0000629: Crash from socks designated for melting

(0001464)
Logical2u (manager)
2010-04-06 08:46

Please upload the save to the Dwarf Fortress File Depot, at http://dffd.wimbli.com/index.php [^]

Might be related to 0000536.

(0001466)
derigo (reporter)
2010-04-06 08:48
edited on: 2010-04-06 09:07

save located at

http://dffd.wimbli.com/file.php?id=2009 [^]

Incidentally, this save also has some plant fiber socks that somehow got themselves designated for melting a la http://bay12games.com/dwarves/mantisbt/view.php?id=609 [^]

But I don't think thats related to the crash, they've been around for years.

(0001499)
Logical2u (manager)
2010-04-06 09:47

Alright, I also get a crash on my laptop, and it doesn't appear to be related to 0000536 - very different crash notifications. (Windows 7 x64, Core 2 Duo T8300, 4gigs of Ram)

Can you provide your OS, processor, and RAM, just for recordkeeping?

(0001507)
derigo (reporter)
2010-04-06 10:13

winxp
AMD Athlon 64x2 DUAL 5000+ 2.6GHz
2gigs of ram

(0001604)
derigo (reporter)
2010-04-06 15:26

Hate to bug you again Logical2u, but I have an update on this I'd like you to verify, if you don't mind.

I just noticed this saved game will also crash for me if any attempt is made to abandon the fort. So effectively, this world is dead for me.

(0001606)
Logical2u (manager)
2010-04-06 15:41

No problem good sir, however I was able to successfully abandon the world - would you like the save?

(Might have crashed due to memory limitations - it lagged on my computer, so I hesitate to think what might've happened to you)

(0001689)
derigo (reporter)
2010-04-06 23:19

Interesting. I tried to abandon 4 times, it didn't seem to lag, it just crashed immediately, without hesitation.

No thanks on the save. I've... moved on. *tear*

(0001966)
derigo (reporter)
2010-04-07 23:44

Argh.

Different fortress, different world, almost identical situation. Fortresses are about the same age, about the same population, and they both have inexplicable melt designated socks.

There's no crash on abandon this time, and the crash time seems a little more consistent than before. (its bout a minute)

save is here:
http://dffd.wimbli.com/file.php?id=2027 [^]

(0002152)
derigo (reporter)
2010-04-08 13:54

just tested to see if this save still crashes in the new version. If anything, it crashes faster.

(0003371)
derigo (reporter)
2010-04-13 19:54

just an update, crash persists in 31.03

(0003374)
Arkaaito (reporter)
2010-04-13 20:05

Sorry if this has already been checked, but is the memory usage exactly at the 2GB mark? If so, the cause is a failure to allocate additional space - tag w/memory.

(0003426)
derigo (reporter)
2010-04-14 01:21
edited on: 2010-04-14 01:51

No, memory use hovers around 500mb. Plenty left over.

Also this crash was verified (by Logical2u) on a box with 4gb ram.

(0003433)
derigo (reporter)
2010-04-14 02:15

Ok, I think I may have tracked down the underlying cause of these crashes. My evidence is anecdotal at best, but its better than none. I believe this crash is caused by inappropriate melt-designation of items which cannot be melted, or the subsequent attempts of the smelter to melt them as descibed in

http://www.bay12games.com/dwarves/mantisbt/view.php?id=609 [^]

So far all of my forts have become unstable and crash in the manner of this bug report around year 2. Now around year 2, you get a big migration bringing your population up to around 60(assuming things have been going well). As a result, I tend to initiate a bunch of projects:

-I start smoothing my fort out with my Engineering Corps.
-My leatherworker and clothier are usually Masters or close by this point, so they start making high quality clothing for soldiers to wear in layers via lots of manager orders.
-I finally have enough smelters to start training my weapon and armor smiths in earnest, so they start cranking out iron leggings and iron corkscrews.
-As they do this, I usually designate the products for melting, and add melting to each of the smelters.

I went through a new fort and did the same stuff I usually do. But when I got to this point I proceeded very carefully. I started these large projects one at a time, and let them go for a while before starting the next. When the crash came, it was a few months after I had started the major melting operation.

So, that's my guess. This crash consistently happens to all of my forts, and effectively ends the game at year 2. So either my system generates faulty forts, or something specific I'm doing causes this. Melting is my guess.

