GC.start
? (lib/gosu.rb
already tries to protect against this, though =/)gdb
to it and see more information about the crash? The C level stack trace does not make any sense at all, why would IO.hpp
be involved in GL rendering?! :)gdb ruby
run path_to_script.rb
backtrace
when things blow up.
/home/ravenskrag/.rvm/gems/ruby-1.9.3-p448/gems/gosu-0.7.50/lib/gosu/swig_patches.rb:37:in
show': method call' called on terminated object (0x0000000190cb98 flags=0x0 klass=0x0) (NotImplementedError)
from /home/ravenskrag/.rvm/gems/ruby-1.9.3-p448/gems/gosu-0.7.50/lib/gosu/swig_patches.rb:37:in
show'<main>'
call
should be called on, are indeed the GL blocks.$gosu_gl_blocks = nil
gosu-0.7.50/lib/swig_patches.rb
, can you change the protected_draw_2
method to... def protected_draw_2
protected_draw
$gosu_gl_blocks_2 = $gosu_gl_blocks
$gosu_gl_blocks = nil
end
gl
blocks inside a global Array
, which is emptied at the end of draw
. But I think that's too early, because the blocks will only be disposed after draw
. With the new hack, they're retained for one more frame.Powered by mwForum 2.29.7 © 1999-2015 Markus Wichitill