It was just adding a level of indirection to Ptr.getClass().
All the call were replaced by that instead. The number of lines changed
is important, but the change itself is trivial, so everything should be
fine. :)
This is usually not needed, because it is not possible to exit dialogue while in a choice. However you can still exit dialogue by loading a different savegame.
Kept some fixes from the first round of review. Found out that several
targets weren't being built with the same basic warnings disabled.
Disabled a few warnings for external libraries specifically, rather than
applying them to all targets.
Handles edge case where it was possible to highlight adjacent
sub-terms without whitespace between them.
Also makes ignoring words not prefixed by whitespace the
assumed behaviour.
Dialog links can no longer be highlighted if they appear in the
middle of the word. This is achieved by confirming that the
character before a match is not alphabetic, so that words
following hyphens can still potentially match.
Most warnings are innocuous (wrong type-specifier for forward
declarations, conversion of literals into unsigned integers, warnings
about methods optimized out), but I believe actual bugs were revealed in
vartypedelegate.cpp and combat.cpp.
The first problem was with the ScrollView skin, which had a full-sized client area. Since the scrollbar starts out visible, MyGUI expects the client area to be smaller to accomodate for the scrollbar width. As a result, the starting canvas size becomes bigger than the view size.
Another bug was with the MWList code: reducing the canvas size for the scrollbar is not needed, since MyGUI is already doing that, and attempting to do it manually interferes with the view offset.
I've always felt that having the version/revision text at the bottom
center in the main menu was a bit out of place. A more common place for
this kind of thing is in one of the corners. I chose bottom right.
Aditionally I right aligned it and changed the v and r in version and
revision to capital letters.
Comments?
The progress is not particularly accurate. It simply uses the current / total number of records written/read as indication. Cell records are currently the largest by far, but there is a good chance that could be optimized using a change tracking system.