Either of these would previously crash (same for VX):
tm = Tilemap.new
at = tm.autotiles
tm = nil
GC.start
at[0] = Bitmap.new(1, 1)
tm = Tilemap.new
at = tm.autotiles
tm.dispose
at[0] = Bitmap.new(1, 1)
Funnily, this makes RMXP itself crash too, but crashing is
never acceptable except for possibly resource exhaustion.
MRI: Filesystem: A garbage collected SDL_RWops file handle
will call close on itself even if 'close' was explicitly
called on it on the script side before.
The RGSS doesn't do it. But doing it shouldn't be a problem, as it's the
correct way. If a game is broken by this commit (unlikely), then the game
needs to be fixed as silent truncation is highly unlikely to be wanted.
This removes the static dependency on fluidsynth being present
at buildtime (even headers aren't needed anymore).
Even though midi is a default format for the RPG XP/VX series,
it has fallen more and more out of use, with VX Ace completely
abandoning it from the RTP and making ogg vorbis the de facto
standard. Midi support is kept for legacy reasons, but isn't
encouraged. On top of all this, fluidsynth together with glib
is a heavy dependency that often times won't even be used.
Making it optional at build time is an attempt to unify and
keep build config fragmentation low.
In RGSS3, fluidsynth / midi is not initialized at all by
default, but rather on demand when either a midi track is
played back or Audio.setup_midi is called.
The gist of it is that for Etc and Font props, the assignment
operator (eg. 'sprite.color=') does not take a reference of the
right hand parameter and replaces its previous one with it (this
was the old behavior). Rather, it keeps its internal property
object and copies the parameter object into it by value.
The getter is unchanged; it still returns a reference to the
internal property object.
s = Sprite.new
c = Color.new
s.color = c
p s.color == c # => true
p s.color.object_id == c.object_id # => false (true before)
c = s.color
p s.color.object_id == c.object_id # => true
Previously, when the script names in the backtrace were in the form
'SectionXXX'/'{XXXX}' (default RMXP behavior), this same name would
be shown in the error message box, whereas RMXP always displays the
actual script name (eg. 'Scene_Map') instead; only with
useScriptNames=true was mkxp's behavior mostly correct.
Fix this by keeping a backtrace format -> actual script name mapping
and always using the script name in the message box.
The script name and line would also be parsed wrongly if the name
contained any colons, fix this by walking the exception string
back to front instead.
Also prepend the script index to the name when useScriptNames=true;
this is invaluable when debugging projects where multiple scripts
carry the same name. The format is 'XXX:Script_Name'.
Even though 3.0.1 is the current version, 3.0.0 is used, because mkxp doesn't
implement the changes like not raising a TypeError for 'Color.new == nil' yet.
3.0.1 (finally) returns false. Other changes don't affect mkxp (I think).
See <http://tkool.jp/support/download/rpgvxace/rpgvxace_update> (Japanese).
Fixes Window_NameBox visual appearance in Skyborn.
Also nuke the second SceneElement constructor that has been
obsolete since the Tilemap mapViewport rewrite.
Jonas Kulla
cremno