(0003841)
DwarfNinja (reporter)
2010-04-16 06:52
edited on: 2010-04-16 06:53

I'm having a similar crash around that population level and time. However I don't have any melt designations, the only weird thing is that I don't have a mayor designated. The dwarves haven't elected one and I haven't been able to appoint one.

This is the second fort that has reached this point and begun crashing at or near that big wave of migrants. Population goes from 40 to 60 and then it crashes. Sometimes it doesn't even get the migrants but it still crashes.

(0003842)
Glaucus (reporter)
2010-04-16 07:10

http://www.bay12games.com/dwarves/mantisbt/view.php?id=861 [^]

(0003934)
JimboOmega (reporter)
2010-04-16 20:33

My version of this (and many other bugs) - crashes in within a minute on both computers I tried it on.

http://dffd.wimbli.com/file.php?id=2127 [^]

(0004424)
Nimrod (reporter)
2010-04-21 01:57

I can reproduce a save crash when i got some (ex goblin) copper items (armor) marked for melting. I removed the melting tag - no more save crash

(0004439)
JimboOmega (reporter)
2010-04-21 07:01

The crash on save/abandon appears to be a different problem - based on Nimrod's suggestion I removed melt designations and can save normally.

However, I still experience this issue - where my fortress crashes after running a brief time.

My memory use also hovers around 500MB when this happens. I wonder if that is related. Also the crash in my save happens just seconds after "No migrants have arrived this season" shows up. Strange.

(0004442)
Langdon (reporter)
2010-04-21 07:14

Just adding, I have this 0.31.03 game that also crashes consistently on saving (bug 0000277) and on abandon: http://dffd.wimbli.com/file.php?id=2185 [^]

I went fairly slow on this fort, and only reached the magma sea by year 6, so there were no melting problems (as reported by Nimrod) with this fort (my previous forts were beside magma pipes, so I had magma smelters early).

My 0.31.02 fort previous to this one abandoned properly, so this is the first save I have with the crash on abandon bug.

(0004784)
derigo (reporter)
2010-04-23 15:01

this is a child of 0001046

(0004788)
Footkerchief (manager)
2010-04-23 15:17

derigo: I'm still not clear on whether the crash from melt-designated socks has the same cause as the other melting crashes. Right now, I have this as a child of 0000609, and 0000609 related to 0001046. I may reorganize it later -- it's still not clear what's going on.

(0004798)
derigo (reporter)
2010-04-23 16:39

I'm fairly certain the crashes in these saves simply result from the fact that I have a major melting operation going on in these forts, and lots of people have confirmed that melting causes crashes. I suspect that the 'melt designated socks' and other stuff that's not supposed to be designated is merely a side-effect of the same bug. It doesn't, by itself, cause any crashing, but it probably will provide toady with some good clues as to what's actually happening. Furthermore, I suspect that if the other reporters of the 'melt orders causing a crash' crash were to inspect their stocks closely, they would probably find some items designated for melting that shouldn't be. It's not the kind of thing you notice easily if you're not looking for it.

(0004805)
Footkerchief (manager)
2010-04-23 18:00

There's a good chance you're right. At any rate, these reports are all linked together, so Toady should see them all at once -- he usually changes the relationships anyway after he figures out what's really going on, so I'm not going to worry about it.

(0006245)
Langdon (reporter)
2010-05-05 22:45

@derigo: My personal opinion is it is a combination of the melting bugs and military/arsenal dwarf bugs. In my last fort (save linked above) I only melted weapons and sold armor to the caravans, while making sure that all my soldiers were set to use specifically steel battleaxes (as I wanted to avoid the crash due to dwarves using large daggers). I avoided crashes as long as I never melted any armor, but once I designated some goblin armor for melting, I see the crashes.

I believe the arsenal dwarf re-assigns armor/weapons every season or so (or when he believes new items are available?), but he might be selecting items assigned for melting, which crashes the game once a dwarf attempts to melt said assigned item. I do need to run more tests (assign specifically steel weapons/armor to dwarves, then only melt iron/copper armor) to make sure.

(0006254)
derigo (reporter)
2010-05-06 00:06

Sorry, but I don't think so. When I'm training my weaponsmith, normally, the most cost efficient thing to craft (if you can melt) is enormous corkscrews, a trap component, not a weapon. I've experienced both the incorrect melt designation and the crash while attempting to melt just corkscrews, nothing else. It is possible that the unauthorized melt designation is choosing items that are already being assigned or worn, but I don't see why even that would cause a crash.

