HP-Laserjet2300dn printer, Postscript problem

[wedgetail]

The hp-laserjet2300dn, has duplex and network capability ( = *dn ).  I seem to have been doing a lot of printer work but more using 'normal' printing ie not using its in built Postscript capability.  I have met problems of various degrees and have settled on using CUPS a lot.

I seem to be steering in the direction of using KDE-mini, various builds and trying to keep minimal software on those. This means sometimes running without printing capability.  So happened that I started some experimentation with minimising number of drivers which led me to look at testing the Postscript on this printer. It is early days but I think I am up against an, for me, unexpected problem.

When CUPS is running with various add on drivers, and because of permutations one can make, one of my first stops is to ask CUPS how many drivers it knows about, by doing the following:

Code: [Select]

[root@localhost gert]# lpinfo -m >/home/gert/Documents/yymmdd-printersavailable.glk

Piping the numbers to a file, opening the file in Kwrite. Hit the F7 key and type "sort" in the field at the bottom. Then I search on the printer I am interested in. I pick out all instances of interest and then return to CUPS and install various queues for experiments.

For one of my KDE-mini systems I did this and found following two Postscript options.

3-distribution/HP/hp-laserjet_2300-ps.ppd.gz HP LaserJet 2300 Postscript (recommended)
3-distribution/HP/hp-laserjet_2300_series-ps.ppd.gz HP LaserJet 2300 Postscript (recommended)

The install is not difficult for me and one of my first actions is to use the Printers > Maintenance > Print Test Page which is normally a good indicator of success if this prints.

To my surprise the operations hangs with trying to connect to printer, extract from lpstat -t.

printer hp-2300dn-ps is idle.  enabled since Mon 27 Feb 2012 04:00:50 PM EST
        Connecting to printer...

I have turned on debug level "debug" ( I am not sure if there is not another one debug2 not got that far.)

[root@localhost gert]# cat /etc/cups/cupsd.conf |grep LogLevel
LogLevel debug
[root@localhost gert]#

The log is a bit overwhelming but reading it from the back it seems that the job went through ok, the details are complex for me and may return to look closer

I may misunderstand the following.  Using the Postscript in the printer my understanding is that less "filters" needed from application to the final queue file to be sent to the printer. That final file I thought was in Postscript format, but I may need to study up on this.  

With this in mind I stop cups to get hold of the file.

Code: [Select]

[root@localhost gert]# cupsdisable hp-2300dn-ps
Print Maintenance "test page" from CUPS Admin

Code: [Select]

[root@localhost gert]# lpstat -o
hp-2300dn-ps-92         root              1024   Mon 27 Feb 2012 04:30:48 PM EST

Code: [Select]

[root@localhost gert]# ls -l /var/spool/cups
total 12
-rw------- 1 root sys  640 Feb 27 16:31 c00092
-rw-r----- 1 root sys  305 Feb 27 16:30 d00092-001 < ------------- gert
drwxrwx--T 2 root sys 4096 Feb 27 16:00 tmp/

[root@localhost gert]# cp /var/spool/cups/d00092-001 /home/gert/120227-printout_test

[root@localhost gert]# chmod 777 /home/gert/120227-printout_test
[root@localhost gert]#


Content of self test page, /home/gert/120227-printout_test
-------
#CUPS-BANNER
Show printer-name printer-info printer-location printer-make-and-model printer-driver-name printer-driver-version paper-size imageable-area
Header Printer Test Page
Footer Printer Test Page
Notice This test page was produced by CUPS 1.4.6.
Image images/cups.png
Image images/color-wheel.png
-------

The file seems to be going to the printer, but nothing happens.  I say it seems to go to the printer because if I repeat the process with stopping CUPS and look at the queue file for a normal printer driver and asking for the Print test page the queue file will show the same content as above but when I then enable CUPS again the queue file will print on the printer. See below, and this is not the Postscript driver in use.

