| | <<<Back 1 day (to 2013/11/27) | 2013/11/28 |
| Robin_Watts | henrys: You can code the excel support then :) | 00:10.06 |
| henrys | I'm on it | 00:12.55 |
| tor5 | it's still there (in /home/git/ghostpdl.git/hooks/update.bak) if we want to re-enable it | 09:59.51 |
| | I also think the cluster wanting a linear history to test (and not having to test everything twice, once for the real commit and once for the merge) was a factor in the policy | 10:00.45 |
| chrisl | tor5: Does it get overridden if we do a forced push? | 10:01.02 |
| kens | Not testing twice sounds like a plan | 10:01.30 |
| tor5 | chrisl: I don't think so | 10:03.08 |
| | we could make it test user names, to only allow specific people to push merges | 10:03.26 |
| chrisl | tor5: let's see what henrys thinks - it doesn't exactly crop up very often. | 10:05.36 |
| | tor5: or we could create a "gs-merge" group on casper with just the artifex employees in the group, and check on that | 10:07.27 |
| tor5 | chrisl: that would be good too | 10:08.21 |
| Robin_Watts | tor5: ping | 11:05.43 |
| | morning paulgardiner | 11:06.02 |
| paulgardiner | Hi | 11:06.16 |
| ghostbot | niihau | 11:06.16 |
| tor5 | Robin_Watts: so, bug 694810 ... an obviously broken PDF which we can handle if we run the xref repair | 11:08.24 |
| | but detect that is broken too late. | 11:08.30 |
| | I think we ought to catch certain classes of object loading errors and rebuild the xref if they trigger | 11:08.52 |
| | also: pong :) | 11:09.10 |
| Robin_Watts | tor5: So, what's the state of play with us taking openjpeg etc from ghostpdl? | 11:10.24 |
| | My memory is so crap :( | 11:10.34 |
| | Shelly reports that that will solve all (or close to all) our remaining SEGVs in the fuzzing files. | 11:11.13 |
| | I have a commit on my repo with such a change in it. | 11:12.06 |
| | I can't remember why it was never pushed. | 11:12.13 |
| tor5 | Robin_Watts: the openjpeg repo branch mess, IIRC | 11:12.58 |
| | but that's only a vague recollection | 11:13.07 |
| | we've got it set up to track upstream svn via git-svn | 11:13.24 |
| | then there are patches on ghostscripts directory copy | 11:13.34 |
| Robin_Watts | tor5: I thought you had the gs copy being mirrored into a separate git repo automatically now ? | 11:14.04 |
| tor5 | Robin_Watts: yeah, I think I set that up. let me just double check. | 11:14.25 |
| | there was also something about tagging so we don't lose mupdf git history since we'll be dropping the branch we used to use in the submodule | 11:16.02 |
| | Robin_Watts: okay, so it looks good to go to me. | 11:20.27 |
| | I probably jsut forgot to give you the okay to push, and then we both forgot about it :) | 11:20.43 |
| | let me just push a tag on the thirdparty/openjpeg.git to where we are now so we keep history | 11:21.10 |
| Robin_Watts | tor5: I'm just running a cluster test now to double check. | 11:22.01 |
| tor5 | and then we should prune the working branches there too, I think | 11:24.35 |
| | so it looks less confusing | 11:24.40 |
| | we have 10 commits that are referenced from mupdf, I'll just check that they're all in a tagged reachable location so they don't get garbage collected | 11:26.22 |
| Robin_Watts | 373 diffs, just checking the bmpcmp now, but that number rings a bell. | 11:47.39 |
| | tor5: There are some other commits on my repo too. | 11:57.19 |
| | 2 in fact. | 11:57.45 |
| | ignore the top 3. | 11:57.50 |
| | tor5: OK, bmpcmp looks good. I'll push it asap. | 12:11.26 |
| | s/asap/as soon as you OK it/ | 12:13.26 |
| kens | Robin_Watts : ping | 13:25.55 |
| tor5 | Robin_Watts: bottom 3 look good (bug fixes and openjpeg) | 13:40.45 |
| paulgardiner | tor5: ping | 13:53.30 |
| tor5 | paulgardiner: yes? | 14:03.56 |
