| | <<<Back 1 day (to 2018/11/11) | 20181112 |
| tor8 | Lovely. Adobe Reader XI requires Flash Player to display PDF portfolio documents.... | 15:07.29 |
| moolc | tor8: didn't they (Adobe) deprecate it (Flash Player) a while back? | 15:30.56 |
| Robin_Watts | tor8: http://git.ghostscript.com/?p=user/robin/mupdf.git;a=commitdiff;h=1b9e3d6b729fe5ba2bd56b75d4ad6213c8b7c1d9 | 18:51.26 |
| | The bmpcmp is convincing. | 18:51.33 |
| | and this for Raed's other bug: http://git.ghostscript.com/?p=user/robin/mupdf.git;a=commitdiff;h=7ef7fbc7e21334091858802b3aad9b33dfd7cb2f | 18:52.18 |
| tor8 | Robin_Watts: Bug 700154 fix LGTM | 19:17.54 |
| | Robin_Watts: and the SVG one I see no obvious problems with, but I don't claim to understand the feBlend voodoo | 19:20.54 |
| Robin_Watts | tor8: Me either :) | 19:21.12 |
| | The in/in2 stuff says what you want to blend. | 19:21.31 |
| | the background that's already been rendered is one. | 19:21.43 |
| | what you are rendering now is another., | 19:21.49 |
| | (you can choose to just filter strokes or fills etc) | 19:22.06 |
| tor8 | yeah. I'm sort of seeing how it fits together from the spec. | 19:22.33 |
| | ugly hack is what it is! | 19:22.44 |
| | the spec, not your code :) | 19:22.49 |
| | Robin_Watts: both commits LGTM | 19:24.48 |
| Robin_Watts | Ta. | 19:25.27 |
| tor8 | two commits on tor/master need nodding at | 19:25.41 |
| Robin_Watts | The mutool clean one... What was the problem there? | 19:28.10 |
| | If I have an ascii encoded image, and I clean it, will I get a file with the raw image data in it? Or will it still be ascii encoded? | 19:28.45 |
| | The second commit is good. | 19:30.07 |
| tor8 | the problem is we don't support returning decoded JPXDecode data from our filter chains | 19:30.30 |
| Robin_Watts | I understand that. | 19:30.38 |
| tor8 | we always short out the JPXDecode step, since the PDF interpreter handles it in a special case | 19:30.54 |
| | but the pdf-write code would think it had been decoded, and removed the /Filter JPXDecode | 19:31.16 |
| Robin_Watts | Oh, when we decompressed images, we are actually giving them the pixel values. | 19:31.25 |
| | OK, lgtm. | 19:31.31 |
| tor8 | yes. | 19:31.33 |
| Robin_Watts | I'd really like it to be able to strip other encodings off and just give us the binary. | 19:31.51 |
| tor8 | and writing the pixel values would break things, since stuff like colorspaces are in the JPX stream | 19:32.06 |
| Robin_Watts | but either it does that already, or it doesn't - your code shouldn't make it worse. | 19:32.09 |
| | yeah, I follow why we can't (trivially) do it. | 19:32.30 |
| tor8 | yeah. I've been meaning to add a 'strip ASCII' filter pass/option to it. | 19:32.30 |
| | we have one to add an ASCII filter to all binary streams | 19:32.39 |
| | but not an option to remove one if it's there | 19:32.46 |
| | Forward 1 day (to 2018/11/13)>>> | |
Log of #mupdf at irc.freenode.net.