That being said, the arsenal dwarf certainly does cause some darn strange stuff to happen. All things considered, it would be unwise to rule anything out where he is concerned, as he's clearly a large source of odd/buggy behavior. Programming bugs are a lot like real ones in that respect, where you see one, there's bound to be a whole nest.

(0007946)
Footkerchief (manager)
2010-06-08 08:57

Seems pretty certain by now that this is a duplicate of 0000609.

Issue History
Date Modified	Username	Field	Change
2010-04-06 08:33	derigo	New Issue
2010-04-06 08:46	Logical2u	Note Added: 0001464
2010-04-06 08:48	derigo	Note Added: 0001466
2010-04-06 08:49	derigo	Note Edited: 0001466	View Revisions
2010-04-06 09:07	derigo	Note Edited: 0001466	View Revisions
2010-04-06 09:47	Logical2u	Note Added: 0001499
2010-04-06 10:13	derigo	Note Added: 0001507
2010-04-06 15:26	derigo	Note Added: 0001604
2010-04-06 15:41	Logical2u	Note Added: 0001606
2010-04-06 23:19	derigo	Note Added: 0001689
2010-04-07 23:44	derigo	Note Added: 0001966
2010-04-08 11:09	derigo	Tag Attached: crash
2010-04-08 13:54	derigo	Note Added: 0002152
2010-04-13 19:54	derigo	Note Added: 0003371
2010-04-13 20:05	Arkaaito	Note Added: 0003374
2010-04-13 21:55	derigo	Note Added: 0003401
2010-04-14 01:21	derigo	Note Added: 0003426
2010-04-14 01:44	derigo	Note Deleted: 0003401
2010-04-14 01:51	derigo	Note Edited: 0003426	View Revisions
2010-04-14 02:15	derigo	Note Added: 0003433
2010-04-16 06:52	DwarfNinja	Note Added: 0003841
2010-04-16 06:53	DwarfNinja	Note Edited: 0003841	View Revisions
2010-04-16 07:10	Glaucus	Note Added: 0003842
2010-04-16 20:33	JimboOmega	Note Added: 0003934
2010-04-17 03:30	alexandertnt	Tag Attached: cyclops
2010-04-17 03:30	alexandertnt	Tag Detached: cyclops
2010-04-19 18:41	Footkerchief	Relationship added	child of 0000609
2010-04-20 09:39	Footkerchief	Summary	inexplicable crash => Crash on save/abandon
2010-04-20 09:39	Footkerchief	Relationship added	parent of 0001367
2010-04-21 01:57	Nimrod	Note Added: 0004424
2010-04-21 07:01	JimboOmega	Note Added: 0004439
2010-04-21 07:12	JimboOmega	Issue Monitored: JimboOmega
2010-04-21 07:14	Langdon	Note Added: 0004442
2010-04-23 15:01	derigo	Note Added: 0004784
2010-04-23 15:12	Footkerchief	Summary	Crash on save/abandon => Crash from socks designated for melting
2010-04-23 15:15	Footkerchief	Relationship deleted	parent of 0001367
2010-04-23 15:17	Footkerchief	Note Added: 0004788
2010-04-23 16:39	derigo	Note Added: 0004798
2010-04-23 18:00	Footkerchief	Note Added: 0004805
2010-05-05 22:45	Langdon	Note Added: 0006245
2010-05-06 00:06	derigo	Note Added: 0006254
2010-06-08 08:57	Footkerchief	Note Added: 0007946
2010-06-08 08:57	Footkerchief	Relationship replaced	duplicate of 0000609
2010-06-08 08:57	Footkerchief	Status	new => resolved
2010-06-08 08:57	Footkerchief	Resolution	open => duplicate
2010-06-08 08:57	Footkerchief	Assigned To	=> Footkerchief
2010-06-09 06:45	Toady One	Status	resolved => closed

Notes
(0001464) Logical2u (manager) 2010-04-06 08:46	Please upload the save to the Dwarf Fortress File Depot, at http://dffd.wimbli.com/index.php [^] Might be related to 0000536.

