| <<<Back 1 day (to 2014/11/04) | 20141105 |
ahos | Good evening, I'm trying compile mupdf's to android, but "make generate" command doesn't do anything. | 03:41.08 |
| Could someone help me here, please? | 03:41.29 |
| When I run that command I receive this message: "File not found - *.[ch] | 03:42.23 |
| File not found -.[ch] | 03:42.35 |
| make: Interrupt/Exception caught (code = 0xc00000fd, addr = 0x4227d3)" | 03:43.44 |
neosimsim | Hey guys. I'm again trying to open an aes encrypted PDF file using. stream = fz_new_stream(...); stream = fz_open_aesd(stream, "c6...d6", 32); Now I get "Fatal signal 11 (SIGSEGV), code 1, fault addr 0xfe11 in tid 32303 (.m_und_h.muhpdf)". I'm working on Android btw. Thanks for help. | 10:25.34 |
Robin_Watts | neosimsim: We're going to need a lot more information than that before we can even begin to guess. | 10:28.25 |
kens | OK be back in a couple of hours.... | 10:34.35 |
neosimsim | I'm trying to display an aes encrypted PDF on an Android device. The file starts with the iv. | 10:35.42 |
| I do not mean the pdf internal entryption. | 10:35.58 |
| For my first approach I simply tried to edit the mupdf.c in the Android jni dir. I just created a new gist. | 10:44.00 |
| https://gist.github.com/anonymous/30c8e637f0780a4a2081 | 10:44.06 |
| I first tried to dilpay a nonencrypted pdf wich works fine. Then I tried the same with an encrypted pdf and added line 54 | 10:45.02 |
| I was hoping this should be enough but obiously I'm missing something. | 10:45.34 |
Robin_Watts | line 53. Should stream be encStream? | 10:49.32 |
| And you're assigning stream->seek = bufferStreamSeek. | 10:50.24 |
| Move line 55 to before line 54. | 10:50.57 |
| And change encStream to stream. | 10:51.01 |
| The aes stream almost certainly doesn't support being seeked within. That will give you problems. | 10:52.52 |
neosimsim | OK thank you I think I'm a step further. Now I get some other problems starting with "error: cannot recognize version marker". | 10:55.33 |
Robin_Watts | Ok, show me the latest version so I can talk you through the problem. | 10:56.16 |
neosimsim | I'm I right that the aesd stream expecets the IV to be the start of the document? | 10:56.18 |
| New gist: https://gist.github.com/neosimsim/62394a699e21d28dd2bf | 11:00.13 |
| I added the errors as comment | 11:00.22 |
Robin_Watts | Ok, so line 53 sets up a stream that will read from a buffer. | 11:00.58 |
| And line 54 tweaks it so that it can cope with seeking within the buffer. | 11:01.17 |
| Line 55 then sets up a decryption stream that is based upon the buffer stream. | 11:01.50 |
| The decryption stream does not support seeking. | 11:02.10 |
| Thus whenever the PDF reading code seeks within the stream (and it does this a lot), the decryption stream will go wrong. | 11:02.43 |
| You might get data errors, you might get crashes. | 11:02.52 |
| So, quite apart from whether the AES stream is going to work with your AES encryption or not, this will not work. | 11:03.30 |
| In general, seeking within encrypted streams is not possible, because to decrypt something at offset X, you have to have decrypted everything before X. | 11:04.24 |
| hence you need to modify your approach. | 11:04.37 |
| Either you need to completely decrypt the incoming file and then store it in a memory buffer somewhere, and then build a stream from that... | 11:05.03 |
| or you need to use a different form of (less secure) encryption where you can seek. | 11:05.27 |
neosimsim | Ok, so I think I'll have to write an own kind of stream which support my needs. | 11:09.26 |
| First I'll overthink my approach. Thanks a lot. | 11:09.57 |
Robin_Watts | np | 11:10.11 |
neosimsim | One more question. Is it correct that the IV should the first in the file? | 11:15.42 |
Robin_Watts | neosimsim: Not a clue, sorry. | 11:22.55 |
nsz | tor8: found a bug in splice but i dont yet understand how the array code works | 15:59.12 |
| a="abcde".split(); a[15]='x'; b=a.splice(0,4); print(a.join("")) | 16:00.23 |