| Robin_Watts | tor5: Thanks. | 14:04.51 |
| kens | Robin_Watts : ping | 14:05.01 |
| tor5 | Robin_Watts: there's a fix on tor/master for you as well | 14:05.02 |
| | actually, I need to amend it first to deal with android, one min | 14:05.32 |
| paulgardiner | tor5: I'm just trying to understand how the library part of the build is provoked for iOS. | 14:05.41 |
| | Since pinging, I discovered build_libs.sh | 14:06.01 |
| | I need to pull in different a different js implementation | 14:06.26 |
| tor5 | paulgardiner: ah, build_libs.sh as a "Run Script" external target | 14:06.41 |
| | with a lot of voodoo to get the right build parameters | 14:06.48 |
| oznt | hi guys | 14:06.59 |
| tor5 | Robin_Watts: okay, now there's one for review | 14:07.21 |
| | please test that the android and win32 builds still work before approving it, I can't test those at the moment | 14:07.45 |
| oznt | i am trying to read pdf's table of contents , and the labels i get for the links are all of set by +1 , is there something i should know about labels in pdfs? | 14:07.51 |
| tor5 | oznt: page numbering starts at 0 | 14:08.14 |
| | in the internal mupdf functions | 14:08.26 |
| paulgardiner | tor5: Xcode is doing its usual excellent job of hiding where the setting live within the menus | 14:08.54 |
| tor5 | the various command line tools offset so that the user-visible pages start at 1 | 14:08.55 |
| Robin_Watts | tor5: will do. | 14:09.13 |
| oznt | tor5, so my page numbering would have been -1 offset | 14:10.00 |
| tor5 | oznt: are you using "mudraw -l" or fz_load_outline? | 14:10.25 |
| oznt | tor5, mudraw is working fine | 14:12.37 |
| kens | Which host has ghopstpdl origin/master ? | 14:12.48 |
| oznt | except that it outputs, page numbers, not labels | 14:12.52 |
| tor5 | oznt: I'm sorry, then I don't really understand what you mean. which labels are wrong (and offset how?) | 14:14.59 |
| kens | a specimen file to look at might help | 14:15.22 |
| oznt | what i mean is: when i open the pdf, it has some pages labeled "i, ii, v, x" (so the numbers are 1,2,5, 10) | 14:15.26 |
| Robin_Watts | kens: casper holds the golden repo. | 14:15.30 |
| | if that's what you mean by 'origin', | 14:15.36 |
| | Do: git remote -v | 14:15.52 |
| kens | Robin_Watts : so why do I get a message about the server key not beiung cached when I try to use git pull ? | 14:16.15 |
| Robin_Watts | kens: hmm. no idea. | 14:17.01 |
| tor5 | kens: sounds like ssh griping, does it warn if you run ssh manually from the git command line? | 14:17.04 |
| Robin_Watts | When did it last work? | 14:17.08 |
| kens | tor5 : I can never run ssh from Msys git | 14:17.16 |
| oznt | tor5, can i output the labels like that? r | 14:17.23 |
| kens | Robin_Watts : it still works on my desktop, but not my laptop | 14:17.28 |
| Robin_Watts | kens: Oh, right, that makes sense. | 14:17.49 |
| kens | I can use Putty to ssh from windows and it did add the server id | 14:17.50 |
| tor5 | oznt: oh, you mean the page number labels? sorry, we don't support those :) | 14:18.08 |
| Robin_Watts | That's cos casper moved IP address a while ago | 14:18.14 |
| tor5 | oznt: we never bother to parse them | 14:18.22 |
| Robin_Watts | do git remote -v and tell me what the line for origin says. | 14:18.45 |
| kens | Robin_Watts : this is a new laptop I'm setting up (or more accurately failing to set up) | 14:18.50 |
| Robin_Watts | I see. | 14:18.59 |
| oznt | tor5, ok, thanks, i was hoping for some salvation with mupdf ... i have to use poppler but it's documented awefuly ... | 14:19.14 |
| kens | I think its a plink error I'm getting | 14:19.27 |
| Robin_Watts | kens: Probably. but I can't solve it until you do git remote -v and tell me what the line for origin says. | 14:19.56 |
