| | <<<Back 1 day (to 2014/10/15) | 20141016 |
| meluha | is there way ghostscript can generate pdf from pdf by using only ASCII85 filter type? | 02:35.28 |
| dinamic | morning folks | 05:46.30 |
| | api ref. for mujs should mentioned that js_getregsitry() puts a undefined object on stack if not exists. | 06:05.51 |
| | glad i implemented a dumpstack helper :) | 06:06.08 |
| | also js_try() / js_endtry() would be nice to have in public api | 06:06.34 |
| Chandler_ | Hi | 08:49.34 |
| 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. | 08:49.34 |
| test__ | halo | 09:35.17 |
| dinamic | hi tor8, what approach should i take if i want to iterate over keys in an object as i can do with js_getindex() for arrays in mujs ? | 10:55.01 |
| | i noticed that the object proto have keys function, should i just get prototype.keys and do a js_call() ? | 10:55.47 |
| | or is there another approach for objects due to arrays have public api for iteration | 10:56.48 |
| tor8 | which I guess is a bit of a harsh limitation that should be fixed | 11:40.21 |
| | dinamic: http://git.ghostscript.com/?p=user/tor/mujs.git;a=commit;h=54afaa6e91330b2edb2255f983cb7ae71a473fa5 | 11:50.05 |
| | bah, typo in that commit. use this instead: http://git.ghostscript.com/?p=user/tor/mujs.git;a=commit;h=790103815fb3c5299c1b4afd2f7075dc0aca4a17 | 11:51.06 |
| | dinamic: otherwise, as you say, js_getglobal("Object"); js_getproperty("getOwnPropertyNames") (or "keys") and js_call that will work just fine | 12:01.43 |
| nsz | tor8 i have a set of easy math fixes for mujs i can send them to you in some form (eg through bugzilla) or what is your prefered way? | 12:23.40 |
| | the string formatting is harder, there are more gratuitous c vs js incompatibility than i expected so i'll postpone that now | 12:24.48 |
| tor8 | nsz: bugzilla or email both work | 12:29.25 |
| nsz | ok | 12:29.39 |
| | (will do it in the evening) | 12:30.14 |
| dinamic | tor8: nice, that was a small fix for a larger problem.. :) | 12:37.15 |
| | tor8: i guess the official mujs repo is ccxvii/mujs, what is the merge policy from yours to upstream ? | 12:53.09 |
| | Aah :), Artifex är svenskt... | 12:53.54 |
| | nevermind, found that git://git.ghostscript.com/mujs.git is the upstream | 12:55.48 |
| tor8 | dinamic: no, I'm the only swede employed by artifex | 13:01.02 |
| | git.ghostscript.com/mujs.git is the official one, I just keep a mirror on github | 13:01.18 |
| | dinamic: the rest of us are spread around the world, mostly in the UK and the US | 13:02.07 |
| pipitas | I'm trying to stitch together a font query that is to be sent to a proprietary PS interpreter. When I test it against Ghostscript, Ghostscript's response is to deny even fontnames I know for sure are in its $FONTPATH. | 13:05.14 |
| | (I've been using Example G.10 from the PLRM2.pdf on page 695. Even this very example extended by the 14 base font names returns as well as by the Nimbus names returns "Font /Fontname: No" for each case.) | 13:06.35 |
| | Could it be that Ghostscript just doesn't support this standard Level 2 feature of querying the interpreter for available font resources? | 13:07.27 |
| kens | WOuld seem more likely you are doing something wrong. WHy are you suing tghe PLRM2 rather than the 3rd edition ? | 13:07.55 |
| | Yeah that's a dumb old way to do it. | 13:09.09 |
| | It assumes fonts are files on disk | 13:09.23 |
| | Use level 3 constructs,it will work better | 13:09.46 |
| ChandlerLiu3 | hello, everybody! | 13:09.56 |
| kens | pipitas see resourceforall | 13:10.13 |
| pipitas | I'm using PLRM2 because the proprietary interpreter to be finally tested (not available here right now) is explicitely Level 2. I'm only practising with Ghostscript for now. | 13:11.03 |
| kens | (*) {==} 256 string cvs /Font resourceforall | 13:11.09 |
| | pipitas, font enumeration in level 2 is device depdendent, because it assumes fonts are files stored on disk. So code that works on one device won't work on another, that's one of the reasons for the existence of resourceforall. | 13:12.01 |
| ChandlerLiu3 | I plan to add some text comment/remark into a PDF view, but I want to save the comments separately, can we do this on MuPDF? if yes, do we any hints, such as which API? I'm working on Android | 13:12.13 |
| | I saw there is annotation features, but I feel a little difficulty to understand the APIs. | 13:13.22 |
