| | <<<Back 1 day (to 2011/08/20) | 2011/08/21 |
| robin_watts_mac | tor8: You use lcms in mupdf, right ? | 11:34.02 |
| tor8 | robin_watts_mac: nope | 11:40.10 |
| | I've had plans to, but never gotten around to it | 11:40.22 |
| robin_watts_mac | ok. | 11:40.27 |
| tor8 | you thinking about memory errors again? | 11:40.41 |
| robin_watts_mac | No. | 11:40.47 |
| | lcms uses a different map from 8 to 16 bit colors than the standard 0xAB -> 0xABAB thing. | 11:41.17 |
| tor8 | 0xAB00? | 11:41.37 |
| robin_watts_mac | and because of that gs gets different results for line art and images. | 11:42.02 |
| | no. More wacky than that. | 11:42.28 |
| | Sorry, no. | 11:43.29 |
| | 8 to 16 is 0xABAB. | 11:43.36 |
| | but 16 to 8 isn't just truncation | 11:43.48 |
| | #define RGB_16_TO_8(rgb) (BYTE) ((((rgb) * 65281 + 8388608) >> 24) & 0xFF) | 11:44.01 |
| | RGB_16_TO_8(RGB_8_TO_16(x)) == x as you'd expect | 11:45.00 |
| | but SIGMA(i=0 to 0xFFFF, abs(i - RGB_8_TO_16(RGB_16_TO_8(i)))) is minimised. | 11:46.37 |
| | brb | 11:48.11 |
| cryptopsy | is it possible to control mupdf somehow through keybindings in readline or tmux/screen ? | 23:25.47 |
| | Forward 1 day (to 2011/08/22)>>> | |
IRC Logs