[http://sourceforge.net/p/ufoai/bugs/2076 Item 2076] imported from sourceforge.net tracker on 2013-01-28 19:18:54

SVN Rev 23538

Run down of the problem:

Start mission and return to main menu (escape) and try the retry option.

This will cause game to return to world view then crash with assert(item->t).

I have found a crude method around this but will look into the real cause as it could be the cause to some other problems in the future.

Method around it:

/base/ufos/menu_singleplayer.ufo

Line 129
- click { cmd "game_go;" }
+ click { cmd "sv win 7;" }

This basicly makes the menu->retry end the mission with a failure (i believe thats what original xcom did). Then able to click retry from this menu to reload mission without any faults.


Below attached gdb bt dump:
http://pastebin.com/m70305bb0

ufoconsole.log file is to large to upload. if requested I can post it or trunk it.

Side note:
I did revert SVN to 23530 (effectively) and same problem occured.

===== Comments Ported from Sourceforge =====

====== rudolfowood (2009-03-16 19:42:34) ======

I can provide another way to reproduce such a crash on item->t: load attached savegame, start and play mission, after won (e.g. with debug_killteam 7) after returning to geoscape reload the save and try to start mission again - will assert on item->t.
File Added: slot2.sav
====== itsdrone (2009-03-17 01:21:49) ======

While trying to track this down some today I determined, atleast for myself, this will only occur if [cvar]developer != 0.

No matter how I try I cannot get it to fail on assert(item->t) if developer is turned off. It also doesn't relate to any specific subclass of developer logging. (any value other then 0 will cause it).


====== itsdrone (2009-03-17 02:00:01) ======

Resolution found...

In fn SCR_BeginLoadingPlaque developer variable was being tested for any value at all and would break out the function before doing anything if found to be of any value other then 0.

Removal of this test/break out cleared any assert(item->t) induced from running game_go commands. (mission retry from main menu, load new mission)

attached is svn diff from # 23542

I'm not quite sure why it was in the function in the first place. no comments as to why it was there.

File Added: error.diff
====== tlh2000 (2009-03-19 06:42:37) ======

i've removed the developer check - i'm absolutly not sure why this should fix it - but it works for me and the check was from the "old days" (tm)
====== sf-robot (2009-04-03 02:20:09) ======

This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).