(0001466) derigo (reporter) 2010-04-06 08:48 edited on: 2010-04-06 09:07	save located at http://dffd.wimbli.com/file.php?id=2009 [^] Incidentally, this save also has some plant fiber socks that somehow got themselves designated for melting a la http://bay12games.com/dwarves/mantisbt/view.php?id=609 [^] But I don't think thats related to the crash, they've been around for years.

(0001499) Logical2u (manager) 2010-04-06 09:47	Alright, I also get a crash on my laptop, and it doesn't appear to be related to 0000536 - very different crash notifications. (Windows 7 x64, Core 2 Duo T8300, 4gigs of Ram) Can you provide your OS, processor, and RAM, just for recordkeeping?

(0001507) derigo (reporter) 2010-04-06 10:13	winxp AMD Athlon 64x2 DUAL 5000+ 2.6GHz 2gigs of ram

(0001604) derigo (reporter) 2010-04-06 15:26	Hate to bug you again Logical2u, but I have an update on this I'd like you to verify, if you don't mind. I just noticed this saved game will also crash for me if any attempt is made to abandon the fort. So effectively, this world is dead for me.

(0001606) Logical2u (manager) 2010-04-06 15:41	No problem good sir, however I was able to successfully abandon the world - would you like the save? (Might have crashed due to memory limitations - it lagged on my computer, so I hesitate to think what might've happened to you)

(0001689) derigo (reporter) 2010-04-06 23:19	Interesting. I tried to abandon 4 times, it didn't seem to lag, it just crashed immediately, without hesitation. No thanks on the save. I've... moved on. tear

(0001966) derigo (reporter) 2010-04-07 23:44	Argh. Different fortress, different world, almost identical situation. Fortresses are about the same age, about the same population, and they both have inexplicable melt designated socks. There's no crash on abandon this time, and the crash time seems a little more consistent than before. (its bout a minute) save is here: http://dffd.wimbli.com/file.php?id=2027 [^]

(0002152) derigo (reporter) 2010-04-08 13:54	just tested to see if this save still crashes in the new version. If anything, it crashes faster.

(0003371) derigo (reporter) 2010-04-13 19:54	just an update, crash persists in 31.03

(0003374) Arkaaito (reporter) 2010-04-13 20:05	Sorry if this has already been checked, but is the memory usage exactly at the 2GB mark? If so, the cause is a failure to allocate additional space - tag w/memory.

(0003426) derigo (reporter) 2010-04-14 01:21 edited on: 2010-04-14 01:51	No, memory use hovers around 500mb. Plenty left over. Also this crash was verified (by Logical2u) on a box with 4gb ram.

(0003433) derigo (reporter) 2010-04-14 02:15	Ok, I think I may have tracked down the underlying cause of these crashes. My evidence is anecdotal at best, but its better than none. I believe this crash is caused by inappropriate melt-designation of items which cannot be melted, or the subsequent attempts of the smelter to melt them as descibed in http://www.bay12games.com/dwarves/mantisbt/view.php?id=609 [^] So far all of my forts have become unstable and crash in the manner of this bug report around year 2. Now around year 2, you get a big migration bringing your population up to around 60(assuming things have been going well). As a result, I tend to initiate a bunch of projects: -I start smoothing my fort out with my Engineering Corps. -My leatherworker and clothier are usually Masters or close by this point, so they start making high quality clothing for soldiers to wear in layers via lots of manager orders. -I finally have enough smelters to start training my weapon and armor smiths in earnest, so they start cranking out iron leggings and iron corkscrews. -As they do this, I usually designate the products for melting, and add melting to each of the smelters. I went through a new fort and did the same stuff I usually do. But when I got to this point I proceeded very carefully. I started these large projects one at a time, and let them go for a while before starting the next. When the crash came, it was a few months after I had started the major melting operation. So, that's my guess. This crash consistently happens to all of my forts, and effectively ends the game at year 2. So either my system generates faulty forts, or something specific I'm doing causes this. Melting is my guess.

(0003841) DwarfNinja (reporter) 2010-04-16 06:52 edited on: 2010-04-16 06:53	I'm having a similar crash around that population level and time. However I don't have any melt designations, the only weird thing is that I don't have a mayor designated. The dwarves haven't elected one and I haven't been able to appoint one. This is the second fort that has reached this point and begun crashing at or near that big wave of migrants. Population goes from 40 to 60 and then it crashes. Sometimes it doesn't even get the migrants but it still crashes.

