| Age | Commit message (Collapse) | Author |
|---|
|
|
|
primarily removes the fixed engine type limits and also happens to allow (with the patch option 'dynamic_engines') multiple NewGRF vehicle sets to coexist.
|
|
more files than needed
|
|
more files than needed
|
|
code.
|
|
|
|
generate.vbs script.
|
|
|
|
generate script.
|
|
project files.
|
|
place to a single location.
|
|
space as necessary. This avoids copying data around for vector's push_back() function.
|
|
threading we have better extendable.
|
|
compiler's decisions about inlining
|
|
|
|
|
|
by default
|
|
openttd.h) to engine_type.h.
|
|
waypoint.h (and openttd.h) to waypoint_type.h.
|
|
openttd.h) to order_type.h.
|
|
openttd.h) to group_type.h.
|
|
|
|
files.
|
|
|
|
someone without a keyboard can enter text too. Patch by Dominik.
|
|
little neighbourhood. Based on a patch by Dominik.
|
|
TortoiseSVN when ran from a 32bit MSVC on win64 system (Jafinto)
|
|
steep slopes.
|
|
section of source.list
|
|
michi_cc).
-Fixed by removing FollowTrack_t structure and all related functions. Their purpose was to allow to call track followers from C code. They were:
1. never used
2. obsoleted by switching to C++
|
|
|
|
users unable to run generate bash script
|
|
with some parts for AI use.
|
|
|
|
|
|
file to spare 4-8kB of binary size
|
|
|
|
|
|
file while sorting. Only the "Absolute" name matters
|
|
It is more descriptive name and it will not upset the generate script anymore
|
|
|
|
|
|
in such a manner that it triggers an internal compiler error without any clue what of the code is wrong. Even trying to bisect the problem does not give a single line of code that causes the trouble.
|
|
GetTunnelBridgeLength() and/or GetOtherTunnelEnd() instead
|
|
|
|
long track
-Codechange: use dedicated pathfinder for signal updating, resulting in better performance and possible future improvements
|
|
needed.
|
|
aren't there anymore
|
|
|
|
|
