| | <<<Back 1 day (to 2011/07/29) | 2011/07/30 |
| mvrhel2 | tkamppeter: you are probably asleep, but I am cluster pushing a fix for RGBW that has it set its profile properly in gdevcups.c. Output looks good with rasterview and requires no special command line | 06:09.10 |
| | it should come back fine, then I will commit and see about having chrisl get all of this into 9,04 | 06:09.36 |
| | chrisl: are you up? | 06:09.42 |
| mvrhel2 | realizes it is the weekend, not to mention time for sleep | 06:12.18 |
| Robin_Watts | Morning tor8. | 09:53.47 |
| | I left a commit for you on casper. | 09:53.53 |
| | http://www.h-online.com/open/news/item/Emacs-has-been-violating-the-GPL-since-2009-1288623.html | 10:00.19 |
| kens | Well, I think I'm going to have to store a lto more information about glyph widths and positions for my text extraction stuff. The reason the 'ratpoison' man page came out incorrectly is because (bizarrely) the PostScript does emit the last character of one word, and the first of the new word as one operation, using ashow to insert the 'space'. | 10:19.36 |
| tor8 | kens: I've tried a different approach to extracting the text, but it's dog slow | 10:25.42 |
| kens | A different approach to what you currently do ? | 10:25.56 |
| tor8 | based on a variant of the 'flood fill' algorithm, to group the text into contiguous blocks | 10:26.18 |
| kens | I was trying to get away without storing the x/y co-ord of each glyph, but I now realise that's not possible. | 10:26.29 |
| | Ah, I'm not trying to extract blocks yet. | 10:26.42 |
| tor8 | I store the bounding box and baseline coordinates for each glyph | 10:26.49 |
| kens | Hmm, more than I was intending.... | 10:27.07 |
| tor8 | the baseline to assemble them back into lines after I've grouped them into blocks | 10:27.10 |
| kens | Yes. | 10:27.16 |
| | Surely you also need text direction ? | 10:27.22 |
| tor8 | yes, I will need that eventually too | 10:27.38 |
| kens | Which depends on CTM and WMode :-( | 10:27.45 |
| tor8 | yeah. i've only done horizontal text for now. you only need to store a 2d vector for the advance direction, not a full CTM and wmode | 10:29.03 |
| kens | Certainly, I merely meant that it was influenced by both | 10:29.26 |
| | Right, I'm off for the weekend, bye all. | 11:16.17 |
| ccarpenterg | hello | 22:46.54 |
| | anyone has experience using elaphe http://code.google.com/p/elaphe/? | 22:47.11 |
| | Forward 1 day (to 2011/07/31)>>> | |
IRC Logs