| the splice does something wrong after a[15]='x' (which adds a lot of undefined) | 16:01.06 |
| the returned value is ok (b) but the wrong values get deleted from a | 16:01.40 |
tor8 | nsz: yes. | 16:04.33 |
| nsz: probably an off-by-one somewhere :( | 16:05.09 |
nsz | should have been .split("") | 16:10.48 |
henrys | marcosw: macpro is now yosemite - the dashboard has said "updating ghostscript" for some time now. Not sure what that's all about. | 16:18.33 |
| marcosw: now it says removing from node pool... guess I should have checked with you first. | 16:20.35 |
Robin_Watts | henrys: That probably means the macpro isn't sending the heartbeat signal to the cluster master any more. | 16:25.09 |
| Have you rebooted since installing Yosemite? | 16:25.28 |
henrys | Robin_Watts: that's just cron right? | 16:25.37 |
Robin_Watts | Yes, I would imagine so. | 16:25.53 |
| Are the cron jobs still running? | 16:26.06 |
| i.e. the crontab didn't get blitzed during the upgrade? | 16:26.19 |
tor8 | nsz: looks like it's not deleting the indexes like it ought. or rather, not succeeding in moving the undefined values to overwrite the deleted part | 16:27.37 |
henrys | crontab report clusterpull.sh is going regularly | 16:27.41 |
| s/report/reports | 16:27.54 |
| marcosw will probably know right off ... I'm not slowing things down to much by it begin down. | 16:28.34 |
| Robin_Watts: otherwise a painless upgrade, usually a few things break for me. | 16:30.05 |
| I guess they didn't do anything interesting ;-) | 16:30.28 |
mvrhel_laptop | morning | 16:30.37 |
Robin_Watts | henrys: Mostly cosmetic this time, unlike mavericks which was all under the hood. | 16:30.54 |
| henrys: seems to be partaking in the cluster test now. | 16:34.36 |
henrys | hi mvrhel_laptop | 16:34.54 |
| mvrhel_laptop: have you tried clusterpushing with xpswrite changes yet? marcosw didn't respond about that. | 16:35.21 |
tor8 | nsz: bug fix pushed | 16:37.24 |
nsz | thanks | 16:38.48 |
Robin_Watts | henrys: As far as I can tell, there are no xpswrite tests in the cluster. | 16:44.00 |
rayjj | chrisl: thanks for responding to Len. I am also somewhat at a loss w.r.t. the 906 performance change. | 17:18.10 |
| chrisl: between you and me, I don't really trust that they are running on h/w that is identical | 17:19.05 |
Robin_Watts | rayjj: Is he not running both the 8.71 and the 9.06 derived code back to back on the same platform? | 17:30.48 |
chrisl | rayjj: I'm only really just getting to do anything - as usual, it took me the best part of the afternoon to get the 9.06 sim to even build :-( | 17:32.14 |
| rayjj: do you see any measurable difference on Windows? | 17:32.57 |
henrys | rayjj: the meeting will be white tie. | 17:42.33 |
rayjj | chrisl: no, but I'm going to move to a much slower machine to see if that makes a difference. I have to install VS 2008 there, and everything is slow there, even installing :-( | 17:42.53 |
| henrys: that's what I thought. I've seen movies of how they dress in England (mostly black and white) and it looked pretty formal ;-) | 17:43.37 |
henrys | lots of starch | 17:43.57 |
rayjj | henrys: but I asked because after a show here in LA one time Scott and Miles and I took a customer to dinner and I had to wear one of the 'loaner' jackets because of the dress code. LA -- can you imagine ? | 17:45.17 |
Robin_Watts | rayjj: There are lots of restaurants in london that have dress codes. | 17:46.24 |
rayjj | of course that was about 8 years ago. I'm sure if I had been a movie star or something, they wouldn't have asked :-/ | 17:46.35 |
Robin_Watts | I don't believe Artifex has the budget to take us all to one :) | 17:46.36 |
| English people generally dress very poorly. | 17:47.12 |
rayjj | Robin_Watts: oh, it isn't just the Artifex staff ? | 17:47.33 |
| ;-) | 17:47.41 |
Robin_Watts | The french dress very well (but forget to wash, so we still have the high moral ground) | 17:47.52 |
rayjj | or at least the higher olfactory ground | 17:48.18 |