| kens | I believe I have verything I need for plink to work, but it keeps complaining the server's host key is no in the cache | 14:20.06 |
| tor5 | oznt: it should be possible to read the /PageLabels entry of the catalog and add the feature relatively easily, but we've never really found a need for it | 14:20.09 |
| | which isn't to say we will refuse a patch for it :) | 14:20.19 |
| oznt | tor5, I am working on a project where the client WANTS it | 14:20.45 |
| kens | Robin_Watts : git remote -v says | 14:20.48 |
| | ken@ghostscript.com:/home/git/ghostrpdl.git (fetch) | 14:20.48 |
| Robin_Watts | right. So do: plink ken@ghostscript.com | 14:21.07 |
| oznt | but ... i don't have enought time during this project for digging in the C code | 14:21.11 |
| kens | Robin_Watts : that looks better | 14:21.38 |
| Robin_Watts | kens: with luck that should ask you about a key. | 14:21.41 |
| | Once it's done that, you can then try git pull again. | 14:21.59 |
| kens | yes and I told it ti accept it, now i can oull, thanks! | 14:22.11 |
| Robin_Watts | fab. | 14:22.15 |
| tor5 | oznt: you could add a bug for it, marked as an enhancement, saying something to the effect of "make an API to access the PDF page labels" | 14:22.25 |
| kens | wonders what other wrinkles I have yet to fall into | 14:22.29 |
| Robin_Watts | tor5: Android works, windows doesn't. | 14:22.32 |
| tor5 | and maybe one of us will get to it on a lazy afternoon | 14:22.35 |
| Robin_Watts | I'll fix it after lunch. | 14:22.39 |
| tor5 | Robin_Watts: okay. | 14:22.50 |
| | or if you have a support contract, I'm sure we could rearrange priorities | 14:23.25 |
| oznt | the project is django based ... so I was thinking to parse the output of mudraw | 14:23.41 |
| | can i ask you guys a general question about software developing ? I am a junior in the field ... so I am wondering | 14:23.41 |
| | I will do ... | 14:24.22 |
| tor5 | oznt: ask away | 14:44.45 |
| oznt | tor5, I read the code of mupdf a couple of time and it's quite well documented (at least the functions header), than I look into other C (FOSS) project and very often there are no comments at all | 14:46.14 |
| kens | Some developers dislike comments 'because they are never current with the code' | 14:46.57 |
| oznt | am I expecting too much? I know the code should be self explaining ... so where is the border ? | 14:47.04 |
| | because it takes effort to write them? | 14:47.24 |
| kens | because people change the code, but not the comments | 14:47.34 |
| | This is 'personal preference' stuff. | 14:47.58 |
| tor7 | oznt: if you'd looked two years ago, you would have found much fewer comments in the mupdf code. we made a big drive to document it all once we reached a certain level of maturity. | 14:48.03 |
| | code maturity, that is | 14:48.09 |
| kens | Put 3 software engineers in a room and you'll get at least 4 opinions | 14:48.16 |
| oznt | well, i have to admit, I your code is good | 14:49.57 |
| kens | You haven't looed at Gh... | 14:50.40 |
| | GS | 14:50.48 |
| chrisl | Abandon hope, all ye who enter GS..... | 14:51.52 |
| Robin_Watts | tor7: How is the jbig2dec handled? | 16:00.18 |
| tor7 | you mean the git? | 16:00.30 |
| Robin_Watts | Is that similarly extracted from gs to the standalone repo? | 16:00.32 |
| tor7 | yeah, exact same way as openjpeg is now. | 16:00.44 |
| Robin_Watts | bug 694362 - apparently there is stuff we haven't pulled in. | 16:00.59 |
| tor7 | crontab job extracts the subdirectories, and pulls them into the gits | 16:01.06 |
| Robin_Watts | probably just requires us to update the submodule. | 16:01.14 |
| | I will look at that. | 16:01.17 |
| | Linux builds are failing :( | 16:01.37 |
| tor7 | yeah, we have to manually bump the mupdf/thirdparty/jbig2dec submodule | 16:01.40 |
