0005732: Site finder doesn't see flux/coal if metal ores are present

(0021919)
dree12 (reporter)
2012-03-30 17:05

Limestone is a flux stone, so the site finder seems to be correct.

(0021920)
the_game_hunt (reporter)
2012-03-30 17:13
edited on: 2012-03-30 17:13

It says the map HASNT any flux stone. However after embark the first stone layer IS limestone.

(0021929)
Footkerchief (manager)
2012-03-30 20:01

Reminder sent to: dree12

Tags like Intentional/Expected are for manager use only.

(0021930)
Footkerchief (manager)
2012-03-30 20:03

I've attempted to clarify the description. Need info about whether other people can reproduce this.

(0021959)
dree12 (reporter)
2012-03-31 15:13

(Sorry about the Intentional/Expected thing, it isn't documented or enforced anywhere and I've used it in the past.)

I still don't quite understand the steps to reproduce: it sounds like one should select Flux stone: NO on the site finder, and be able to find a site without flux - but the site still has flux?

(0021971)
Kogut (reporter)
2012-04-01 01:08

Exactly.

(0021972)
Maristo (reporter)
2012-04-01 01:15

I just fired up DF 0.34.07 for the first time. After generating 5 worlds, I have yet to encounter a world that has flux stone in it, according to the site finder. I have not embarked to determine whether or not flux stone actually exists in the world, but considering I've generated several large worlds with minerals set to "Everywhere" I think it's reasonable to expect that there be some flux stone somewhere. Regardless, the site finder always indicates that there is none.

(0021976)
Kogut (reporter)
2012-04-01 02:51
edited on: 2012-04-01 02:51

minerals set to "Everywhere" may result in killing flux as flux is not a ore

(0022013)
Steb (reporter)
2012-04-02 04:27

Perhaps the finder is indicating flux deposits but not flux layers? If Kogut is correct and minerals "Everywhere" kills off flux deposits then the only flux available would be where it forms a whole layer.

(0022014)
Kogut (reporter)
2012-04-02 04:31

Somebody posted on forums that his/her world returned 0 hits for places with flux and 0 hits for places without flux.

(0022068)
tatterdemalian (reporter)
2012-04-03 21:16

Had that bug too.

Site finder => Flux: Yes, No Match
Site finder => Flux: No, No Match
DFHack => Prospect: 300k chalk, 200k marble
Embark: Chalk everywhere.

(0023070)
Thoranius (reporter)
2012-06-20 21:26

Just tested and confirmed this bug still exists in version 34.11.

(0023118)
pedrolance (reporter)
2012-06-28 02:32

This issue happens to me when following the Getting Started with Dwarf Fortress book steps which leds to not being able to find a single suitable place using the finder function.

(0023130)
AlienDjinn (reporter)
2012-06-30 10:28

Yeah, just got the O'Reilly book yesterday and I'm having the same issue following the tutorial. IT doesn't find any Flux to build a starting location.

(0023172)
drakythe (reporter)
2012-07-06 06:16

I too am in the O'Reilly boat. I went ahead and turned minerals down to the level below the max, and now the site finder can find Flux quite easily.

(0023983)
Quietust (reporter)
2013-05-27 12:57
edited on: 2013-05-27 13:01

I've done a bit of disassembly diving, and I think I see what's going wrong here:

When you're browsing embark locations, it displays a list of attributes (Shallow metal[s], Deep metal[s], Soil, Clay, Aquifer, etc.), and one of those attributes is "Flux". As it counts up metal ores, it first checks the layer material itself (in case you modded a layer stone to be smeltable into metal) and then all of the inclusions within it, and once it's done that it checks the "material" for [REACTION_CLASS:FLUX]; unfortunately, the material it checks is the last one it checked for metal ores, so it will only report Flux being present if either the last vein stone in the list is Calcite (which is unlikely) or if there are zero inclusions and the layer is made of flux.

It is likely that the site finder uses the same logic and thus has the same bug.

Effectively, this is what the code appears to be doing:



foreach geo_biome->layers as layer {

   inorganic_material *mat;

   mat = LoadMaterial(layer->material, layer->matgloss);

   if (mat->metal_ores.size()) { increment shallow_metal or deep_metal depending on depth; }

   foreach layer->veins as vein {

      /* this overwrites the above variable */

      mat = world.LoadMaterial(vein->material, vein->matgloss);

      if (mat->metal_ores.size()) { increment shallow_metal or deep_metal depending on depth; }

   }

   /* if there were any veins above, "mat" still refers to the last one, otherwise it refers to the layer itself */

   if (mat->reaction_class == "FLUX") { has_flux = 1; }

   // also check if the layer material is made of clay

}

(0024004)
Xolroc (reporter)
2013-06-14 08:07

If it is that simple, then all that Toady (or a person writing a binary patch) would have to do is move that check for flux into the foreach, right? Before the mat = world.LoadMaterial(...); so that it'll check the layerstone as well. Though I think that might leave out the very last vein..? So perhaps copy it, and leave the original check where it is.

