samsung SCX-4521F connect to TC almost there

[beerstein]

Thank you for this samK:
I have not done this and forgot about it. You recommended it in post #23 in this other thread. I only remember that I was not able to rename and delete that foomatic-rip thing. So I stopped following this line.
 I will do this now and post results:
Try something like...

1)
Copy /tmp/tcloop/foomatic-rip/usr/local/bin/foomatic-rip to a location that allows the properties to be changed  e.g.  /opt/foomatic-rip/foomatic-rip

Delete the existing link at  /usr/local/lib/cups/filter/foomatic-rip

Create a replacement link at    /usr/local/lib/cups/filter/foomatic-rip   pointing to    /opt/foomatic-rip/foomatic-rip

Change the ownership of   /opt/foomatic-rip  recursively to be owned by root:root

IOn the samba part - I think foomatic-rip is the filter which needs smb.
Thank you so much for the concise explanations here samK
   

[SamK]

IOn the samba part - I think foomatic-rip is the filter which needs smb.

I am not familiar with either the physical printer or the PPD file you are using but would not expect this reliance on Samba to be part of the PPD.  What makes you suspect Samba is required by the PPD file?  Is it documented somewhere?

For the purposes of troubleshooting, I would lean towards using the most simple setup possible in order to obtain a local working configuration.  In this regard I am inclined towards the advice offered by, Juanito in reply #14.  Dropping all Samba elements may simplify the initial set-up stage.  Once the printer is working correctly as a local printer, it may then be easier to make it operate as a networked printer.
   

[beerstein]

Hi: I did the following:
------------------------------------------------------------------------------------------------------------------------------------------------------------
Copy /tmp/tcloop/foomatic-rip/usr/local/bin/foomatic-rip to a location that allows the properties to be changed  e.g.  /opt/foomatic-rip/foomatic-rip
Delete the existing link at  /usr/local/lib/cups/filter/foomatic-rip
Create a replacement link at    /usr/local/lib/cups/filter/foomatic-rip   pointing to    /opt/foomatic-rip/foomatic-rip
Change the ownership of   /opt/foomatic-rip  recursively to be owned by root:root
----------------------------------------------------------------------------------------------------------------------------------------------------------
although it took a while to learn how to link a file (symlink or hard link) I used sym link in ROX Filer.
When I try to print a test page - no error messages - job completed but no print out.
I tried to use the LPT1: and the USB port.

Now I do not know what to do any more. Maybe somebody has an idea?

[Juanito]

I can see that the existence of the cups1311 and cups extensions and their different deps and different "raison d'etre" is causing some grief.

I believe that one of the ideas behind the cups1311 extension (and thus foomatic-rip, which is not required for a lot of/most/all printer drivers using the cups extension) was that the user did not need to be root to perform some of the tasks - this probably explains why there seems to be a permissions issue.

As I recall, the foomatic-rip extension required libgs_ps (as opposed to libgs), is it loaded?

Things may work with libgs, but you'll probably need a symlink /usr/local/lib/libgs.so.8 -> libgs.so.9

[SamK]

@Juanito

I believe that one of the ideas behind the cups1311 extension (and thus foomatic-rip, which is not required for a lot of/most/all printer drivers using the cups extension) was that the user did not need to be root to perform some of the tasks - this probably explains why there seems to be a permissions issue.

Not meaning to be in any way contentious, I may simply have missed your point...

Only cups.tcz (CUPS 1.4.5) is used here and at least one printer requires Foomatic-Rip to be coaxed into life.  This produces the permissions symptoms reported by the OP.  Relocating to /opt (as previously described) resolves the matter completely and enables printing to be conducted as expected.



@beerstein

although it took a while to learn how to link a file (symlink or hard link) I used sym link in ROX Filer.

Not sure why this is so.  In http://forum.tinycorelinux.net/index.php/topic,9065.15.html:

• Reply #23 gave the steps that you quote in this current thread
• Reply #24 requests guidance on how to conduct the steps
• Reply #25 gives the commands to conduct the steps

When I try to print a test page - no error messages - job completed but no print out.
I tried to use the LPT1: and the USB port.