| pipitas | kens: Well, Adobe calls this query "a typical font query" and says about it this: "[it] ... presumes that a file system is available to the Post- Script interpreter. This is not the case in all implementations" -- and I thought Ghostscript surely has a file system available here | 13:15.13 |
| | kens: thanks for the hints anyway... | 13:15.24 |
| kens | pipitas, yes but Ghostscript is a level 3 interpreter, and so is not restricted to the conventions of a level 2 ionterpreter. It does not store its fonts in /fonts on disk. | 13:17.26 |
| | Which is what that code assumes | 13:17.37 |
| | While that query may be 'typical' its by no means universal, and definitely *won't* work on a wide vareity of PostScript interpreters, including pretty much every level 3 interpreter ever. | 13:18.25 |
| pipitas | kens: My assumption was that L3 interpreters are backward compatible and can always handle L2 code. And for this query, when I sent it to Ghostscript I added '-sFONTPATH=/Library/Fonts' where some of these fonts are picked up by GS in other instances (when I do real stuff)... | 13:20.57 |
| tor8 | ChandlerLiu3: save the comments separately, you mean in a different format in a separate file? | 13:21.18 |
| ChandlerLiu3 | tor8: I can handle the data structure that MuPDF needs to render the comments | 13:23.02 |
| tor8 | if you want to store the comments in the PDF file you want to use annotations | 13:23.48 |
| | otherwise you can always ask mupdf to render more content to the page bitmap, much like we do with annotations (render them in a separate pass after the main page content) | 13:24.19 |
| kens | pipitas, well that's an incorrect assumption anyway, even for the language. copypage semantics are different in level 3. In addition what you are looking at its not a defined portion of the language, its a convention which was more or less true in level 1 and much less true in level 2. | 13:24.21 |
| | pipitas adding FONTPATH won't help, the code explicitly looks on the disk for a directory called /fonts and assumes the files stored there have names which reflect the name of the font. THis is because early Adobe rips, based on Unix-like file systems stored tehir fonts this way. Of course that *never* worked on interpreters based on (eg) DOS or Windows. | 13:25.31 |
| | FontDirectory does work though | 13:26.38 |
| | THoguh it returns an empty dictionary :-) | 13:26.56 |
| | Basiclaly that PostScript makes assumptions which were to some degree legitimate for some level 2 interpreters. Its the best that can be done in level 2 because it lacks better support in the language. It's implementation dependent and pretty much guaranteed not to work on a level 3 interpreter, which has much better ways to do the same job. | 13:28.18 |
| | So it may (or very possibly may not) work on the printer you intend to use this on, it won't work with Ghostscript and its not a bug in Ghostscript | 13:28.47 |
| pipitas | kens: I had noticed that the query looked for fonts/ and changed this in my version. I also included some items in the query using the filenames from the disk but it still didn't work. | 13:35.24 |
| | kens: But OK, anyway -- thanks a lot for your help | 13:38.31 |
| dinamic | so flare1 isnt bound at all to SDL | 13:58.49 |
| | sorry ^ | 13:58.58 |
| kubilayrd | Hi all, can I change conversion naming starting from 0 instead of 1? | 15:09.22 |
| | with a single line of command, of course | 15:09.46 |
| kens | You'regoing to have to be much more specific. WHich product, and starting *what* from 0 ? | 15:09.49 |
| kubilayrd | sure, converting a PDF files pages to JPEG images | 15:10.13 |
| kens | Using what product ? | 15:10.29 |
| kubilayrd | ghostscript | 15:10.37 |
| kens | Presuimably you mean the %d file naming syntax, and the answer is no | 15:10.56 |
| kubilayrd | well, thanks | 15:11.12 |
| HoloIRCUser2 | Hey folks, any mupdf devs online? | 15:34.02 |
| Robin_Watts | HoloIRCUser2: Yes. | 15:34.14 |
| HoloIRCUser2 | Grand, I've encountered a slight issue, not expected it to be fixed anytime soon but the bug report page was down for me so I thought I'd log on here | 15:35.35 |
| | *expecting | 15:35.43 |
| Robin_Watts | really? | 15:35.48 |
| | bugs.ghostscript.com seems to be working for m.e | 15:36.10 |
| HoloIRCUser2 | Well it's working again now, it was probably just my end | 15:37.02 |
| | Anyhow it's just a minor thing, when dealing with some pdfs that are single pages split into two halves, if I try to copy any text the app doesn't recognise the 'break' in the page and selects across the line | 15:39.02 |
