1a6a8b52f7
Original revision: 5123b1bf68777ffa86e651f178046b26a87cf2d9 MIT Licensed. Some games still crash and others are unplayable due to issues with controls. Still need a "real" polygon filling algorithm. Currently builds one plugin per puzzle (about 40 in total, around 100K each on ARM), but can easily be made to build a single monolithic overlay (800K or so on ARM). The following games are at least partially broken for various reasons, and have been disabled on this commit: Cube: failed assertion with "Icosahedron" setting Keen: input issues Mines: weird stuff happens on target Palisade: input issues Solo: input issues, occasional crash on target Towers: input issues Undead: input issues Unequal: input and drawing issues (concave polys) Untangle: input issues Features left to do: - In-game help system - Figure out the weird bugs Change-Id: I7c69b6860ab115f973c8d76799502e9bb3d52368
25 lines
650 B
Bash
Executable file
25 lines
650 B
Bash
Executable file
#!/bin/sh
|
|
|
|
# Generate a screenshot from a puzzle save file. Takes the
|
|
# following arguments, in order:
|
|
#
|
|
# - the name of the puzzle binary
|
|
# - the name of the save file
|
|
# - the name of the output image file
|
|
# - (optionally) the proportion of the next move to redo before
|
|
# taking the screenshot.
|
|
#
|
|
# This script requires access to an X server in order to run, but
|
|
# seems to work fine under xvfb-run if you haven't got a real one
|
|
# available (or if you don't want to use it for some reason).
|
|
|
|
binary="$1"
|
|
save="$2"
|
|
image="$3"
|
|
if test "x$4" != "x"; then
|
|
redo="--redo $4"
|
|
else
|
|
redo=
|
|
fi
|
|
|
|
"$binary" $redo --screenshot "$image" --load "$save"
|