tron [Fri, 03 Mar 2006 19:42:09 +0000] rev 3132
(svn r3747) Change HASBIT() to return 0/1 instead of 0/value of tested bit, because the name suggests it does the former and current behavior broke in some places in very subtle ways (for example HASBIT(x, 0) != HASBIT(y, 1) doesn't work, returning a bool after HASBIT(x, 9) neither)
Darkvater [Fri, 03 Mar 2006 11:27:18 +0000] rev 3131
(svn r3746) - [7/7] Moving the callback functions for the setting-changes from settings_gui.c to settings.c. Also enable the callback functionality through the console.
- Fix a nasty bug with improper counter addition. (*i)++ and NOT *i++ when setting a patch value through the console.
tron [Fri, 03 Mar 2006 07:51:41 +0000] rev 3130
(svn r3745) Fix two buffer overflows, one case of undefined behavior (the destination buffer of sprintf() may not alias with one of its arguments) and some other minor stuff introduced in r3740
bjarni [Thu, 02 Mar 2006 22:21:27 +0000] rev 3129
(svn r3742) -codechange: [OSX] changed an assert to use the new error window and added another two to show if the screen buffer fails for some reason in cocoa driver (instead of just crashing)
bjarni [Thu, 02 Mar 2006 22:00:57 +0000] rev 3128
(svn r3741) -Fix: [i686 OSX] reverted the change in rev 3670 for non PowerPC (done with #ifdef) since it appeared to crash intel based macs
This fix is not a good solution and might not work, but if it works, it's better than nothing until we get the real solution
bjarni [Thu, 02 Mar 2006 21:43:09 +0000] rev 3127
(svn r3740) -Feature: [OSX] added OS version, CPU type and newGRF settings to assert window and a message to tell people to add that to a bug report
It also adds a new error window, which can be used just like assert, but it is also active when asserts are turned off
This is useful for places where it's really important to check even if assert is turned off. It's not used in the code yet
bjarni [Thu, 02 Mar 2006 21:27:20 +0000] rev 3126
(svn r3739) -Fix: [Big endian CPUs] fixed bug introduced in rev 3720 where BE computers would read insane map sizes and crash as a result
tron [Thu, 02 Mar 2006 14:45:35 +0000] rev 3125
(svn r3733) Fix a bug in r3719, which caused a crash when the default value for a SLE_VAR_CHAR setting was used
Darkvater [Thu, 02 Mar 2006 09:57:28 +0000] rev 3124
(svn r3732) - Fix two warnings. Stupid MSVC didn't even complain :s (Thank Tron and peter1138)
- NOTE! There are a few warnings left in settings_gui because tha callback functions are disabled. I still need a good place for them.
celestar [Thu, 02 Mar 2006 08:55:12 +0000] rev 3123
(svn r3730) Multistop modifications:
-Codechange: Completely rewritten the slot assignment system. It now consumes less CPU cycles and memory
-Codechange: Increased maximum number of roadstops to 16.
-Fix: Several conditions where a slot becomes unliked from a vehicle
-Codechange: ClearSlot now only takes one parameter, the vehicle
-Feature: Console command 'clearslots' to clear ALL currently assinged slots. debug usage only
-Feature: vehicles that cannot get a slot now wait on the road instead of planlessly blocking stops or circling around
-Codechange: Adjusted debug levels
TODO: Make the slot finder compatible with (a) pathfinder(s).