| | <<<Back 1 day (to 2015/09/16) | 20150917 |
| valkanium | someone can help me | 01:59.10 |
| | i'm receiving error message: ghostscript failed to distill PS | 01:59.31 |
| kens | valkanium (for the logs) that is not a Ghostscriipt error message. | 11:52.43 |
| mvrhel_laptop | That actually is a gsview error message... | 12:05.37 |
| kens | Intriguing | 12:05.55 |
| | Needs a bug report though. | 12:06.32 |
| jogux | RobinWattsLenovo: any chance you can ask someone to knock the video off? almost completely loss audio :( | 15:17.07 |
| kens | marcos is fiddling with it now | 15:17.33 |
| jogux | ta | 15:17.40 |
| | if anyone just started a large upload/download I might want to hit them ;-) | 15:17.49 |
| jogux | has completely dropped out now | 15:18.02 |
| kens | I think marcos may have killed stuff, not sure | 15:18.18 |
| | he's sti doing stuff | 15:18.27 |
| | welcome back | 15:19.05 |
| rayjj | jogux: how hard would it be to have the ATS work with the NUI ? | 16:35.25 |
| jogux | hard | 16:35.36 |
| | we need to do an API test suite for the core | 16:35.47 |
| | then we need to do automation tests for each OS I reckon | 16:35.55 |
| duch | good evening | 16:53.04 |
| | Are there any postscript (ghostscript) gurus overhere? | 16:53.43 |
| kens | Yes but we are all in a staff meeting. As the message says, if you have a question just ask it | 16:54.15 |
| duch | I shall give a short explenation about what I'm trying to do | 16:54.52 |
| chrisl | And we're not generally in the game of teaching Postscript programming...... | 16:54.52 |
| kens | If this is a PostScript programming question you might want to consider using Stack Overflow | 16:55.23 |
| duch | Normaly I use ghostscript to concatenate different kind of pdf together and generate a new one | 16:55.25 |
| | But sometimes this results in very large PDF files | 16:55.37 |
| kens | No you don't. You are *not* concatenating PDFs | 16:55.42 |
| duch | I know it builds a complete new pdf | 16:56.10 |
| kens | Not only that, it does not contain any part of the input | 16:56.28 |
| | The ony goal is that the visual appearance of the output should be essentially the same as the input | 16:56.52 |
| duch | I was wondering if it is possible to control ghostscript so that I for example can say to make a file that is about 10mb big and then start generating a new one | 16:57.00 |
| kens | No, that is not at all possible | 16:57.18 |
| duch | I know that the option exists to split everything in different pages | 16:57.28 |
| | Ah ok to bad | 16:57.49 |
| | So there exists no possibility to ask ghostscript through postscript how big the output file is? | 16:59.19 |
| kens | No | 16:59.27 |
| | We don't kow how big it is until we write it, we don't write it until the device is closed | 16:59.48 |
| duch | Everything is done in memory? Because ghostscript produces a lot of temporary files | 17:00.35 |
| kens | Some stuff is in memo, some is files on diks, these are not neccesarily compressed, and will be in the final output, etc etc | 17:01.06 |
| duch | is it an option for me to program something like this into the ghoscript engine with c++ or is it a good as impossible task to do? | 17:02.22 |
| kens | Yu can''t use C++, GS is written in C | 17:02.40 |
| | Ghostscript itself is difficult to understand, the pdfwrite device even more so | 17:02.59 |
| duch | uhm ok | 17:03.21 |
| kens | If you have a year or so to spare you might be able to do something like that (but not in C++) | 17:03.23 |
| | And some of it is simply impossible (fonts are not subset until the PDF is written, so how big are they ?) | 17:03.56 |
| duch | C should also not be a problem for me ... but a year is :-) | 17:03.59 |
| rayjj | duch: you could send files in one by one and get a rough idea by sizing the temp files, but as kens says, they won't directly correspond to what gets written to the output PDF. | 17:04.20 |
| kens | jj that bears almost no relation to the size of the final file | 17:04.48 |
| chrisl | Given the way that PDF (and pdfwrite) work, it's hard to see how you could do this efficiently - you'd effectively have to generate a complete output PDF after every page, check the size, and discard it if it's smaller than the target size. | 17:04.57 |
| kens | for instance we often write images several times using different compressions | 17:05.05 |
| duch | What I now do it use the %d option (if I remember correctly) and then put one file together and then look to the output file... but that is very time consuming | 17:05.13 |
| | Any idea if there are some other tools to do something like this? itextsharp... mupdf ... etc... | 17:06.24 |
| | poppler | 17:06.41 |
| kens | Google is ytour friend | 17:06.53 |
| duch | ok... well thanks for the answers and have a nice evening | 17:07.27 |
| rayjj | duch: and maybe try pdftk | 17:07.35 |
| duch | and keep up the good work with ghostscript | 17:07.37 |
| | pdftk uses itextsharp internaly | 17:07.50 |
| | And I used to use pdftk years ago but then switched to ghostscript because of problems I had with pdftk | 17:08.30 |
| | ghostscript just generates the best output pdf files so far | 17:09.06 |
| | The nice thing about ghostscript that it also fixes errors in pdf files and just generates the best result... we get pdf files from everywhere | 17:10.16 |
| | In my opinion the biggest problem with the pdf standard is that everyone can make it's on driver and not everyone is good in implementing it the correct way (according to standards) | 17:11.03 |
| | I see the most strange errors in pdf's | 17:11.23 |
| kens | You shoud see the ones we get | 17:13.48 |
| duch | have a nice evening | 17:13.48 |
| | bye | 17:13.53 |
| | Forward 1 day (to 2015/09/18)>>> | |
IRC Logs