| <<<Back 1 day (to 2016/11/07) | 20161108 |
Robin_Watts | ray_laptop: If you read the logs, sheel is trying to "flatten" PDF form data, not transparency. | 00:31.30 |
k-man | hi | 01:06.23 |
ghostbot | Welcome to #ghostscript, the channel for Ghostscript and MuPDF. If you have a question, please ask it, don't ask to ask it. Do be prepared to wait for a reply as devs will check the logs and reply when they come on line. | 01:06.23 |
k-man | i have a program that generates pdfs and emails them to our customers, recently i've been getting reports of the PDFs fail to open correctly. mupdf reports "errors found on page" when i open the pdf but doesn't elaborate more | 01:07.32 |
| command line output http://sprunge.us/NdfY | 01:08.11 |
kens | k-man : I would suggest you open a bug report and attach an offending file, there's not a lot we can tell fro the output. Clearly MuPDF thinks there's a problem with your PDF file, but we'd need to see it to determine if that is true. | 07:58.09 |
k-man | kens, thanks, i don't think its an issue with MuPDF really, more an issue with the PDF | 09:54.42 |
kens | I suspect that's true, but its impossible to tell without seeing the file | 09:55.05 |
| If you want to put it somewhere I can look at it | 09:55.18 |
| Or just open a bug report and someone will take a peek at the file and tell you waht's wrong with it | 09:55.35 |
k-man | i just don't think it appropriate to file a bug on MuPDF as i don't think its an MuPDF issue is what I mean | 09:55.58 |
kens | At a guess I'd say there's an Indexed colour space where the lookup table is missing or incorrect, but that's just a guess | 09:56.06 |
| k-man : we get that a lot :-) | 09:56.16 |
k-man | ok | 09:56.30 |
kens | BUt if you stick the file somewhere I'll take a look, or mail it to me | 09:56.31 |
k-man | thanks kens, | 09:56.39 |
kens | If you don't want it public | 09:56.39 |
k-man | yes, that is also an issue | 09:57.05 |
kens | OK mail it to ken.sharp at artifex.com | 09:57.21 |
| I promise to delete it afterwards :) | 09:57.43 |
k-man | so i also narrows the issue down, i post process the pdf using the perl PDF::API2 module, the original file doesn't report issues in MuPDF, but my post processed one does | 09:58.11 |
| anyway, give me a few minutes, I'll email you something | 09:58.24 |
kens | g | 09:58.32 |
| NP | 09:58.48 |
k-man | sending now | 10:04.49 |
kens | got it | 10:05.10 |
| give me a few minutes | 10:05.14 |
| Well Acrobat is happy with both files | 10:05.35 |
| MuPDF likes them both for me too | 10:06.23 |
k-man | err what? | 10:06.29 |
| hmmm | 10:06.32 |
kens | WHich version of MuPDF is exhibiting a problem ? | 10:06.35 |
| and on which platofrm ? | 10:06.50 |
k-man | kens, i'm at home now and i grabbed the files from work, so i feel a little unable to properly re-test and answer your questions | 10:07.08 |
| i'll do my best though | 10:07.13 |
kens | No worries | 10:07.18 |
k-man | mupdf on debian: Installed: 1.5-1+deb8u1 | 10:07.59 |
kens | I'm using an interim version here, a comparatively old version of 1.9a | 10:08.04 |
Robin_Watts | 1.5 ? | 10:08.17 |
kens | Hmm. You maybe want to use a more recent version | 10:08.22 |
k-man | apparently | 10:08.24 |
| this is what comes with debian jessie i believe | 10:08.31 |
kens | I don't have a colpy of 1.,5 to hand, I'll try 1.4 and 1.6 | 10:08.47 |
Robin_Watts | k-man: Complain to debian then - that's ages. | 10:08.50 |
k-man | ok, this is heartening news though. perhaps the correct answer when people say they can't open the PDF is "get a newer pdf viewer" | 10:09.05 |
Robin_Watts | We release every 6 months, and we're about to do 1.10 now. | 10:09.06 |
kens | 1.4 doesn't like new4.pdf | 10:09.09 |
| Does open it but complains | 10:09.18 |
k-man | ok | 10:09.22 |
Robin_Watts | so that's 3 years of bugfixes you're missing. | 10:09.23 |
k-man | Robin_Watts, ok :) | 10:09.31 |
| i know, its debian jessie | 10:09.36 |
| they like to do things that way | 10:09.42 |
| is it hard to compile from source? | 10:09.52 |
kens | 1.6 didn't like it, but 1.7 is fione, so I'd have to say its been bug-fixed | 10:10.18 |
Robin_Watts | k-man: Not at all. | 10:10.27 |
kens | It should nto be hard to compile from source, I've done it lots | 10:10.32 |
k-man | ok, i'll test here on my home debian machine | 10:10.38 |
| thanks | 10:10.42 |
Robin_Watts | You can download the 1.9 source on mupdf.com. Then "make build=release" :) | 10:10.57 |
kens | deletes the files | 10:10.57 |
k-man | thanks kens, much appreciated | 10:11.11 |
| i mean, I appreciate the testing, and deletion! haha | 10:11.23 |
kens | NP | 10:12.16 |
| :D | 10:12.18 |
k-man | btw, i really love MuPDF | 10:13.57 |
| its so awesomely minimal | 10:14.05 |
| i'm downloading the git version | 10:19.28 |
| fwiw, debian stretch is on 1.9a currently | 10:20.22 |
kens | current code is always nice, luckily its near release, so it should be all good | 10:20.31 |
k-man | i get these errors when compiling: http://sprunge.us/MZdh | 10:28.46 |
| also the README didn't mention i might need the libharfbuzz-dev library | 10:29.48 |
| just in case you want to update the READEM | 10:29.57 |