| | Should I put in a bug report or is it a known issue with certain pdfs? | 15:39.35 |
| | Also hi, how are you, hope you're well | 15:40.13 |
| | And thanks for an otherwise stellar reader | 15:40.37 |
| Robin_Watts | HoloIRCUser2: You mean pages with text in 2 columns ? | 15:42.39 |
| HoloIRCUser2 | Yeah | 15:42.54 |
| Robin_Watts | That's a known bug. | 15:42.58 |
| HoloIRCUser2 | Ah okay | 15:43.18 |
| Robin_Watts | but a bug including an example PDF would be useful to ensure it's not forgotten. | 15:43.19 |
| HoloIRCUser2 | Will do, I'll submit one now with an attachment. Product & Component for the report? | 15:47.25 |
| | Ghostscript/pdf interpreter? | 15:48.01 |
| Robin_Watts | MuPDF. | 15:48.10 |
| | MuPDF is completely separate from Ghostscript. | 15:48.26 |
| HoloIRCUser2 | That would make more sense. | 15:48.26 |
| | Submitted the bug (No. 695603) | 15:59.28 |
| | Hope it's of some help | 15:59.39 |
| | Oh and I included an example pdf with it | 16:03.03 |
| gav7 | hi | 16:36.24 |
| 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. | 16:36.24 |
| gav7 | I'd like to add 3 suggestions for the MuPDF program (on windows 7): 1) ability to print the document on paper, 2) scrollbar on the side and the ability to turn it on/off, 3) continuous mode and the ability to turn it on/off | 16:41.16 |
| Robin_Watts | gav7: Checkout the new gsview app. | 16:49.34 |
| | Not officially released yet. | 16:49.44 |
| | mvrhel_laptop: Is there a place people can get the current beta version? | 16:49.58 |
| gav7 | MuPDF bug (on windows 7): after exiting full screen the backround (the surrounding) is not rerenderd | 17:06.43 |
| mvrhel_laptop | Robin_Watts: I really want to get the fixes in for xpswrite first. I have not put it up yet | 17:20.13 |
| | printing is slow slow when images are present | 17:20.20 |
| Robin_Watts | mvrhel_laptop: It's roughly 1000 times better than the alternative already though, so... | 17:20.56 |
| | and there will always be "just 1 more thing" | 17:21.11 |
| mvrhel_laptop | I suppose. OK. who has access to getting gsview.com set up? we have the domain name that is all that I know | 17:21.49 |
| | if I knew where to put the web stuff, I could get it set up or if someone else wants to do it that is fine too | 17:22.28 |
| Robin_Watts | mvrhel_laptop: Presumably gsview.com would be hosted on casper? | 17:23.00 |
| chrisl | It is - /var/www/gsview.com | 17:23.30 |
| mvrhel_laptop | aha | 17:23.37 |
| | just the man to set it the webpage... ;) | 17:23.51 |
| | s/it/up | 17:23.57 |
| chrisl | I'll set something up tomorrow | 17:24.07 |
| mvrhel_laptop | ok. cool. I will make the latest x64 and win32 installers | 17:24.26 |
| | and get those to you late today/tonight | 17:24.37 |
| | thanks chrisl | 17:24.42 |
| Robin_Watts | mvrhel_laptop: call the installers gsview-0.01-EARLY-ACCESS-TEST.msi or something :) | 17:24.45 |
| mvrhel_laptop | yes. good idea | 17:24.56 |
| chrisl | mvrhel_laptop: if you just stick them in your home dir on casper and let me know the names | 17:25.13 |
| mvrhel_laptop | chrisl: ok. I will do that. thanks | 17:25.45 |
| henrys | chrisl: changes for your review in my repo - welcome back! | 17:37.45 |
| chrisl | henrys: thanks. I did have a look at your changes. My comment was that having separate accessor methods for pcl and pxl seemed over complicated - wouldn't a common "pl_" one do? | 17:39.32 |
| henrys | chrisl: yeah I guess I could change that for the other stuff as well. | 17:43.39 |
| | icc etc. | 17:43.48 |
| chrisl | henrys: beyond that stylistic observation, I'm good with the changes | 17:44.56 |
| henrys | thanks chrisl | 17:45.13 |
| tomty89 | chrisl_away: hello | 18:26.30 |
| mvrhel_laptop | bbiab | 18:31.50 |
| rayjj | hmm... I wonder if the company M board 'posync' has problems. I can run all the PCL jobs with -dBGPrint=true and they work, although there is a BIG performance hit on page 8 of J11_acrobat.prn (that I _don't_ see on Windows) | 23:18.38 |
| | but running PS and PDF, I get all of the Outputpage start ... Outputpage end messages, then "GPL Ghostscript GIT PRERELEASE 9.16: Attempt to delete a memfile still open for read: 0x1a032b8" | 23:19.42 |
| | except (like English spelling rules) one file works -- PLRM_100.ps | 23:20.57 |
| | Except for the anomalous PCL j9_acrobat.prn page 8 and slightly worse (0.8x) PLRM_100.prn the performance is better, but with the errors I don't know what to trust | 23:24.45 |
| | Forward 1 day (to 2014/10/17)>>> | |
IRC Logs