| | Robin_Watts: on origin/master? | 16:01.59 |
| Robin_Watts | I suspect your commit and my openjpeg commit didn't agree with one another. I will fix. | 16:02.02 |
| | on golden/master. | 16:02.05 |
| | I will fix and force-push. | 16:02.12 |
| tor7 | yeah, something's bust | 16:02.36 |
| | yeah, the merge of Makethird is probably off with the OPENJPEG_CFLAGS | 16:03.15 |
| Robin_Watts | I see it. | 16:04.22 |
| | kens, chrisl: Were lexmark t640's among the b0rked printers that had problems with postscript? | 16:16.06 |
| mace | kens/chrisl: robin kindly suggested i ask this here, on the off chance you guys can suggest a fix.. i'm trying to get a lexmark t640 set up with cups, but i'm getting wierd behaviour.. first test page prints ok, second test page prints ok too but is preceeded by a page with "ageLevel: 2" at the top. this makes me suspect that the first job isn't being fully processed, and robin mentioned that there was a hack/workaround needed | 16:16.15 |
| | Robin_Watts is more succinct ;) | 16:16.23 |
| chrisl | tbh, I can't remember..... | 16:17.12 |
| Robin_Watts | mace: Ah, well, I didn't know all the details :) | 16:17.17 |
| kens | We've fixed a few problems, working around broken PostScript implementations,but that doesn't sound like any bug we've seen | 16:17.24 |
| Robin_Watts | ISTR that the problems we had reported were more like indefinite hangs, or much worse crashes than that. | 16:17.40 |
| mace | right | 16:17.47 |
| kens | what happens if you run the PostScript file through GS on its own ? | 16:17.49 |
| chrisl | Robin_Watts: No, we've had strange errors, too | 16:17.58 |
| Robin_Watts | tkamppeter_ is the CUPS guru. | 16:18.01 |
| | chrisl: OK, I'll shut up here then :) | 16:18.08 |
| mace | kens: i've not tried, will try and figure out if i can do that ;) | 16:18.21 |
| kens | mace can you send us the intermediate PostScript file ? | 16:18.32 |
| chrisl | Robin_Watts: well, you right in that *most* have been freezes or crashes | 16:18.41 |
| kens | OOh, and are you sending PostScript or PDF to Cups ? | 16:18.46 |
| | Also, where/how are you generating the test page ? | 16:19.14 |
| mace | i've tried two methods, hitting the "test page" button in cups, and running the lp command to print a text file | 16:19.52 |
| kens | Hmm, I have no idea what cups does if you ask it to printa test page :-) | 16:20.20 |
| mace | nor i :) | 16:20.27 |
| chrisl | I suspect it has a PDF stashed somewhere, and tosses that to the printer - I'll have a quick look | 16:21.02 |
| kens | Probably the first thing to do is capture the file that gets sent to your printer | 16:21.07 |
| | THat will (shoudl be) PostScript, presumably generated by GS. I should also ask which version of GS you are using ? | 16:21.37 |
| | That 'PageLevel' kind of looks like a DSC comment | 16:22.07 |
| mace | ageLevel, not PageLevel.. quick look at the postscript spec makes me suspect LanguageLevel, but I know nothing :) | 16:22.51 |
| | found some pdf's, sec | 16:23.13 |
| chrisl | The cups test page does seem to be a stashed PDF file | 16:23.40 |
| | mace: this should give you the steps to capture the PS going to the printer: https://wiki.ubuntu.com/DebuggingPrintingProblems#Getting_the_data_which_would_go_to_the_printer | 16:24.46 |
| mace | will do | 16:25.31 |
| kens | That 'ageLevel' could be the end of a 'LanguageLevel' DSC comment | 16:25.32 |
| | A quick check and we do emit a %%LanguageLevel: 2 comment in the ps2write output | 16:26.42 |
| chrisl | Hrm, the "printer settings" on this computer seems to have gone AWOL..... | 16:28.16 |
| mace | just trying the 'raw' printing thing via lp to see if it's reproducable with that | 16:28.35 |
