An attempt to run main rope code only every 8th tick:
- A big win for inner loop which is O(n^2) and thus it takes 64 times less cpu now.
- Look and feel is almost identical (well, to my non-pro view)
- Yes, the speed is limited to 7 px per accounting tick which is potentially very bad to collisions
- Ruin shoppa
#
# Export file for libvorbisidec
#
# Only the symbols listed in the global section will be callable from
# applications linking to libvorbisidec.
#
@PACKAGE@.so.1
{
global:
ov_clear;
ov_open;
ov_open_callbacks;
ov_test;
ov_test_callbacks;
ov_test_open;
ov_bitrate;
ov_bitrate_instant;
ov_streams;
ov_seekable;
ov_serialnumber;
ov_raw_total;
ov_pcm_total;
ov_time_total;
ov_raw_seek;
ov_pcm_seek;
ov_pcm_seek_page;
ov_time_seek;
ov_time_seek_page;
ov_raw_tell;
ov_pcm_tell;
ov_time_tell;
ov_info;
ov_comment;
ov_read;
vorbis_info_init;
vorbis_info_clear;
vorbis_info_blocksize;
vorbis_comment_init;
vorbis_comment_add;
vorbis_comment_add_tag;
vorbis_comment_query;
vorbis_comment_query_count;
vorbis_comment_clear;
local:
*;
};