Map crashes when adding triggers

[Verdun]

I dunno. It's a strange thing I don't really understand right now.

I am making good progress on an elaborate 480x480 map, but I am worried I might lose it.

I imported triggers from earlier versions of the project, on other maps. Some of them work no problem, but others, even after I adapt them, crash the map when I test it, with an error message (something along the lines of "runs out of memory") or the map loads and after a few seconds it goes, in a word, bonkers - worker units become unable to harvest, and may even pop in and out of existence at random positions. Deleting or disabling the last trigger I worked on usually restores the map to functionality.

A key problem seems to be map initialization triggers - anything other than the minimum standard trigger kills the map.

One work-around I found is delay a bunch of triggers, have them start 5 seconds after intilialization. That, the map can usually handle. But yesterday I had to delete one trigger completely.

Thing is, I am a bit worried because right now I am fixing things "blind", not really knowing what's going on and going by trial and error, not knowing why the things that seem to be helping are helping. I only work with what I think are "simple" GUI triggers that worked fine on their original maps.

It slways seems to be linked to the trigger editor - another weird crash I got was when I tried to replace a region variable by selecting a new region on the map, and I got another "ran out of memory!!" scream in the world editor itself, not even a test run.
(When I repeated the same action after reloading the map, the editor didn't complain.)

So my question is, any ideas for trouble shooting, or what the core issue may be?
Could it be that the map itself is corrupt in some way, and how would one find out?

 

[Verdun]

Another weird issue I fear might be related is that a bunch of abilities I imported via object data import are not working, even though in the object editor they look fine. Worse, there are abilities that are the same in almost every way (example: head attachment effect rather than chest attachment effect) and one "survives" import and the other doesn't. I really fear there might be some general bug.

 

[Pyrogasm]

I can't speak to actually resolving your issues, but my advice to make a rolling backup of your map starting now if you suspect instability with the file. Using the 'map as a folder' option for the map may help resolve any weird crashes from imports behaving poorly.

 

[Verdun]

I can't speak to actually resolving your issues, but my advice to make a rolling backup of your map starting now if you suspect instability with the file. Using the 'map as a folder' option for the map may help resolve any weird crashes from imports behaving poorly.

Yeah backups are my lifeline, I believe I already have over 30 versions of the map or so. I just hope it's not some fundamental corruption that already set in in, like, version 10; I kept adding and working on triggers and object data and imported files while continuing to place hundreds of elements (though I suppose worst-case, the doodads and units might be roughly salvageable with copy-paste into an older version).

 

[Verdun]

Anway, here is an example of a trigger that crashes the map as described, if turned on.
What the trigger does, or is supposed to do, and did correctly in its old home, was that a little cloud dummy unit with locust ability, flying between the north and south poles, creates an appropriate weather effect whenever it hits a specific climate zone.
The regions have been recreated in the new map.

• Bad Weather Tropical
  • Events
    • Unit - A unit enters Tropical to Equatorial <gen>
  • Conditions
    • (Unit-type of (Entering unit)) Equal to Raincloud
  • Actions
    • Unit - Kill Thunderhag
    • Unit - Kill Tornadohag
    • Custom script: call RemoveLocation(udg_Fleeceposition)
    • Set Fleeceposition = (Position of (Entering unit))
    • Region - Center Rainstorm <gen> on Fleeceposition
    • If (All Conditions are True) then do (Then Actions) else do (Else Actions)
      • If - Conditions
        • (Random real number between 0.00 and 10.00) Greater than or equal to 5.00
      • Then - Actions
        • Unit - Create 1 Lightning for (Owner of (Entering unit)) at Fleeceposition facing Default building facing degrees
        • Set Thunderhag = (Last created unit)
      • Else - Actions
    • Custom script: call RemoveLocation(udg_Fleeceposition)
    • Environment - Turn (Last created weather effect) Off
    • Environment - Remove (Last created weather effect)
    • If (All Conditions are True) then do (Then Actions) else do (Else Actions)
      • If - Conditions
        • (Random real number between 0.00 and 10.00) Greater than or equal to 4.00
      • Then - Actions
        • Environment - Create at Rainstorm <gen> the weather effect Ashenvale Rain (Heavy)
      • Else - Actions
        • Environment - Create at Rainstorm <gen> the weather effect Dungeon White Fog (Heavy)
    • Environment - Turn (Last created weather effect) On

UPDATE: I played around with the "Tropical to Equatorial" region, most notably, I made it smaller. That saved the trigger.
I tried to see if any specific sector of the region was problematic, maybe because of one faulty unit or so. But no, it's literally just the size. Maybe it gets triggered too often if it is too large? But "Turn off this trigger" at the start didn't help.

So I decided, okay, then I make all climate zones smaller. But no, the map cannot seem to handle more than one climate zone trigger even existing side by side. Really weird.

Re-UPDATE: I turned the regions into very thin tripwires, and now they all work. The map couldn't for some reason monitor multiple huge regions.
Let's see if I can avoid other glitches too.