Date: Mon, 14 Apr 2014 15:44:24 +0400 From: Boris Samorodov <bsam@passap.ru> To: "O. Hartmann" <ohartman@zedat.fu-berlin.de> Cc: Max Brazhnikov <makc@freebsd.org>, freebsd-questions@freebsd.org, FreeBSD Ports <freebsd-ports@freebsd.org> Subject: Re: print/cups: since update to 1.7.1: error : Send-Document client-error-document-format-not-supported Message-ID: <534BCA18.5030508@passap.ru> In-Reply-To: <20140414113632.46dc4bc1.ohartman@zedat.fu-berlin.de> References: <20140408154210.1aec26a8.ohartman@zedat.fu-berlin.de> <534476FE.4020109@passap.ru> <20140411130833.0d4cd2f2.ohartman@zedat.fu-berlin.de> <2456953.yOakfpjjYJ@lumiwa.farms.net> <20140412151604.0c90adc4.ohartman@zedat.fu-berlin.de> <534AC039.5080109@passap.ru> <20140414113632.46dc4bc1.ohartman@zedat.fu-berlin.de>
next in thread | previous in thread | raw e-mail | index | archive | help
14.04.2014 13:36, O. Hartmann пишет: > On Sun, 13 Apr 2014 20:50:01 +0400 > Boris Samorodov <bsam@passap.ru> wrote: > >> 12.04.2014 17:16, O. Hartmann пишет: >>> On Fri, 11 Apr 2014 08:19:45 -0400 >>> Ajtim <lumiwa@gmail.com> wrote: >>> >>>> On Friday 11 April 2014 13:08:33 O. Hartmann wrote: >>>>> On Wed, 09 Apr 2014 02:23:58 +0400 >>>>> >>>>> Boris Samorodov <bsam@passap.ru> wrote: >>>>>> 09.04.2014 00:25, O. Hartmann пишет: >>>>>>> On Tue, 08 Apr 2014 23:16:33 +0400 >>>>>>> >>>>>>> Boris Samorodov <bsam@passap.ru> wrote: >>>>>>>> 08.04.2014 17:42, O. Hartmann пишет: >>>>>>>>> Since the update of print/cups from 1.5.X to 1.7.1 >>>>>>>> >>>>>>>> How did you do it? >>>>>>> >>>>>>> As it is reported in /usr/ports/UPDATING. I delete first cups-image, >>>>>>> then did the update which reeled in all the new stuff automatically. >>>>>> >>>>>> OK, lets start from some obvious things. >>>>>> Did you restart cupsd while experimenting? >>>>> >>>>> No. >>>>> >>>>>> Give an output for: >>>>>> ----- >>>>>> % type lpr >>>>> >>>>> lpr is /usr/local/bin/lpr >>>>> pkg which /usr/local/bin/lpr >>>>> /usr/local/bin/lpr was installed by package cups-base-1.7.1 >>>>> >>>>>> % ls -l /usr/local/etc/cups >>>>> >>>>> -r--r--r-- 1 root wheel 2807 11 Apr 11:58 cups-browsed.conf >>>>> -rw-r----- 1 root cups 3197 8 Apr 17:15 cups-files.conf >>>>> -rw-r----- 1 root cups 3197 8 Apr 17:15 cups-files.conf.bak >>>>> -rw-r----- 1 root wheel 3137 11 Apr 12:02 cups-files.conf.sample >>>>> -r--r--r-- 1 root wheel 9521 11 Apr 11:50 cups-pdf.conf >>>>> -r--r--r-- 1 root wheel 9521 11 Apr 11:50 cups-pdf.conf.sample >>>>> -rw-r----- 1 root cups 3442 8 Apr 17:22 cupsd.conf >>>>> -rw-r----- 1 root cups 5098 8 Apr 17:20 cupsd.conf.O >>>>> -rw-r----- 1 root cups 3442 8 Apr 17:22 cupsd.conf.bak >>>>> -rw-r----- 1 root wheel 4492 11 Apr 12:02 cupsd.conf.default >>>>> -r--r--r-- 1 root wheel 4492 11 Apr 12:02 cupsd.conf.sample >>>>> drwxr-xr-x 2 root wheel 512 11 Apr 12:02 interfaces >>>>> -r--r--r-- 1 root wheel 1875 8 Apr 19:17 mime.convs >>>>> -r--r--r-- 1 root wheel 1874 11 Apr 12:02 mime.convs.sample >>>>> -r--r--r-- 1 root wheel 6456 8 Apr 19:18 mime.types >>>>> -r--r--r-- 1 root wheel 6455 11 Apr 12:02 mime.types.sample >>>>> drwxr-xr-x 2 root cups 512 11 Apr 12:02 ppd >>>>> -rw------- 1 root cups 4134 8 Apr 17:33 printers.conf >>>>> -rw------- 1 root cups 4134 8 Apr 14:58 printers.conf.O >>>>> -rw------- 1 root cups 4134 8 Apr 17:22 printers.conf.bak >>>>> -rw-r--r-- 1 root cups 946 11 Apr 11:53 pstoraster.convs >>>>> -r--r--r-- 1 root wheel 778 11 Apr 12:04 pstotiff.convs >>>>> -r--r--r-- 1 root wheel 2084 11 Apr 12:04 pstotiff.types >>>>> -r--r--r-- 1 root wheel 284 11 