[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Update of bug #13679 (project freeciv): Release: => 2.3.0 Planned Release: => 2.3.1,2.4.0 ___ Reply to this item at:

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Follow-up Comment #8, bug #13679 (project freeciv): > Note that autogame regression testing is made harder by event > cache where timestamps are saved. Is there a way to disable > this feature while doing autogame testing, or is one supposed > to filter it out from savegames afterwards? Apart

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Update of bug #13679 (project freeciv): Category:None => general Status: Ready For Test => Fixed Assigned to:None => cazfi Open/Closed:

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Update of bug #13679 (project freeciv): Status:None => Ready For Test ___ Reply to this item at: ___ Message sen

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Additional Item Attachment, bug #13679 (project freeciv): File name: rt.servSize:0 KB ___ Reply to this item at: ___ Message sent via/b

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Follow-up Comment #6, bug #13679 (project freeciv): No, I don't use scenarios. Problem is very much reproducible here. I'm attaching autogame, though I don't see how some seeds could trigger the bug and others not. Number of players do matter. More players -> more startpositions generated -> more

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Follow-up Comment #5, bug #13679 (project freeciv): I did a big test with defined mapseed and gameseed today - besides the name of the savegame all 11 runs are identical at the end (turn > 550). Do you start using a scenario (start positions?) (file #11941) __

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Follow-up Comment #4, bug #13679 (project freeciv): I debugged this to find at least one of the problems. When starting positions are inserted to hash, tile pointer (memory address) is used as key -> different bucket is used when tile memory is allocated from different address -> iteration over

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Follow-up Comment #3, bug #13679 (project freeciv): you could disable the event cache - 'set ec_turn 0' ___ Reply to this item at: ___ Nachricht geschickt von/d

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Follow-up Comment #2, bug #13679 (project freeciv): I still get randomly different savegames with identical seeds. Note that autogame regression testing is made harder by event cache where timestamps are saved. Is there a way to disable this feature while doing autogame testing, or is one suppos

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

Follow-up Comment #1, bug #13679 (project freeciv): Does it still occur in current svn? I didn't observe this. May it has been fixed for bug #14415? ___ Reply to this item at: ___

[Freeciv-Dev] [bug #13679] Autogames with same gameseed & mapseed differ

URL: Summary: Autogames with same gameseed & mapseed differ Project: Freeciv Submitted by: cazfi Submitted on: Friday 06/12/2009 at 22:23 Category: None Severity: 3 - Normal