Posts: 6,482
	Threads: 176
	Joined: Jan 2012
	
Thanks: 131
	Given 1085 thank(s) in 857 post(s)
	 
	
	
		Slime loading, line 13: you're converting a pointer into a reference, I would expect that not to compile at all; then you're leaking the slime object if it doesn't load successfully. Use the same technique as in skeleton-loading - auto_ptr. I don't see why the skeleton loading should fail, no apparent problems there. Maybe push it to a branch so that I can give it a try with the debugger.
	
	
	
	
	
 
 
	
	
	
		
	Posts: 954
	Threads: 16
	Joined: May 2013
	
Thanks: 68
	Given 112 thank(s) in 93 post(s)
	 
	
		
		
		11-16-2013, 10:15 PM 
(This post was last modified: 11-16-2013, 10:16 PM by tigerw.)
		
	 
	
		Code #2:
It runs, I can sometimes even see the skeleton, but it immediately stops responding and after a while the deadlock detector asserts. I'll push to a branch soon, thanks!
	
	
	
	
	
 
 
	
	
	
		
	Posts: 6,482
	Threads: 176
	Joined: Jan 2012
	
Thanks: 131
	Given 1085 thank(s) in 857 post(s)
	 
	
	
		So is a_Entity a pointer or a reference? On line 1 you're using it as a pointer, on line 5 as a reference. If it's a pointer, cast it to cMonster * and it should work.
	
	
	
	
	
 
 
	
	
	
		
	Posts: 6,482
	Threads: 176
	Joined: Jan 2012
	
Thanks: 131
	Given 1085 thank(s) in 857 post(s)
	 
	
	
		The mobsaving code was a mess, mixing references with pointers. I've fixed it and it seems to work; didn't test much though.
	
	
	
	
	
 
 
	
	
	
		
	Posts: 1,162
	Threads: 68
	Joined: Mar 2013
	
Thanks: 245
	Given 128 thank(s) in 101 post(s)
	 
	
	
		I tested it and it's working very well. Why don't merge it?