Issue History
Date Modified	Username	Field	Change
2012-03-30 17:00	the_game_hunt	New Issue
2012-03-30 17:05	dree12	Note Added: 0021919
2012-03-30 17:05	dree12	Tag Attached: Intentional/Expected
2012-03-30 17:13	the_game_hunt	Note Added: 0021920
2012-03-30 17:13	the_game_hunt	Note Edited: 0021920	View Revisions
2012-03-30 17:16	the_game_hunt	Tag Attached: site finder
2012-03-30 20:01	Footkerchief	Issue Monitored: dree12
2012-03-30 20:01	Footkerchief	Note Added: 0021929
2012-03-30 20:01	Footkerchief	Tag Detached: Intentional/Expected
2012-03-30 20:02	Footkerchief	Description Updated	View Revisions
2012-03-30 20:03	Footkerchief	Note Added: 0021930
2012-03-31 15:13	dree12	Note Added: 0021959
2012-03-31 15:53	Logical2u	Note Added: 0021960
2012-03-31 15:54	Logical2u	Note Deleted: 0021960
2012-04-01 01:08	Kogut	Note Added: 0021971
2012-04-01 01:15	Maristo	Note Added: 0021972
2012-04-01 02:51	Kogut	Note Added: 0021976
2012-04-01 02:51	Kogut	Note Edited: 0021976	View Revisions
2012-04-01 09:08	Footkerchief	Summary	Site finder doesnt have correct flux stone detectation => Site finder too pessimistic about presence of flux stone
2012-04-02 04:27	Steb	Note Added: 0022013
2012-04-02 04:31	Kogut	Note Added: 0022014
2012-04-02 07:40	Beeskee	Note Added: 0022029
2012-04-02 07:54	Beeskee	Issue Monitored: Beeskee
2012-04-03 02:23	Beeskee	Note Deleted: 0022029
2012-04-03 06:27	Dame de la Licorne	Issue Monitored: Dame de la Licorne
2012-04-03 21:16	tatterdemalian	Note Added: 0022068
2012-05-15 07:19	Footkerchief	Relationship added	has duplicate 0005885
2012-06-15 09:05	Footkerchief	Relationship added	related to 0002440
2012-06-15 09:06	Footkerchief	Relationship added	has duplicate 0006024
2012-06-15 09:06	Footkerchief	Issue Monitored: Zara
2012-06-15 09:09	Footkerchief	Summary	Site finder too pessimistic about presence of flux stone => Setting mineral occurrence to EVERYWHERE results in flux/coal getting crowded out
2012-06-15 09:10	Footkerchief	Relationship added	has duplicate 0004425
2012-06-20 21:26	Thoranius	Note Added: 0023070
2012-06-28 02:32	pedrolance	Note Added: 0023118
2012-06-30 10:28	AlienDjinn	Note Added: 0023130
2012-07-01 12:27	AlienDjinn	Issue Monitored: AlienDjinn
2012-07-06 06:16	drakythe	Note Added: 0023172
2013-05-27 12:57	Quietust	Note Added: 0023983
2013-05-27 13:01	Quietust	Note Edited: 0023983	View Revisions
2013-06-14 08:07	Xolroc	Note Added: 0024004
2013-11-25 15:00	Dwarfu	Sticky Issue	No => Yes
2013-11-25 15:03	Dwarfu	Additional Information Updated	View Revisions
2013-11-25 15:03	Dwarfu	Additional Information Updated	View Revisions
2013-11-29 14:53	Footkerchief	Tag Attached: binary patch
2014-01-15 14:46	Kirig Stonebeard II	Issue Monitored: Kirig Stonebeard II
2014-01-17 10:08	Kirig Stonebeard	Issue Monitored: Kirig Stonebeard
2014-03-25 13:40	Dwarfu	Assigned To	=> Dwarfu
2014-03-25 13:40	Dwarfu	Status	new => acknowledged
2014-07-13 17:49	Footkerchief	Relationship added	has duplicate 0007241
2014-07-13 17:50	Footkerchief	Summary	Setting mineral occurrence to EVERYWHERE results in flux/coal getting crowded out => Site finder doesn't see flux/coal if metal ores are present
2014-07-18 11:55	Toady One	Status	acknowledged => resolved
2014-07-18 11:55	Toady One	Fixed in Version	=> Next Version
2014-07-18 11:55	Toady One	Resolution	open => fixed
2014-07-18 11:55	Toady One	Assigned To	Dwarfu => Toady One
2014-07-18 13:01	Dwarfu	Sticky Issue	Yes => No
2014-07-18 16:51	Dame de la Licorne	Issue End Monitor: Dame de la Licorne
2014-07-26 12:07	Dwarfu	Relationship added	related to 0004451

Notes
(0021919) dree12 (reporter) 2012-03-30 17:05	Limestone is a flux stone, so the site finder seems to be correct.