| README | 10:29.59 |
kens | I'll have to defer to RIobin and tor on build issues | 10:30.28 |
| Or even RObin | 10:30.39 |
k-man | oh, it looks like this: http://bugs.ghostscript.com/show_bug.cgi?id=693518 | 10:31.30 |
kens | You're using system shared libraries ? Because it ought to compile with the third party libraries we ship. Did you grab those from Git as well ? | 10:32.17 |
k-man | oh.... no | 10:33.30 |
kens | You'll want to do that too | 10:33.45 |
k-man | ok, missed that bit in the readme, sorry | 10:33.58 |
kens | somehting like git submodule update, but I can never remember the syntax | 10:34.00 |
k-man | yep, its in the readme, i just missed it | 10:34.28 |
kens | Well at least its there, so you don't have to depend on my flaky memory :) | 10:35.02 |
k-man | hehe | 10:35.37 |
| so is the development sponsored by a company? | 10:36.30 |
kens | MuPDF is dual licenced, like Ghostscript, there's an open source (AGPL) version and a commercial licence | 10:37.27 |
| Artifex licences both products commercially | 10:37.39 |
| Obviously the commercial licneces are what pays for us developers ;-) | 10:39.59 |
k-man | yep, that's great | 10:40.09 |
sheel | ray_laptop & Robit_watts: Thanks a lot for your inputs. | 14:25.45 |
| We do want to retain the view but want to make sure there is no malware hidden in an image or embedded object. | 14:27.08 |
| ray_laptop: It seems what you suggested will work for us as long as we do not loose any view. I have to confirm whether we need the text to be searchable. | 14:32.13 |
Guest22490 | Hello I am needing to install GSView 5.0 but I am unable to, where can I find this install? | 14:51.28 |
Robin_Watts | Guest22490: GSView 5.0 was produced by GhostGum. | 14:54.49 |
chrisl | Guest22490: a quick google search will bring up a download site | 14:55.15 |
Robin_Watts | We (Artifex, the company that maintains and develops Ghostscript) now offer GSView 6.0, which is better in our opinion. | 14:55.22 |
sheel | Why isn't lib included in ghostscript-9.20-linux-x86_64.tgz | 15:52.17 |
| It seems I need to get the source code for the lib/viewpbm.ps | 15:53.03 |
Robin_Watts | That's a question for chrisl, and he's in a meeting now. | 15:53.27 |
chrisl | sheel: it's never come up before, tbh | 15:54.23 |
| sheel: The purpose of the Linux binaries was to allow people to test the latest binary easily, they're not meant as a "proper" distribution of Ghostscript | 15:56.04 |
sheel | chris1: ok; thanks. What is the standard way to get the library? directly copy from source? | 15:58.30 |
chrisl | sheel: On linux, either by source, or by the distro Ghostscript package. Unfortunately, shipping "proper" binaries for Linux is a bit a nightmare, because of all the variation in the distributions | 15:59.54 |
| sheel: Most distros these days, I think, put the Ghostscript related stuff in /usr/share/ghostscript/* | 16:01.46 |
sheel | chris1: ok; thanks | 16:14.39 |
| ray_laptop: you command with -r300 reduces the size by many scales. Is there a way to preserver the original size? | 16:17.41 |
| ray_laptop, Robin_Watts: The command 'gs -q -sstdout=/dev/null -sDEVICE=ppmraw -o /tmp/xxx input.pdf ; gs -q -r300 -sDEVICE=pdfwrite -o output.pdf -r300 -dSCALE=1 -- lib/viewpbm.ps /tmp/xxx' produces ver tiny image - fits 1/8th of a standard 8/11 page | 19:45.25 |
| Any help on how to make the image fit the standard page or original size will be appreciated. | 19:46.27 |
Robin_Watts | So you write input.pdf to a ppm at the standard 72 dpi. | 19:46.47 |
| Then you try and put that ppm back into a PDF at 300dpi, and wonder why it's small? :) | 19:47.07 |
| Put a -r300 before the input.pdf :) | 19:47.24 |
| Why do you mention -r300 twice in the latter command? | 19:48.12 |
sheel | that was the command given by ray_laptop. | 19:49.24 |
Robin_Watts | yeah, ray obviously put the -r300 in the wrong place :) | 19:50.56 |
sheel | Robin_Watts: That worked! Thanks a lot! | 19:50.57 |
Robin_Watts | no worries. | 19:51.04 |
sheel | I really appreciate the help provided by the ghostscript team You guys rock! | 19:52.17 |
Robin_Watts | sheel: No worries. Did Kens talk to you about your usage of this stuff? | 19:55.21 |
| are you using it for a personal project, or a work one? | 19:55.34 |
sheel | Robin_Watts: The ghostript conversion cause the size balloon from 266KB to 2.8MB, is that expected? | 19:55.43 |
Robin_Watts | And if it's a work one, is it for use 'in house' or for 'distribution'? | 19:55.50 |
| sheel: Of course that's expected. | 19:56.07 |
| You've converted from a nice compact "draw a red circle here", "put the string 'fred' here" type representation to a single bitmap image of each page. | 19:57.03 |
sheel | Robin_Watts: OK, makes sense. It is for work. it is for in house. | 19:58.06 |
Robin_Watts | You can either do that at a low resolution (in which case the file will look crap when you zoom in, but will stay small), or you can do it at a high resolution (in which case the file will look fine as long as you don't zoom too far, but it'll be larger) | 19:58.07 |
| Ok. | 19:58.27 |
| Forward 1 day (to 2016/11/09)>>> | |