Age | Commit message (Collapse) | Author |
|
Slightly simplify the code which determines whether to build a rail or a road bridge
|
|
Do not try to invalidate a window which just got deleted. It is pointless
|
|
Do not call ClearSlot() just before calling DeleteVehicle(), which does it too
|
|
Do not explicitly pass the station specification and the station to ResolveStation(). They are already contained in the ResolverObject
|
|
-Regression (r8314): Reevaluate the disabled elrail setting after loading because the railtype doesn't get saved anymore
|
|
STR_AVAILABLE_ROAD_VEHICLES for clarificaton (Tron)
|
|
global vehicle lists to 'Available <vehtype>' as it is a view-only list, not one from which you can purchase (rolling) stock.
|
|
brazilian_portuguese - 1 changed by fukumori (1)
korean - 1 changed by Nios (1)
ukrainian - 1 fixed, 23 changed by mad (24)
|
|
-Simplifies assignment from raw pointers
-Should be harder to crash the program by incorrect assignment into it.
-Should help with MorphOS compilation errors
|
|
(seems that g++ 2.95 defines it anyway)
|
|
assert_compile(!(x == y)) does.. don't ask me why!
|
|
because MorphOS wants UBYTE arrays and we use char arrays. Solution is a bit hackish.
|
|
-Fix: Fix an out of bounds array access when assigning the engine type in the _engine array
This usually hits the _engine_name_strings array and causes an invalid StringID, though it depends on the compiler which data structure gets placed after the _engines array.
Most probably this was exposed by removing railtype from Engine, which changed the size of this struct.
|
|
queue wrapped back to the oldest item, showing all news again.
-Regression [FS#573] (r8049): ShowLastNewsMessage could show an out-of-bounds news item because it did not checked if a previous item actually existed the first time it is called (forced news is INVALID_NEWS).
|
|
korean - 2 changed by leejaeuk5 (2)
slovak - 1 changed by lengyel (1)
slovenian - 290 changed by Necrolyte (290)
ukrainian - 1 fixed, 55 changed by mad (7), znikoz (49)
|
|
|
|
and changed a one-member enum into a static const. Simplify their naming and add some doxygen-comments to RoadStop
|
|
rename it to 'IsValid'
-Codechange: While I'm at it, put 'RoadStop::AllocateRaw' into the protected section
|
|
Simplify r8399
|
|
CmdBuildRoadStop(), but it was not deleted if DC_EXEC flag was not set or if the station creation failed
|
|
macros.h as it is no longer needed.
|
|
same location as the unload button
instead assert if the widget array place them at different locations
|
|
estonian - 1 fixed by vermon (1)
finnish - 1 fixed by kerba (1)
greek - 13 fixed by Kesnar (13)
norwegian_bokmal - 1 fixed by oletk (1)
romanian - 1 fixed by kneekoo (1)
spanish - 1 fixed by eusebio (1)
ukrainian - 31 fixed, 24 changed by znikoz (4), mad (51)
|
|
so the order of the widget array fits the placement on the screen
|
|
|
|
-Codechange: CheckCompatibleRail() is only called for tiles which are known to contain a piece of rail. Simplify the function accordingly by eliminating unnecessary checks.
-Fix (?): Also fix an inconsistency in deciding what a compatible rail type is between level crossings and other rail tiles. It is unknown if this caused any problems.
|
|
Use HASBIT() instead of implementing it manually
|
|
stations work
|
|
Now it only works on first and last widget to resize
It now works with offsets correctly (no longer assumes the left is 0)
It's no longer needed to have a widget right of the ones you resize
Can handle any number of widgets
|
|
-Regression (r8314): I only did half the necessary changes to move railtype from Engine to RailVehicleInfo. Now do the rest.
|
|
for each window, a global ResizeButtons() is added
|
|
(requested by Wolf01)
|
|
This fixes the issue where OSX included stdint.h though other includes before defining __STDC_LIMIT_MACROS (and failed to compile because of this)
|
|
refitable detection. It had two variables doing the same thing, but not always set to the same (oops)
|
|
this wrong argidx issue, we decided to remove the {P ..} from STR_TONS
|
|
DrawVehiclePurchaseInfo() and renamed them into type specific names
|
|
the details with one line since some newgrf trains needs it
|
|
only add "(refitable)" if the engine in question is refitable
|
|
|
|
brazilian_portuguese - 2 changed by fukumori (2)
esperanto - 14 changed by LaPingvino (14)
simplified_chinese - 6 fixed, 10 changed by Fishingsnow (16)
slovenian - 394 changed by Necrolyte (394)
ukrainian - 68 changed by mad (61), znikoz (7)
|
|
describes the
required value better anyway. However, INT64_MAX comes from stdint.h which
doesn't seem to exist on MSVC or MorphOS, so it's defined manually for them.
|
|
|
|
openttd.h, which does not exist in the masterserver/updater.
|
|
file of it's own
Note: it's the same as r8363, only this time the svn history of autoreplace should be preserved
|
|
|
|
descriptive names, and add some more comments.
|
|
datasets should be drawn, and use HASBIT for testing against them.
|
|
is invalid for both an int64 and a uint.
|
|
fix two longstanding bugs.
1) If the graph contained negative values, the y axis labels were half the size they should have been.
2) If the maximum value was larger than the height of the graph in pixels << 32, all datapoints were scaled to 0.
|
|
informations, add attribute m6.
More to come on that matter...
|