rather then forcing applications to set this to a single bitmap. this
results in much nicer looking secret doors that don't randomly flip
bitmaps at all during and after opening.
fixes secret door bitmaps seeming to switch around randomly before
coming to a stop. they still will switch once after opening, but it will
now only switch to the expected bitmap.
otherwise it wasn't actually doing anything with it. palette loading
was done as part of bitmap loading only, but this allows a specific
palette to be set which may make it easier to notice bitmap palette
mismatches in some scenarios
- medit.dtf is assumed to always be beside mapedit.exe
- assets listed in the desc file being loaded are all assumed to be
beside that same desc file
paths for both the mapedit.exe and desc file are discovered
automatically from the command line arguments received.
this, in my opinion, represents the most intuitive/logical behaviour.
before this change, you would have needed to have copied mapedit.exe
and medit.dtf to the same directory as your map desc file and all the
assets being loaded. kind of icky in my opinion.
with the dos map editor only, it's basically required to have all these
to be able to edit maps and rebuild the 'pics.dtf' resource files to
actually use your edited maps with the way this version of ack3d works
the map_edit source on the book's cd included a prebuilt version of
the old watcom version of ack3d (with headers) which defined UINT to be
'unsigned short' while our version of ack3d defines UINT to be
'unsigned int'
the map_edit sources from the cd seemed to include two different
versions of a map editor. the sources i've left here seem to most
closely match the version of mapedit.exe included on the book's cd
these were on the book cd under /ACK/DOS/BORLAND. this version of
ackiff.c is SIGNIFICANTLY faster then the windows one ported back to dos
(i don't think i did anything wrong when i did that, but who knows...)
despite the book itself stating that this code was not included due
to the old GIF patent back in the 90's, it was actually included on the
cd under /ACK/DOS/WATCOM/ACK_LIB. oops!
with the timer speedup call left in, it feels much closer (speed-wise)
to the original watcom fdemo executable from the book cd. but still not
quite a perfect match in my opinion
i didn't recognize _getds from anywhere so assumed it was something
commented out for a reason, but turns out that it was part of the
original modplayer code (which was originally written for watcom
anyway). _getds is an internal function from the watcom standard
libraries.
i'm convinced that the modplayer doesn't work as-is. start/stop of bg
music was commented out in the newer borland sources, and neither the
borland 'nor watcom fdemo executable on the book's cd ever played
bg music for me. enabling it (with the mod file present) causes some
crashes.
borland conditionals have been removed. asm sources converted fully
to tasm ideal mode. keyboard and timer interrupt handlers moved to c
code copied from fdemo watcom sources
sources taken from book CD:
ack_lib -> /ACK/WIN/ACK_LIB
fdemo -> /ACK/DOS/FDEMO/SOURCE (and /ACK/DOS/BORLAND as needed)
mall -> /ACK/DOS/MALL/SOURCE (and /ACK/DOS/BORLAND as needed)
some source files were missing for the demo projects and needed to be
copied from /ACK/DOS/BORLAND (as indicated above)
Gered