Code: [Select]

[root@localhost gert]# cupsdisable HP_Laserjet_2300dn
[root@localhost gert]# ls -l /var/spool/cups
total 44

-rw-r----- 1 root sys 16169 Feb 27 17:09 d00096-001
-rw-r----- 1 root sys   305 Feb 27 19:10 d00097-001
drwxrwx--T 2 root sys  4096 Feb 27 17:09 tmp/
[root@localhost gert]# cp /var/spool/cups/d00097-001 /home/gert/120227-printout_test_default_printer

[root@localhost gert]# chmod 777 /home/gert/120227-printout_test_default_printer
[root@localhost gert]# cupsenable  HP_Laserjet_2300dn
[root@localhost gert]#

Below is some more information.

[gert@localhost:~] $lpoptions -p hp-2300dn-ps
auth-info-required=none
copies=1
device-uri=lpd://192.168.1.6
finishings=3
job-hold-until=no-hold
job-priority=50
job-sheets=none,none
marker-change-time=0
number-up=1
printer-commands=AutoConfigure,Clean,PrintSelfTestPage
printer-info=KDE-mini
printer-is-accepting-jobs=true
printer-is-shared=false
printer-location=here
printer-make-and-model='HP LaserJet 2300 Postscript (recommended)'
printer-state=4
printer-state-change-time=1330322946
printer-state-reasons=connecting-to-device,com.apple.print.recoverable-warning
printer-type=10522708
printer-uri-supported=ipp://localhost:631/printers/hp-2300dn-ps
[gert@localhost:~] $

For the Postscript it seems there is just one step missing, something like ps2xxx filter perhaps.

Suggestions welcome.  

I am going to take a break and then go over my procedure again.  There is something about the text files I copied from the queue I am unhappy about, they are not Postscript, they contain the text information to be expected on the Printer Test Page and the text refer to two images, which I recognise as present on a 'proper' test page.  

[djohnston]

So happened that I started some experimentation with minimising number of drivers which led me to look at testing the Postscript on this printer.

One thing you haven't mentioned. Do you have task-printing-server installed?

[wedgetail]

djohnston
The short answer no.  The reason being that this brings in a foomatic ensemble of approx 20 MB. Foomatic is a massive system and it integrates well with CUPS but it is not needed for printing in general. Well that is my opinion, I do understand why the foomatic is considered needed for CUPS.

The foomatic does have drivers that may solve the problem, but I am deliberately avoiding them at the moment.

task-printing-server also bring in printer-tools which from what I understand is more for use with the inkjet printers.

In desperation I have just installed task-printing-server, but it does not seem to have made any difference so far.   

[AS]

wedgetail,

/usr/share/printer-testpages/testpage-a4.ps
/usr/share/printer-testpages/testpage.ps

the above are two Postscript test pages, A4 and Letter.

try to send it directly to the printer, without filtering:

Code: [Select]

lpr -P hp-2300dn-ps -o raw /usr/share/printer-testpages/testpage.ps
there will be no filtering here, because of "-o raw" option. 

AS

[wedgetail]

AS
Thanks, love the look of that.

For my own education I have taken the queue files to my workhorse kde-mini-OS where I have some 17 printer queues already and used the default 2300dn printer here which is not Postscript, when I print the copied queue content (from the trouble OS) I get the correct result.    

So the Print Test Page on the Postscript driver, result in the queue, copied to a file and then printed on the other system prints out the test page associated with the new "driver".

I also made a ksnapshot of the file (listing in a previous post) which appears to be a text file, the resulting *.png file I also printed out fine, the graphics a bit rough but ok.

My conclusion is that the content of the queue is correct but fails to transfer to the printer.  

I have to re-boot here to the other system but will come back into this post and continue with your result of your suggestion.  

Had some confusion here, the queue you used was actually on my workhorse, oh well it got a reboot. I feel I have some great embarrassment coming to me.  Still no good.

