| <<<Back 1 day (to 2017/09/19) | 20170920 |
chw | hi kens, thanks for taking a look at the report. | 07:04.48 |
| would it be helpful for me to supply some more files? | 07:05.11 |
kens | I don't think so, one is enough to be going on with. I suspect the problem is the U umlaut, and at the moment I am suspicous that the problem is with the validation tools rather than the data. | 07:05.59 |
| But I need to sit down and analyse it, and right now I'm trying to get free of a conversation which has spiralled completely out of control :-( | 07:06.28 |
chw | oh, ok. it's just that i have another file, without umlauts, that has a similar, if not exact same issue. the meta contains the trademark sign. | 07:07.40 |
kens | That's still a character outside the ASCII range | 07:07.55 |
| So we will endoce it with PDFDocEncoding in the dicitonayr and UTF-something in the XML | 07:08.11 |
| sorry I mean encode | 07:08.22 |
chw | got it. i guess the only thing i haven't tried yet is a pdf file within the ASCII range. | 07:09.23 |
| may give that a try, just to make sure it actually works. | 07:09.46 |
kens | I suspect that will validate. Because PDFDocEncoding includes the normal ASCII range, and UTF-8 is exactly the same as ASCII for ASCII characters | 07:10.00 |
| So the two strings will be identical | 07:10.12 |
| My suspicion is that the validation tools aren't capable of resolving the different encodings to realise that the actual content is the same | 07:10.38 |
| But, I need to actually work out what's in the XML before I can say that for sure | 07:11.04 |
chw | sounds plausible. i'll try a few more files, later today. may i ask what exact version of acrobat x pro you have? as in is it patched to the lastest, or just 1.0. that way i can just use that right away. | 07:14.03 |
kens | Umm, good question :-) | 07:14.15 |
| I'll launch it and ask it, I suspect its unpatched, I try to avoid patches from Adobe | 07:14.33 |
| Its says its 10.1.6 | 07:15.00 |
| grr | 07:15.05 |
| 10.1.16 | 07:15.12 |
| Seems my keyboard doesn't like me this morning | 07:15.22 |
chw | ok, so it's the latest version, good to know. | 07:15.48 |
kens | :-) | 07:16.04 |
chw | the thing is, you never know what they patch, so who knows if validation is one of those things... | 07:16.41 |
| better just use the same version. | 07:16.58 |
kens | As I recall, the validation was bought in from an outside supplier, and was not updated for a number of releases. It does sound like they've changed it in the most recent release. | 07:17.44 |
chw | as in Acrobat DC, right? | 07:18.15 |
| i'm not really up to date on adobe versions | 07:18.33 |
kens | Yeah, I've been avoiding the cloud one, I don;t want to pay Adobe a subscription | 07:18.42 |
chw | most people don't. also the autoupdater is straigt up garbage | 07:19.26 |
kens | Another reason to avoidit :-) | 07:19.37 |
| I'll probably have to update when/if Adobe ever support PDF 2.0 | 07:19.58 |
chw | right, working with this stuff, makes it a quasi requirement | 07:21.03 |
| sadly | 07:21.08 |
kens | :-) I was about to say that.... | 07:21.16 |
chw | also, looking at the release notes for the acrobat updates there is acrobat 2017. i've never even heard that was a thing.. | 07:22.25 |
kens | Hmm, maybe that's the one that supports PDF 2.0 then, its a new one on me | 07:23.00 |
| Released July 2017 apparently, so pretty new | 07:23.30 |
| Ah, its the desktop version of DC | 07:23.43 |
| Nothing of interest in the headline features though | 07:24.28 |
chw | yeah, seems kind of stale. i don't see a point in upgrading | 07:25.19 |
kens | wonders why a search tool to find your tools isn't regarded as a broken UI | 07:25.24 |
chw | ha, even wikipedia gave up on keeping up with the acrobat version history. | 07:27.49 |
kens | :-) | 07:28.06 |
chw | to be fair though, auto updates without a changelog make it quite hard, if not impossible... | 07:28.52 |