Apr 11:49 snmp.conf >>>>> -r--r--r-- 1 root wheel 284 11 Apr 12:02 snmp.conf.sample >>>>> drwx------ 2 root cups 512 11 Apr 12:02 ssl >>>>> >>>>>> % make -C /usr/ports/print/cups-client pretty-print-config >>>>> >>>>> -GNUTLS (whoops ... this is not the default, isn't it?) >>>>> >>>>> ==> corrected that with a new recompilation with rmconfig preceded. >>>>> >>>>>> % grep CUPS /etc/make.conf >>>>> >>>>> NULL (menas: no output) >>>>> >>>>>> % pkg info -x cups hp foo gut >>>>> >>>>> cups-base-1.7.1 >>>>> cups-client-1.7.1 >>>>> cups-filters-1.0.52 >>>>> cups-image-1.7.1 >>>>> cups-pdf-2.6.1_1 >>>>> cups-pstoraster-8.15.4_7 >>>>> cups-samba-6.0_7 >>>>> gutenprint-cups-5.2.8_1 >>>>> libgnomecups-0.2.3_5,1 >>>>> linux-f10-cups-libs-1.3.11_1 >>>>> foomatic-db-hpijs-1.4 >>>>> hplip-3.14.1 >>>>> kdevelop-php-1.6.0_1 >>>>> kdevelop-php-docs-1.6.0_1 >>>>> php5-5.4.27 >>>>> swhplugins-0.4.15_4 >>>>> foomatic-db-20140331 >>>>> foomatic-db-engine-4.0.11,2 >>>>> foomatic-db-hpijs-1.4 >>>>> foomatic-filters-4.0.17 >>>>> gimp-gutenprint-5.2.8 >>>>> gutenprint-base-5.2.8 >>>>> gutenprint-cups-5.2.8_1 >>>>> gutenprint-ijs-5.2.8 >>>>> >>>>>> ----- >>>>>> >>>>>> Stop cupsd, move temporary /var/log/cups/*_log to another location. >>>>>> Start cupsd and look for suspicious messages at /var/log/cups/* files. >>>>> >>>>> Nothing suspicious so far. With print/cups-filters installed, the whole cups >>>>> printing system is corrupted and doesn't print a single sheet of paper >>>>> (PDF/PS) normal es expected. >>>>> >>>>> >>>>> Prior to this task, I recompiled, as you suggested, first all cups ports and >>>>> afterwards hplip/foomatic. I also installed print/cups-filters before >>>>> recompiling hplip/foomatic. >>>> >>>> It works for me now (FreeBSD 10.0-RELEASE): >>>> >>>> I did deinstall cups* and hplip. Than installed cups and patched hplip. First >>>> I used as usual hp-business_inkjet_3000-hpijs-pcl3.ppd.gz and it didn't work >>>> It printed:%PDF-1.4 >>>> and jobs processing never stopped. Than I removed device and installed again >>>> with -3000-ps.ppd.gz and it works but cannot print test page. >>>> >>> >>> I tried alternatives, but it is with all (known to me) usefull drivers for the >>> specific printer the same result: empty pages, print job stuck in queue. I also tried >>> most recent hplip-3.14.4 but I doubt this is the reason. >>> >>> I can print PDF and PS, as reported, when deinstalling/removing print/cups-filters >>> using clients like xpdf, xdvi or printing directly via lpr -PPRINTER_NAME jobfile.ps. >>> This fails when print/cups-filter is installed. >> >> This is rather strange. But may be this is because you use hplip which >> I've never used before. > > I used hplip before with CUPS 1.5.X with the printer of mine, the HP OfficeJet Pro 8600 > N911a, as I mentined. The hplip driver suite privided the only prperly working driver for > my. But at this very moment, I'm not sure anymore whether I used the proper driver for > that kind of printer with HP Officejet Pro 8600 hpijs, 3.14.4, since I might have used > that kind of dribver WITHOUT the "hpijs" - but I can not confirm since I have not system > anymore with the old CUPS 1.5.X. > >> >>> Have you tried to deinstall by intention cups-filters and check whether the formerly >>> used driver works for you? >> >> Yes, I tested cups-1.7.0 for several months with no ill-effects (HP, >> OKI and Epson printers). > > I see no sign of ill effects either with all of the HP laser printers we use and those I > use in two diffreren departments. > >> >>> I