| | ok that seems to cause other problems | 16:29.29 |
| chrisl | mace: what are you sending "raw" to the printer? | 16:30.39 |
| mace | text file | 16:31.04 |
| kens | You cna't send a text file to a PostScript printer | 16:31.17 |
| | It will give you all kinds of errors | 16:31.25 |
| mace | well its going via cups still, but without whatever filters -oraw disables | 16:31.50 |
| | i digress anyhow, will get the intermediate postscript for you | 16:32.08 |
| kens | I would imagine -oraw means 'send this directo to the printer' | 16:32.08 |
| chrisl | mace: -oraw disables *all* the filter - it sends exactly the input to the printer | 16:32.21 |
| mace | right, that's what it looked like | 16:33.03 |
| chrisl | It *might* "work" if the printer also supports PCL5 | 16:33.25 |
| kens | Yes true | 16:33.35 |
| | It says it handles PCL6 | 16:34.30 |
| | Also PDF 1.5 and PS 3 | 16:34.42 |
| | as well as something called PersonalPrinter Data Stream | 16:34.57 |
| chrisl | Yeh, it depends whether it needs PJL and any other preamble before the "PCL" | 16:35.06 |
| kens | I would guess it probably does, given the range of possible inputs | 16:35.28 |
| chrisl | A lot of printers default to PCL5 if they can't identify the input by other means | 16:36.10 |
| kens | OK PPDS ios some kind of IBM specific thingy | 16:36.16 |
| | Its a mere 32 years old | 16:36.38 |
| mace | ok i've gotten this output, what would you like me to do with it? :) | 16:37.55 |
| kens | Put it somewhere that we can look at it | 16:38.16 |
| mace | good answer | 16:39.26 |
| chrisl | mace: if you haven't got somewhere convenient, you can mail it to me, and I'll put it somewhere | 16:39.54 |
| mace | http://dsis.it/printout | 16:40.52 |
| chrisl | Forbidden You don't have permission to access /printout on this server. | 16:41.15 |
| mace | fixed, sorry | 16:41.43 |
| kens | That's an old version of GS, 9.05 | 16:41.56 |
| | You could try a newer one | 16:42.02 |
| mace | it's the one in current debian stable | 16:42.14 |
| kens | Yep, but its still old | 16:42.24 |
| mace | i'll have a look | 16:42.28 |
| | hm, 9.05 is the latest even in debian unstable | 16:43.09 |
| chrisl | 9.10 was release back in August.... | 16:43.27 |
| mace | neat, last unstable release was july, damn debian developers :) | 16:43.58 |
| kens | Well, the one page file processes correctly on 3 different interpreters for me | 16:44.06 |
| mace | out of interest, if i send that file to the printer raw, will that work as expected? | 16:44.36 |
| kens | Try it, it should | 16:44.43 |
| | You say you're getting *2* pages out of this ? | 16:45.00 |
| chrisl | But it should work anyway | 16:45.09 |
| kens | mace, how many pages does your printer print ? | 16:45.31 |
| mace | ok sending that raw to the usb lp port gives 1 page as it should | 16:46.13 |
| | i was getting 1 page the first time around, then 2 pages the second time around | 16:46.29 |
| | via cups etc | 16:46.34 |
| kens | That's strange. | 16:46.47 |
| | WHat was on the two pages when you got 2 pages out ? | 16:46.56 |
| mace | first page had just "ageLevel: 2" | 16:47.25 |
| | second page looked correct | 16:47.30 |
| kens | I'm at a loss to explain that..... | 16:47.49 |
| chrisl | That smacks of a comms error rather than a GS type problem - although, who knows.... | 16:48.28 |
| | mace: does the printer have a network connection? | 16:48.53 |
| mace | ah, i'm making some progress now | 16:48.58 |
| | chrisl: sadly not | 16:49.01 |
| | ok so catting that PS to the lp port multiple times works fine | 16:49.16 |
| | printing that PS via lp -oraw does the 2 page thing | 16:49.42 |
| | so i guess that means it's something to do with cups itself, not with ghostscript et al? | 16:50.07 |