kens | Yes, that would pretty much kill things. Lack of a ChangeLog is evil | 07:29.15 |
chw | yep. that's one of the things i like most about open source. | 07:35.16 |
kens | OK looks like there may be a genuine problem creating the UTF data in the XMP, not entirely sure why yet. | 08:04.33 |
| I'm seeing three binary bytes for the u umlaut, when I only expected 2 | 08:05.55 |
chw | one thing i haven't mentioned yet: | 08:16.46 |
| when i dump the xml with poppler pdfinfo -meta, the encoding autoguessed by all text editors i've tried is ISO-8859-1. | 08:17.03 |
| the dump on the input file is guessed as UTF-8. | 08:17.10 |
| it's probably just another symptom of the same problem. | 08:17.20 |
kens | Yes I expet its the same thing, I'll have to debug through the code to see where its generating it, I didn't write this bit of code so I'm not familiar with it | 08:20.57 |
chw | ok, good to hear it's not just on my end. | 08:27.55 |
kens | Hmm, well that's not right. Its writing 0xfc directly instead of encoding it | 08:28.02 |
chw | yeah, i've seen it on the pdfinfo dump with the ISO-8859-1 encoding. | 08:29.17 |
kens | Whoever wrote this code made a serious screwup here | 08:29.38 |
| Anything over 0xAE doesn't get encoded | 08:29.46 |
| Which is nonsense for UTF-8 | 08:30.05 |
chw | and it's been like that for quite some time, right? i've tried gs from 9.19 up to current git, so the problem is not a new one. | 08:32.35 |
kens | Looks like I'm going to have to generate the missing half of a lookup table, this will take me a few hours and I'm going out in an hour or so for the rest of the morning. Should get it done by this afternoon though | 08:32.40 |
| chw as far as I can tell its *always* been like that | 08:32.50 |
| It certainly predates me owning this code, which is near 10 years now | 08:33.03 |
chw | ouch | 08:33.13 |
kens | I've never had occasion to look at this piece of code before, but its clearly wrong | 08:33.31 |
| It looks like th author of the code only filled in the values from 0x80 to 0xAE, anything above that is ignored and written unchanged | 08:35.22 |
chw | well, that explains just about everything i've seen then. | 08:37.06 |
kens | I'd imagine so yes | 08:37.17 |
chw | except pdf_info.ps | 08:37.25 |
kens | Oh well, that's probably some kind of error in pdf_info.ps | 08:37.39 |
| I'm rather less concerned by that :-) | 08:37.46 |
chw | probably. but it would be rather nice if it didn't die on every pdf gs generated. | 08:39.44 |
| anyway, thanks a lot for taking a look. | 08:39.59 |
| i really appreciate that | 08:40.05 |
kens | It might not after I fix this, but I'd prefer to treat it as a separate problem | 08:40.11 |
| NP thanks for reporting it | 08:40.20 |
chw | sure, np. | 08:40.37 |
| i'll recompile as soon as the fix lands, to make sure ;) | 08:41.00 |
kens | That would be good, thanks | 08:41.09 |
| Hopefully will have it done today, might be tomorrow, depends if the lilypond folk get back to me again | 08:41.31 |
chw | that's not a problem, take all the time you need. | 08:43.20 |
| oh, btw. the fix would land for 9.22, right? i'm not familiar with your release process... | 08:44.26 |
kens | Probably not, though that's not up to me | 08:44.38 |
| We've already done RC1 | 08:44.47 |
chw | yeah, i've seen. no backports? | 08:45.17 |
kens | So normally we would only add critical fixes to that, and since this has been outstanding for, well quite some time, it can't really be considered critical | 08:45.19 |
| We never backport | 08:45.25 |
| Well, unless someone is paying us... | 08:45.41 |
chw | ok, understood. i'll be sure to use git master then. | 08:46.22 |
kens | Usually the best policy, we don't often screw that up :-) | 08:46.42 |
hj | is it possible that there is problem when using page selection with ghostscript | 14:36.22 |
kens | All things are possible in an infinite universe | 14:36.46 |