I cleared the joblist on the queue, before doing

[gert@localhost:~/Documents] $lpr -P hp-2300dn-ps -o raw /usr/share/printer-testpages/testpage.ps
[gert@localhost:~/Documents] $

From the CUPS webadmin, bit of imagination needed to fill in the graphics.

Code: [Select]

hp-2300dn-ps-102 testpage.ps gert 13k Unknown processing since Tue 28 Feb 2012 12:23:10 PM EST
Since the queue is hanging I have found that I can pick up the queue content quite easily:

[root@localhost KDE-mini_printing_Investigation_post_script]# ls -l /var/spool/cups
total 64
.............
-rw------- 1 root sys   775 Feb 28 12:23 c00102
-rw-r----- 1 root sys 12466 Feb 28 12:23 d00102-001
drwxrwx--T 2 root sys  4096 Feb 28 12:22 tmp/
[root@localhost KDE-mini_printing_Investigation_post_script]# cp  /var/spool/cups/d00102-001 120828-as_test1

[root@localhost KDE-mini_printing_Investigation_post_script]# chmod 777  120828-as_test1

[root@localhost KDE-mini_printing_Investigation_post_script]#

Added note: I have repeated the exercise with the A4 sheet size, just in case as I do have the A4 option set"

PageSize/Page Size: Letter LetterSmall Executive DoublePostcard Legal LegalSmall *A4 A4Small A5 ISOB5 B5 w612h935 w558h774 Env10 EnvMonarch EnvDL EnvC5 EnvISOB5 Custom.WIDTHxHEIGHT

I also cancelled the 'letter' size test just in case it was blocking for the A4 test. I think the Printer display panel normally makes a noise if the the sheet size does not match.

The file opens in Okular, very nice colour picture and it appears complete.  The printer of course is not a colour version but I would not expect this to be a problem, had gray scale conversion I expect.

Just downloaded the manual for the printer,  perhaps you can't just hit it with a Postscript driver and expect it to work.

Because of the size of the cups_error_log, what should I look for in here?

[wedgetail]

I am not sure if this has any importance, but I have opened the ppd file and copied out some of the beginning that I thought might be of help.

The file is one of those with Language options so it gets very long and boring apart from the front end.

*%=================================================
*%       Device Capabilities
*%=================================================
*ColorDevice: False
*DefaultColorSpace: Gray
*FileSystem: False

*LanguageLevel: "2"
*Throughput: "20"
*TTRasterizer: Type42
*?TTRasterizer: "
   save
      42 /FontType resourcestatus
      { pop pop (Type42)} {pop pop (None)} ifelse = flush
   restore
"
*End

*%=================================================
*%       Emulations and Protocols
*%=================================================
*Protocols: TBCP

*SuggestedJobTimeout:  "0"
*SuggestedWaitTimeout: "120"
*PrintPSErrors: True

[AS]

How is connected the printer ? LAN ? USB ?

[wedgetail]

LAN
device-uri=lpd://192.168.1.6

lpd has been my standard for a very long time, until I saw you did not use that so I am thinking about changing to "port 9100" but not just now

[AS]

LAN
device-uri=lpd://192.168.1.6

lpd has been my standard for a very long time, until I saw you did not use that so I am thinking about changing to "port 9100" but not just now

switch to jetdirect ... port 9100 ... just more reliable than lpd, (and one less parameter: the lpd queue name). 

[wedgetail]

Yes I will set up a 9100 queue as well. Jet Direct

From the manual

Fast printing performance, advanced font and graphics technologies, and advanced imaging capabilities are benefits of the PCL 6 printer language.

The PCL 6 technology also includes 45 scalable TrueType™ fonts and one bitmapped line printer font.

Other languages and fonts include HP PCL5e and PostScript® (PS) 3 emulation, which includes 35 built-in PS language fonts.

In a previous post in the ppd I notice it is Level 2 Postscript, but I find it hard to believe this would be a problem.
The printer automatically determines and switches to the appropriate language for the print job.

[AS]

Yes I will set up a 9100 queue as well. Jet Direct

From the manual

Fast printing performance, advanced font and graphics technologies, and advanced imaging capabilities are benefits of the PCL 6 printer language.