| tkamppeter | mace, also follow the instrctions on https://wiki.ubuntu.com/DebuggingPrintingProblems#PostScript_.28PDF.29_printer_chokes_on_the_PostScript_.28PDF.29_coming_from_Ubuntu | 16:50.07 |
| chrisl | mace: hmm, yes, that rather points the finger at cups rather than GS | 16:50.21 |
| kens | Well, either the output is different the second time cups sends it, or its something odd with comms | 16:50.25 |
| mace | kens: i tried printing the GS file whilst redirecting to file.. the new output was identical to the original | 16:51.09 |
| Robin_Watts | mace, so it sounds like gs is behaving as it should. | 16:52.08 |
| | and it's something in cups that's bu{ff,gg}ering the file as it sends it to the printer. | 16:52.27 |
| mace | yeah, as expected.. i trust you folks more than i trust cups :) | 16:52.30 |
| kens | Its hard tro see how it could be GS, or the PostScript produced by GS, given the tests you;ve run | 16:52.47 |
| chrisl | I'm surprised there is no PJL in the data being sent to the printer, given the multiple PDLs the printer supports | 16:53.04 |
| Robin_Watts | chrisl: Well, maybe cups is introducing such extra stuff, and that's what's going wrong? | 16:53.28 |
| chrisl | Robin_Watts: this *should* the exact data going to the printer, cups additions and all | 16:53.52 |
| mace | tkamppeter: that step 10 works fine | 16:53.57 |
| | tkamppeter: lp -opsdebug on that GS file reliably prints 1 page each time | 16:54.22 |
| chrisl | Indeed, the device specific setup stuff is in the PS (which comes from cups not GS) | 16:54.40 |
| kens | chrisl sending the PS file directly seemed to work. Maybe the pritners processor can detect PS files | 16:54.53 |
| Robin_Watts | tor7: How often does the gs -> jbig2dec thing run? | 16:55.02 |
| chrisl | kens: yeh, it must be - I'm just surprised. | 16:55.17 |
| | mace: it could be a USB issue, too. Maybe adding/removing a hub, different port etc? | 16:55.57 |
| tor7 | Robin_Watts: daily | 16:56.03 |
| | but I can trigger a manual run should you want it | 16:56.14 |
| | (it's a pretty heavy job, I don't want to waste too many CPU hours on it) | 16:56.41 |
| Robin_Watts | tor7: could you trigger one please? | 16:57.11 |
| tkamppeter | mace, chrisl, what we can do is adding a quirk exception to the pdftops filter in cups-filters to send uncompressed PS for Lexmark printers. chrisl, if there is any better solution, please tell me. | 16:57.32 |
| tor7 | Robin_Watts: it's running... | 16:57.33 |
| Robin_Watts | Thanks. | 16:57.45 |
| tor7 | check the gits in a couple of minutes | 16:57.47 |
| mace | chrisl: that might be it actually.. i'm seeing usb disconnects in the kernel log | 16:57.50 |
| Robin_Watts | Shelly reports that the problem should be solved with a couple of patches - but all those patches are now included and I think I'm still seeing the problem. | 16:58.30 |
| chrisl | tkamppeter: but sending the compressed PS directly to the port worked correctly, it seems, so that's not likely to be the root of the problem | 16:58.52 |
| kens | I would have to agree, I don't see how it can be the PostScript itself | 16:59.24 |
| chrisl | tkamppeter: I'm wondering if the two methods (directly to the port, and via cups) are driving the USB subsystem in subtly different ways - hence one way seeing the problem, the other not | 17:00.05 |
| | But that's well beyond my area of expertise...... | 17:00.22 |
| | Well, if the worst comes to the worst, I can get a refurbed T640 for 75UKP....... | 17:02.33 |
| kens | At the oment I have to say it doesn't seem like its us. | 17:03.58 |
| chrisl | No, I don't think so. But if sending uncompressed data gets around mace's (guessing!) USB issue, then he could stick with that | 17:05.39 |
| mace | nods | 17:05.53 |