(0003842) Glaucus (reporter) 2010-04-16 07:10	http://www.bay12games.com/dwarves/mantisbt/view.php?id=861 [^]

(0003934) JimboOmega (reporter) 2010-04-16 20:33	My version of this (and many other bugs) - crashes in within a minute on both computers I tried it on. http://dffd.wimbli.com/file.php?id=2127 [^]

(0004424) Nimrod (reporter) 2010-04-21 01:57	I can reproduce a save crash when i got some (ex goblin) copper items (armor) marked for melting. I removed the melting tag - no more save crash

(0004439) JimboOmega (reporter) 2010-04-21 07:01	The crash on save/abandon appears to be a different problem - based on Nimrod's suggestion I removed melt designations and can save normally. However, I still experience this issue - where my fortress crashes after running a brief time. My memory use also hovers around 500MB when this happens. I wonder if that is related. Also the crash in my save happens just seconds after "No migrants have arrived this season" shows up. Strange.

(0004442) Langdon (reporter) 2010-04-21 07:14	Just adding, I have this 0.31.03 game that also crashes consistently on saving (bug 0000277) and on abandon: http://dffd.wimbli.com/file.php?id=2185 [^] I went fairly slow on this fort, and only reached the magma sea by year 6, so there were no melting problems (as reported by Nimrod) with this fort (my previous forts were beside magma pipes, so I had magma smelters early). My 0.31.02 fort previous to this one abandoned properly, so this is the first save I have with the crash on abandon bug.

(0004784) derigo (reporter) 2010-04-23 15:01	this is a child of 0001046

(0004788) Footkerchief (manager) 2010-04-23 15:17	derigo: I'm still not clear on whether the crash from melt-designated socks has the same cause as the other melting crashes. Right now, I have this as a child of 0000609, and 0000609 related to 0001046. I may reorganize it later -- it's still not clear what's going on.

(0004798) derigo (reporter) 2010-04-23 16:39	I'm fairly certain the crashes in these saves simply result from the fact that I have a major melting operation going on in these forts, and lots of people have confirmed that melting causes crashes. I suspect that the 'melt designated socks' and other stuff that's not supposed to be designated is merely a side-effect of the same bug. It doesn't, by itself, cause any crashing, but it probably will provide toady with some good clues as to what's actually happening. Furthermore, I suspect that if the other reporters of the 'melt orders causing a crash' crash were to inspect their stocks closely, they would probably find some items designated for melting that shouldn't be. It's not the kind of thing you notice easily if you're not looking for it.

(0004805) Footkerchief (manager) 2010-04-23 18:00	There's a good chance you're right. At any rate, these reports are all linked together, so Toady should see them all at once -- he usually changes the relationships anyway after he figures out what's really going on, so I'm not going to worry about it.

(0006245) Langdon (reporter) 2010-05-05 22:45	@derigo: My personal opinion is it is a combination of the melting bugs and military/arsenal dwarf bugs. In my last fort (save linked above) I only melted weapons and sold armor to the caravans, while making sure that all my soldiers were set to use specifically steel battleaxes (as I wanted to avoid the crash due to dwarves using large daggers). I avoided crashes as long as I never melted any armor, but once I designated some goblin armor for melting, I see the crashes. I believe the arsenal dwarf re-assigns armor/weapons every season or so (or when he believes new items are available?), but he might be selecting items assigned for melting, which crashes the game once a dwarf attempts to melt said assigned item. I do need to run more tests (assign specifically steel weapons/armor to dwarves, then only melt iron/copper armor) to make sure.

(0006254) derigo (reporter) 2010-05-06 00:06	Sorry, but I don't think so. When I'm training my weaponsmith, normally, the most cost efficient thing to craft (if you can melt) is enormous corkscrews, a trap component, not a weapon. I've experienced both the incorrect melt designation and the crash while attempting to melt just corkscrews, nothing else. It is possible that the unauthorized melt designation is choosing items that are already being assigned or worn, but I don't see why even that would cause a crash. That being said, the arsenal dwarf certainly does cause some darn strange stuff to happen. All things considered, it would be unwise to rule anything out where he is concerned, as he's clearly a large source of odd/buggy behavior. Programming bugs are a lot like real ones in that respect, where you see one, there's bound to be a whole nest.

(0007946) Footkerchief (manager) 2010-06-08 08:57	Seems pretty certain by now that this is a duplicate of 0000609.