also deinstalled everything related to hplip and cups (cups, cups-XXX, hplip, qpdf, >>> foomatic-XXX) and reinstalled first print/cups which reels in all cups-ports and then >>> print/hplip, which reels in qpdf and foomatic-filters. This should ensure the propper >>> chain of dependencies. But it fails. I then recompiled approx 260 ports by "portmaster >>> -f hplip" on my workstation and ~ 170 ports the same way on the server (gutenprint and >>> other gimp realted stuff aren't installed on the server). I had no success. >>> >>> I'd like to figure out what part of the duo cups and hplip is the culprit, but it >>> seems not to be easy if one is dependend on a working printing system. >> >> Since I didn't use (and test) print/hplip I gave it a try. I didn't >> manage to print via command line using hplip (i.e. just >> "lpr -P <hplip_printer> test_file" ended with almost blank line and >> the tast waiting at the printer's query. > > Exactly what I'm facing here. The print job prints out blank sheets, sometimes ending in > "?%" - nothing more. The job is stuck in the queue, never get ready and canceld out, I > have to delete it manually. > > > I did also another thing: I downloaded the hplip original raball from sourceforce. I > found two different folders withing the tree: one hpijs which contains the drivers I see > with hplip 3.14.1 or 3.14.4, but there is also a folder with drivers called "hpcups". > Creating a testqueue and taking the hpcups-driver instead of the hpijs seems to make > things working again, but I hadn't time enough to properly test it. I opened a PR > (ports/188509) and got response from the maintainer (which I added for convenience to the > CC list). > >> >> Further analysis showed that foomatic-rip-hplip doesn't like PDF files >> which new CUPS uses by default. It expects a PS file: >> ----- >> File: <STDIN> >> Reading PostScript input ... >> PID 29404 (/usr/local/libexec/cups/filter/pstopdf) exited with no errors. >> Job does not start with "%!", is it PostScript? >> ----- >> >> After switching from HP's foomatic-rip-hplip to native foomatic-rip >> all is fine: >> ----- >> PID 75535 (/usr/local/libexec/cups/filter/pstopdf) exited with no errors. >> Filetype: PDF >> ----- >> >> So, I fixed an error which I managed to reproduce. Please, try the >> attached file, recompile print/hplip and report back. The patch uses >> a native foomatic-rip where foomatic-rip-hplip was used. >> > > > I applied also your patch and I must say: you made my day! Thanks. Even the hpijs drivers > seem to work properly now. > > I can print via xpdf, dvips, gv, directly via lpr as well as via LibreOffice! > > It seems that your patch has fixed the problem. Great, glad to be helpful. May be the patch is not right and the problem should be resolved in another way. But anyway the problem seems to be localized and I'm sure Max will do the right thing. > A side observation is, that the printing queue takes much longer now to perform the > printout when commiting the printjob: lpr waits a while, stuck, the graphical clients get > a kind of stuck, not refreshing their graphical contents (xpdf, gv, except LibreOffice, > which prints immediately). > > Another side observation that the printer now is acting like a racoon: evertime it prints > out a single page or some pages, it starts a cleaning turn (consuming ink ...). -- WBR, Boris Samorodov (bsam) FreeBSD Committer, http://www.FreeBSD.org The Power To Serve
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?534BCA18.5030508>