The PCL 6 technology also includes 45 scalable TrueType™ fonts and one bitmapped line printer font.

Other languages and fonts include HP PCL5e and PostScript® (PS) 3 emulation, which includes 35 built-in PS language fonts.

In a previous post in the ppd I notice it is Level 2 Postscript, but I find it hard to believe this would be a problem.
The printer automatically determines and switches to the appropriate language for the print job.

Print the "status sheet" from the printer control panel, both about printer config and about network config (if separate ...), there you will see exactly how the printer is configured, which emulations are enabled and which not ... BTW, a PS3 printer is compatible with PS2 and PS.

[wedgetail]

AS
Just the right time to mention "status form" for the printer.  Not sure at what stage I would have arrived here by myself because I was so sure in my mind that the IP was 192.168.1.6 and address this printer has worked on before, however quite some month ago the HP Laserjet 4V printer on 192.168.1.4 broke down and eventually I put the 2300 in its place, and I kept insisting that the address was *.6.

That is an embarrassment.  I have gone back and replaced the IP in the lpd statement and the queues start up immediately.

1..  Well I had clues. No real error in cups_error_log, to me it looked like everything was fine.
2..  Connecting to printer .......
3..  hp laser printers normally just works, almost with any driver (may miss out on some facilities, if the driver is not proper )

Ok, the socket 9100 I have not yet managed to get going.

[wedgetail]

On the socket 9100 I misunderstood "localhost",  because my OS systems run as "localhost" I thought 127.0.0.1:9100 would be the needed string.

Looking at the Status Page from the laserjet 2300dn I noticed it had

hostname:  NPI9FE6CE
IP Addres:  192.168.1.4  < ---- now it does  

Using the following:  socket://192.168.1.4:9100 will do the trick. For some reason using NPI9FE6CE did not work. Hmmmm....-- how is it now one finds all hosts on a LAN. I have even posted on that, grrrrr  

That was too much with that puzzle.  I installed "traceroute" and found that the hostname of the printer:

[root@localhost gert]# traceroute 192.168.1.4
traceroute to 192.168.1.4 (192.168.1.4), 30 hops max, 60 byte packets
 1  NPI9FE6CE.local (192.168.1.4)  1.564 ms  2.035 ms  2.479 ms
[root@localhost gert]#

Now, that was not indicated on the status page from the printer. I am using the following in CUPS and that works fine too. Could probably have been fixed with a "hosts" entry.

socket://NPI9FE6CE.local:9100

Added notes for myself
Reading the Manual I found that the printer has embedded webserver great number of details can be accessed by using Konqueror Web Browser, seems Opera fails to respond, this just brings up CUPS Web page?? The Webserver is a bit slow but easier to handle than the printer panel itself.

Further a fair bit of work trying to find out why this high end printer insists on delivering paper from tray 2 when I want it from tray 1, that is ridiculous.  Tray one is an easy single feed or few sheets feed and I like to be able to feed a single sheet through, for example to add a couple of extra sections to an existing page, having to pull out a bigger difficult tray to force selection to go to the last remaining source of paper

I even went and downloaded the HPLIP and fished out the ppd files for the printer, but this turned out to be same as already available via repo.  Did a bit comparison between the files.  I don't quit understand the programming inside.  Could not convince myself how the the decision on what tray was going to be used was made.

