2015-12-11 23:48	Toady One	Fixed in Version	=> Next Version
2015-12-11 23:48	Toady One	Resolution	open => fixed
2016-05-10 12:15	lightman	Issue End Monitor: lightman
2016-07-19 03:11	Kirkegaard	Issue End Monitor: Kirkegaard

Notes
(0003205) Greyhawk (reporter) 2010-04-12 22:36 edited on: 2010-04-12 22:57	I thought for a moment that it just gives higher priority to dumps near pits than those not near pits, but I was mistaken. Still chooses farther dumps even after I got rid of all garbage dumps near pits.

(0003346) udontknowme (reporter) 2010-04-13 17:57	Afair that isn't new, I realized that in 40d and only use one garbage dump at a time since.

(0003347) Cryten (reporter) 2010-04-13 18:00	This did happen in 40d when the dumps where differnt levels but closer in range ignoring z levels. The simplified pathfinding most likely is to put less strain on the cpu for mundane tasks. Is this the case here?

(0007917) philomory (reporter) 2010-06-08 02:04	This bug really puts a damper on my 'garbage chute' idea, with a vertical tunnel that has a dwarven atom smasher at the bottom. All the dwarves run down the stairs to bottom-most pit (1 z-level above the DAS) and toss their stones into that, rather than the one that the stone was next to, 11 z-levels above :(

(0027293) lightman (reporter) 2014-07-23 00:47	I'm seeing this in 0.40.03: I created a garbage dump zone by some stones that I wanted removed, inside the fortress. I marked the stones for dumping. The dwarves proceeded to take the stones outside and dump them in a hole (not declared as a pit).

(0028803) Footkerchief (manager) 2014-08-11 09:30	Awaiting a save from 0.40.06+.

(0029080) Kirkegaard (reporter) 2014-08-15 01:29	Save from 0.40.8: http://dffd.wimbli.com/file.php?id=9399 [^]

(0030175) Kirkegaard (reporter) 2014-09-13 13:54	@Footkerchief. See save above.

(0030177) Footkerchief (manager) 2014-09-13 14:51	Thanks!

(0033748) Toady One (administrator) 2015-12-10 15:34 edited on: 2015-12-10 15:37	I understand the frustration with Z levels, but it's hard to calculate those distances quickly -- if I deal with that, it'll be part of a large pathing rewrite that has to account for central stairwells in some special way. Are there any saves where this occurs with the horizontal distances? The original report indicates there were, unless they meant 30 tiles over and 30 tiles back.

(0033750) Talvieno (reporter) 2015-12-10 16:21 edited on: 2015-12-10 16:21	Would it not be possible to simply say, for each garbage dump, get distX + distY + distZ, and return the closest? It would ignore walls, and it would be somewhat messy, but it [should] work in a way that's a lot cleaner than how it works now (ignoring z-levels entirely).

(0033792) Toady One (administrator) 2015-12-11 21:34	It doesn't ignore Z levels entirely. That's why I'm asking.

(0033793) Talvieno (reporter) 2015-12-11 23:16	Ah, misunderstood then, apologies.

(0033794) Toady One (administrator) 2015-12-11 23:48	I went ahead and messed around with the job z-calcs a bit, enough to address this problem (I think). Hopefully I didn't break anything else.