(0021920) the_game_hunt (reporter) 2012-03-30 17:13 edited on: 2012-03-30 17:13	It says the map HASNT any flux stone. However after embark the first stone layer IS limestone.

(0021929) Footkerchief (manager) 2012-03-30 20:01	Reminder sent to: dree12 Tags like Intentional/Expected are for manager use only.

(0021930) Footkerchief (manager) 2012-03-30 20:03	I've attempted to clarify the description. Need info about whether other people can reproduce this.

(0021959) dree12 (reporter) 2012-03-31 15:13	(Sorry about the Intentional/Expected thing, it isn't documented or enforced anywhere and I've used it in the past.) I still don't quite understand the steps to reproduce: it sounds like one should select Flux stone: NO on the site finder, and be able to find a site without flux - but the site still has flux?

(0021971) Kogut (reporter) 2012-04-01 01:08	Exactly.

(0021972) Maristo (reporter) 2012-04-01 01:15	I just fired up DF 0.34.07 for the first time. After generating 5 worlds, I have yet to encounter a world that has flux stone in it, according to the site finder. I have not embarked to determine whether or not flux stone actually exists in the world, but considering I've generated several large worlds with minerals set to "Everywhere" I think it's reasonable to expect that there be some flux stone somewhere. Regardless, the site finder always indicates that there is none.

(0021976) Kogut (reporter) 2012-04-01 02:51 edited on: 2012-04-01 02:51	minerals set to "Everywhere" may result in killing flux as flux is not a ore

(0022013) Steb (reporter) 2012-04-02 04:27	Perhaps the finder is indicating flux deposits but not flux layers? If Kogut is correct and minerals "Everywhere" kills off flux deposits then the only flux available would be where it forms a whole layer.

(0022014) Kogut (reporter) 2012-04-02 04:31	Somebody posted on forums that his/her world returned 0 hits for places with flux and 0 hits for places without flux.

(0022068) tatterdemalian (reporter) 2012-04-03 21:16	Had that bug too. Site finder => Flux: Yes, No Match Site finder => Flux: No, No Match DFHack => Prospect: 300k chalk, 200k marble Embark: Chalk everywhere.

(0023070) Thoranius (reporter) 2012-06-20 21:26	Just tested and confirmed this bug still exists in version 34.11.

(0023118) pedrolance (reporter) 2012-06-28 02:32	This issue happens to me when following the Getting Started with Dwarf Fortress book steps which leds to not being able to find a single suitable place using the finder function.

(0023130) AlienDjinn (reporter) 2012-06-30 10:28	Yeah, just got the O'Reilly book yesterday and I'm having the same issue following the tutorial. IT doesn't find any Flux to build a starting location.

(0023172) drakythe (reporter) 2012-07-06 06:16	I too am in the O'Reilly boat. I went ahead and turned minerals down to the level below the max, and now the site finder can find Flux quite easily.

(0023983) Quietust (reporter) 2013-05-27 12:57 edited on: 2013-05-27 13:01	I've done a bit of disassembly diving, and I think I see what's going wrong here: When you're browsing embark locations, it displays a list of attributes (Shallow metal[s], Deep metal[s], Soil, Clay, Aquifer, etc.), and one of those attributes is "Flux". As it counts up metal ores, it first checks the layer material itself (in case you modded a layer stone to be smeltable into metal) and then all of the inclusions within it, and once it's done that it checks the "material" for [REACTION_CLASS:FLUX]; unfortunately, the material it checks is the last one it checked for metal ores, so it will only report Flux being present if either the last vein stone in the list is Calcite (which is unlikely) or if there are zero inclusions and the layer is made of flux. It is likely that the site finder uses the same logic and thus has the same bug. Effectively, this is what the code appears to be doing: foreach geo_biome->layers as layer { inorganic_material mat; mat = LoadMaterial(layer->material, layer->matgloss); if (mat->metal_ores.size()) { increment shallow_metal or deep_metal depending on depth; } foreach layer->veins as vein { / this overwrites the above variable / mat = world.LoadMaterial(vein->material, vein->matgloss); if (mat->metal_ores.size()) { increment shallow_metal or deep_metal depending on depth; } } / if there were any veins above, "mat" still refers to the last one, otherwise it refers to the layer itself */ if (mat->reaction_class == "FLUX") { has_flux = 1; } // also check if the layer material is made of clay }

(0024004) Xolroc (reporter) 2013-06-14 08:07	If it is that simple, then all that Toady (or a person writing a binary patch) would have to do is move that check for flux into the foreach, right? Before the mat = world.LoadMaterial(...); so that it'll check the layerstone as well. Though I think that might leave out the very last vein..? So perhaps copy it, and leave the original check where it is.