hj | if I use -dFirstPage=1 -dLastPage=1 on the commandline i got a crash | 14:36.51 |
kens | Well, that would be a bug then | 14:36.59 |
hj | ./gs -r96 -sDEVICE=pngalpha -dFirstPage=1 -dLastPage=1 -o test.png /nas/support/customers/justcroft/id1205310003-bug693398-vmi20.ps GPL Ghostscript 9.19 (2016-03-23) Copyright (C) 2016 Artifex Software, Inc. All rights reserved. This software comes with NO WARRANTY: see the file PUBLIC for details. Loading NimbusMono-Regular font from %rom%Resource/Font/NimbusMono-Regular... 4097288 2725412 3894288 2602740 1 done. Querying operating | 14:37.33 |
kens | Yeha that's not helpful | 14:37.48 |
chrisl | Well, the "GPL Ghostscript 9.19" suggests one thing...... | 14:38.05 |
kens | If you want someone to look at it, you're going to have to open a bug report and attach the PDF file | 14:38.08 |
| True.... | 14:38.13 |
hj | I will open a bug report. | 14:38.25 |
kens | You should also try the currnt release first | 14:38.35 |
hj | If I leave out the -dFirstPage -dLastPage it converts | 14:38.52 |
kens | Not the point. You are using an old version of Ghostscript. | 14:39.03 |
hj | I tried 9.22rc as well | 14:39.08 |
kens | OK well that's helpful, in that case yes, open a bug | 14:39.28 |
hj | ./gs -sDEVICE=pngalpha -dFirstPage=1 -dLastPage=1 -r96 -o test.png vmi20.ps GPL Ghostscript RELEASE CANDIDATE 1 9.22 (2017-09-12) Copyright (C) 2017 Artifex Software, Inc. All rights reserved. This software comes with NO WARRANTY: see the file PUBLIC for details. Loading NimbusMonoPS-Regular font from %rom%Resource/Font/NimbusMonoPS-Regular... 4335724 2931975 3894424 2605901 1 done. Querying operating system for font files... Segment | 14:39.50 |
| I will open a bug and attach the file | 14:40.13 |
kens | You don't need to spam the channel with logs, I believe you | 14:40.21 |
| chw I just commited a (trivial) fix and updated the bug report | 14:42.01 |
chw | thanks ken. i'll recompile and retest the pdf files first thing tomorrow morning. | 14:44.31 |
kens | No rush | 14:44.45 |
rammanoj___ | I want to render a pdf to png and want to set the quality of it using ImageMagick i.e by using convert command I searched and got the below command gs -sDEVICE=jpeg -sOutputFile=- -dFirstPage=1 -dLastPage=1 -r150 -dBATCH -dNOPAUSE -q "inputfile" | convert -depth 8 -"outputimage" but gave error in my terminal is there any error in the above command if any can you say the command so that both can be done | 14:48.51 |
| simultaneously? | 14:48.51 |
kens | Well you haven't said what the error is | 14:49.29 |
| So how can I tell ? | 14:49.37 |
rammanoj___ | convert: no images defined `-imagename' @ error/convert.c/ConvertImageCommand/3254. | 14:50.12 |
kens | Well that's not a Ghostscript error | 14:50.22 |
| Its a convert error, we cna't help you with that. | 14:50.33 |
rammanoj___ | okay thankyou | 14:50.43 |
hj | sorry for the log spam. bug is reported. 698568 | 14:53.17 |
kens | OK I'll get it in a bit | 14:53.25 |
| Hmm PostScript input, that's very old school :-) | 14:53.51 |
chrisl | rammanoj___: it may be pasting error, but in your convert command, there's no space in '-"outputimage"' - if that's the case, then probably try '- "outputimage"' | 14:55.30 |
hj | I was making a thumbnail of the first page and then it crashes. | 14:57.26 |
kens | I won't get to it today, or possibly this week, I'm very busy right now. You'll get email frmo the bug tracker when I can get round to looking at it, and comment on it though | 14:58.21 |
rammanoj___ | thankyou for help chrisl | 15:10.13 |
chrisl | Did it help? | 15:11.04 |
rammanoj___ | the command is correct as said by you but there was some problem in convert command running so even a simple convert statement is not working | 17:16.28 |
| I think it was some config fault but thanks for help | 17:17.16 |
renee | How do you clear a Microsoft Visual C++ Runtime Library Error? | 17:36.02 |
| Forward 1 day (to 2017/09/21)>>> | |