[gert@localhost Downloads]$ lpoptions -p  Laserjet2300dn-PostScript-1  -l
HPOption_Tray3/Tray 3: *False Tray3_500 Tray3_250
HPOption_Duplexer/Duplex Unit: *True False
HPPaperPolicy/Fit to Page: *PromptUser NearestSizeAdjust NearestSizeNoAdjust
InstalledMemory/Total Printer Memory: *32MB 48MB 64MB 96MB 160MB 256MB
HPStraightPaperPath/Use Straight Paper Path: True *False
Smoothing/Resolution Enhancement: *PrinterDefault On Off
PageSize/Page Size: Letter LetterSmall Executive DoublePostcard Legal LegalSmall *A4 A4Small A5 ISOB5 B5 w612h935 w558h774 Env10 EnvMonarch EnvDL EnvC5 EnvISOB5 Custom.WIDTHxHEIGHT
MediaType/Media Type: *None Plain Preprinted Letterhead Transparency Prepunched Labels Bond Recycled Color Rough Card_Stock Envelope
HPBookletFilter/Format Output As Booklet: True *False
HPBookletBackCover/Last Page Is Back Cover: True *False
HPBookletPageOrder/Page Order: *Normal Asian
HPBookletScaling/Scaling: *Proportional FitPage
HPBookletPageSize/Paper For Booklet: Letter LetterSmall Executive DoublePostcard Legal LegalSmall *A4 A4Small A5 ISOB5 B5 w612h935 w558h774
InputSlot/Paper Source: Auto *Tray1 Tray2 Tray3_500 Tray3_250
ManualFeed/Tray 1 (Manual): True *False
HPHalftone/Levels of Gray: *PrinterDefault Enhanced Standard
Resolution/Printer Resolution: 1200x1200dpi *1200dpi 600x600dpi
HPEconoMode/Economode: *PrinterDefault True False
Duplex/2-Sided Printing: *None DuplexNoTumble DuplexTumble
[gert@localhost Downloads]$

The red shows that I have default input slot for Tray 1, and there is paper in the tray.  This printer does not have a Tray 3 (false)

The command below allows me to move the default input slot, I tested with using Tray2 and the * moved. I am not sure what takes precedence cli or ppd.

[gert@localhost Downloads]$ lpoptions -p  Laserjet2300dn-PostScript-1 -o InputSlot=Tray1
[gert@localhost Downloads]$

The main test was printing a page as follows, because of the simplicity. CUPS had set the defaults beforehand

[root@localhost gert]#lp -d Laserjet2300dn-PostScript-1 /usr/share/printer-testpages/testpage.asc

In desperation I opened the same file in Kwrite, it is just a simple acii text file, selected the printer and checked that paper source should be Tray 1, but no Tray2 supplies the sheet.

Note:
Reading the manual and searching the net, there appears to be a difference between the settings that the CUPS driver can perform and the actual settings available when using the Printer Display Panel (or via the embedded webserver) with nothing to loose I have found the following settings to work for me:

Tray 1 Papersize = ANY
Tray 1 Media = ANY

Tray 2 Papersize = A4
Tray 2 Media = Plain

Tray 1 is "single sheet feed" sometimes open and when open will take precedence over Tray 2. Mostly closed and only opened to feed a few sheets manually for whatever reason.

Tray 2 is the 250 sheet feeder, normally used and is filled with Plain A4 size paper

The system now works as I wished it to, but will not work if Tray one is set to Size=A4, Media=Plain which I considered natural as I generally only use A4 plain paper.

Further using 'normal' ie. not postscript drivers seem to loose a number of capabilities.  I think I have confirmed that the recommended CUPS PostScript driver for this printer can not be fully configured from CUPS default settings or command line input. I would like to look a bit closer at that but now that I have the configuration I wished for I think I may leave it as is.

The two identical ppd PostScript Level 2 files are not particularly CUPS made, they are just ppd and there is a funny not that indicates it is made for Macintosh only?? A curiosity but HP recommends this for use in PCLOS HPLIP and I think this is fine.

I should not forget trying to run the tray selection via CUPS command line either, think I have seen something in lpoptions. From above this could be tricky and depend on if Printer Panel settings are already in your favour.

I should check out the Foomatic ppd Postscript version just to complete the PostScript section.  A quick look confirmed that the size of this file was not likely to support the full range of Default settings from CGI (CUPS Grahics Interface) management.  For example "Media" can not be set and doing it as command line interface there appear no easy way to see if it had any effect

Grrrrrrrr