Robin_Watts | The italians are obsessed with dressing well. Haven't mastered talking without waving their hands about, and governing without rampant corruption, but they do dress nicely. | 17:49.01 |
| Now the germans dress very badly. Look at Angela Merkel and her velour romper suits. | 17:49.43 |
rayjj | Robin_Watts: I suspect it's related to how they drive -- they want to look nice when they go to the undertaker | 17:49.59 |
Robin_Watts | will stop now before continuing to alienate any more of europe. | 17:50.06 |
rayjj | Robin_Watts: and the driving correlates with the German dress -- they drive relatively well | 17:50.39 |
Robin_Watts | They drive for the fatherland. No speed limits on autobahns. | 17:51.30 |
| Italians drive like they talk. Very erratically, very quickly, and far too close to other people. | 17:52.09 |
| With lots of hand gestures. | 17:52.16 |
rayjj | Robin_Watts: yeah, I know. I had a really fun time in a high-end rental cruising at 140 mph from Munich to Innsbruck | 17:52.21 |
Robin_Watts | If you can find it, I highly recommend an Italian TV series called "Gommorah". | 17:53.00 |
rayjj | Robin_Watts: it's on Netflix, I think | 17:54.06 |
| yep -- if it's the one about a Naples mob | 17:56.08 |
Robin_Watts | That's it. | 17:58.15 |
arafath | hello | 18:22.42 |
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. | 18:22.42 |
arafath | hope you all are fine | 18:23.00 |
| I am trying to work with MuPdf in my android app | 18:23.34 |
| I have few pdf files with single page for each , I have to view each one in my embed the pdf views in my girdview | 18:24.20 |
| how can I proceed , can anybody please help me , I have tried to put MuPDFPageView in the getView of my custom grid adapter.. but it through's some error | 18:25.11 |
Robin_Watts | arafath: So, just to restate your problem, so I can be sure I am understanding you... | 18:30.23 |
| You want to do a MuPDF based Android app, in which you will display multiple 1 page PDF files in a gridview. | 18:30.57 |
| Such a thing is certainly possible using MuPDF. | 18:32.05 |
| Indeed, for a customer, we did a version of MuPDF that would show n 1-page PDF files, rather than 1 n-page PDF file. | 18:32.36 |
| But the changes will almost certainly require changes at the mupdf.c level. | 18:33.15 |
| i.e. it won't be something you can do purely in java. | 18:33.25 |
arafath | Robin_Watts : Thanks for your reply. | 18:35.57 |
| can you please tell me that, I have seen that, MuPDFReaderView is used as docViewer in the MuPDFActivity | 18:38.07 |
| in which MuPDFPageView is being added for each of the page | 18:38.27 |
| so .. MuPDFPageView can be added into my own layout placeholder right | 18:40.41 |
| ? | 18:40.42 |
Robin_Watts | arafath: I wouldn't like to swear to it. | 18:55.24 |
fredross-perry | My $0.05 on the Qt/windows gsview build discussion. (a) It's not very un-windowsy, and I like having the same UI in all three places. (b) of the 99 Mbytes, 43 of them are ours, not from Qt. (c) One could build Qt as static libraries, which may be smaller, or may not. (d) I do have "save as XPS", but is it useful for files with lots of text? (e) I think some aspects of the Qt version are better than the existing | 21:01.59 |
| Windows version. | 21:01.59 |
nsz | tor8: you committed a debug printf in the array delindex fix | 22:32.05 |
avih | tor8: you've left a printf at http://git.ghostscript.com/?p=mujs.git;a=commitdiff;h=98744d09a42fc3591825c06ce37bfb2041687e35 . also without following the full logic, you sure it should be idx < 0 ? idx - 1 : idx rather than 1 - idx at http://git.ghostscript.com/?p=mujs.git;a=commitdiff;h=9ed091470480513808c8f1cdb2707b2ceb501c90 ? | 23:05.22 |
| oh, i should have read the line above me :) | 23:08.25 |
| hmm.. how is the behavior defined with negative index? | 23:11.21 |
| also, unrelated, how soon is the github repo updated? is it automatic? | 23:16.15 |
| ah. idx -1 is ok since you just pushed the number. | 23:18.48 |
| Forward 1 day (to 2014/11/06)>>> | |