| | just trying various cable swaps | 17:05.59 |
| | still getting usb disconnects | 17:06.31 |
| chrisl | mace: another thought would another computer if you have one handy to test | 17:06.45 |
| mace | yeah i might try that next | 17:06.54 |
| chrisl | 'Course, it might be a problem with the printer's USB, in which case...... | 17:07.56 |
| mace | open window. discard ;) | 17:08.08 |
| Robin_Watts | chrisl: The fact that the printer can be sent data safely just by raw catting would seem to suggest that the printer itself is OK. | 17:08.39 |
| | Ah, the last few changes DO seem to have fixed it. | 17:09.08 |
| | (the MuPDF jbig2 thing, not mace's thing, obviously) | 17:09.27 |
| chrisl | Robin_Watts: like I said, unless the two "routes" for data are driving the USB in slightly different ways - don't know enough about USB to know if that makes any sense | 17:10.07 |
| | mace: erm, did you notice the next section in the web page a post above? "USB printer does not print or prints garbage" | 17:11.56 |
| Robin_Watts | tor7: I've got another mupdf commit ready. Its just another submodule update for jbig2dec. Seems to solve the problems. | 17:12.40 |
| mace | chrisl: i did, i tried the first workaround, but not the second, will try that in a minute. i've just tried connecting to a different computer and get the same usb disconnects when cups prints | 17:13.38 |
| chrisl | mace: okay, just checking - I would hate for us to miss something "obvious" | 17:14.58 |
| mace | either way, i think gs is innocent | 17:15.25 |
| | ah, i think they are not usb disconnects, but usblp being detached | 17:16.09 |
| | so, the second usb hack there seems to be working ok from linux | 17:20.27 |
| chrisl | Fingers crossed - I like happy endings :-) | 17:21.52 |
| mace | :) | 17:22.08 |
| | now to try from xp | 17:22.12 |
| chrisl | Oh, well we definitely can't take the blame there! | 17:22.31 |
| mace | nope, but you have been most helpful thanks :) | 17:22.44 |
| chrisl | NP. I'm just glad it wasn't us | 17:23.16 |
| Robin_Watts | oh, it's Thanksgiving today. | 17:24.34 |
| kens | Yes | 17:24.45 |
| chrisl | Oh, 'course it is.... | 17:25.07 |
| mace | xp works too | 17:40.04 |
| | what a mission :| | 17:40.09 |
| chrisl | mace: if the problem had turned out to the GS output, *that* would have been a real mission! It's a nightmare to debug problems like that, and much, much worse doing it by remote (lack of) control | 17:41.19 |
| mace | nods | 17:44.25 |
| | well thanks again :) | 17:44.33 |
| mace | moves onto the next impossible task | 17:44.42 |
| chrisl | Have fun ;-) | 17:44.52 |
| tkamppeter | chrisl, mace, perhaps it is a USB oproblem, directly to the port usually goes through the usblp kernel module, CUPS uses libusb. | 18:02.53 |
| | mace, can you follow the instructions of https://wiki.ubuntu.com/DebuggingPrintingProblems#USB_printer_does_not_print_or_prints_garbage | 18:03.29 |
| mace | tkamppeter: which aspect? | 18:07.17 |
| | tkamppeter: i applied step 3 and it seems to have fixed the problem | 18:07.59 |
| Robin_Watts | tor7: Yet another jbig2dec update on robin/master | 18:45.28 |
| | mvrhel_laptop: Happy thanksgiving. | 18:45.42 |
| mvrhel_laptop | morning Robin_Watts thanks | 18:46.05 |
| tkamppeter | mace, thank you, this means that CUPS' USB backend will get a quirk rule to solve your problem. | 19:09.51 |
| | mace, what is your exact printer model and what is the output of "lsusb" (USB vendor and product ID of your printer)? | 19:10.40 |
| | mace, and which version of CUPS are you using? | 19:11.42 |
| Robin_Watts | hi tor7. | 20:05.25 |
| | 1 more jbig2 import thing on robin/master | 20:05.45 |
| tor7 | Robin_Watts: okay, jbig2dec changes LGTM | 21:58.31 |
| | Forward 1 day (to 2013/11/29)>>> | |
IRC Logs