When printing via the parallel port the kernel module lp is required.  To confirm whether it is loaded:

Code: [Select]

lsmod
If it is not loaded have a look at:
http://forum.tinycorelinux.net/index.php/topic,9112.0.html

Is the printer known to work OK?  Does it produce a paper print using only its inbuilt self-test function?
   

[Juanito]

Not meaning to be in any way contentious, I may simply have missed your point...

Only cups.tcz (CUPS 1.4.5) is used here and at least one printer requires Foomatic-Rip to be coaxed into life.  This produces the permissions symptoms reported by the OP.  Relocating to /opt (as previously described) resolves the matter completely and enables printing to be conducted as expected.

That's good to know - looks like we need a foomatic-rip1311 and foomatic-rip extension - the latter having the correct permissions (i.e. those produced by make install).

..or we bin cups1311 in tc-4.x

[SamK]

That's good to know - looks like we need a foomatic-rip1311 and foomatic-rip extension - the latter having the correct permissions (i.e. those produced by make install).

Preventing the permissions issue will be a welcome step.

..or we bin cups1311 in tc-4.x

Doing this in isolation would still leave the permissions issue with the current cups.tcz in conjunction with foomatic-rip.
   

[beerstein]

Hi:
I eliminated almost all problems. - with your help - thank you

I fixed the ownership problem.
gs.tcz is installed on boot
the error log looks good exept the following:
============================================================

envp[28]="FINAL_CONTENT_TYPE=printer/samsung"
D [01/Sep/2011:17:23:51 +0000] [Job 4] Started filter /usr/local/lib/cups/filter/pdftops (PID 5893)
D [01/Sep/2011:17:23:51 +0000] [Job 4] Started filter /usr/local/lib/cups/filter/foomatic-rip (PID 5894)
D [01/Sep/2011:17:23:51 +0000] [Job 4] Started backend /usr/local/lib/cups/backend/parallel (PID 5895)
D [01/Sep/2011:17:23:51 +0000] [Job 4] STATE: +connecting-to-device
D [01/Sep/2011:17:23:51 +0000] [Job 4] Started filter gs (PID 5896)
D [01/Sep/2011:17:23:51 +0000] [Job 4] Started filter pstops (PID 5897)
D [01/Sep/2011:17:23:51 +0000] [Job 4] Set job-printer-state-message to "Unable to execute gs program: No such file or directory", current level=ERROR
D [01/Sep/2011:17:23:51 +0000] [Job 4] PID 5896 (gs) stopped with status 1!
D [01/Sep/2011:17:23:51 +0000] [Job 4] PID 5897 (pstops) stopped with status 1!
D [01/Sep/2011:17:23:51 +0000] [Job 4] Getting input from file
D [01/Sep/2011:17:23:51 +0000] [Job 4] foomatic-rip version 4.0.2.211 running...
D [01/Sep/2011:17:23:51 +0000] [Job 4] Parsing PPD file ...
D [01/Sep/2011:17:23:51 +0000] [Job 4] Added option Manualfeed
D [01/Sep/2011:17:23:51 +0000] [Job 4] Added option Resolution
D [01/Sep/2011:17:23:51 +0000] [Job 4] Added option Economode
======================================================
this is just part of the error log with strange behavior.
BTW: I need perl also on boot.

I have the newer cups running.libgs, fibgsf and libgs_ps are also on boot.

I think I am close. I will not give up.
Any ideas how I can continue?

[Juanito]

You need either libgs or libgs_ps, but not both.

I would suggest you try loading the ghostscript extension and, if that doesn't work, load the ghostscript_common extension as well

[beerstein]

Hi - thanks for the help.
I cleaned up the on-boot stuff and now have only the following cups related tcz´s left:

cups 1.4.5
ghostscript
foomatic-rip
libgs_ps

all modules related to lp are loaded. Now I have the ownership problem again. After reboot the "real" foomatic-rip file
in /opt/foomatic-rip/foomatic-rip shows root/staff instead of root/root ownership.
The same applies to the link (/usr/local/lib/cups/filter/foomatic-rip). Is there a way to keep the ownership alive after reboot?

BTW: I was able to print to a cups server on a remote machine - but only the print out of the test page worked. I was not able to print from a TC application on my machine. On that remote server a HP 5P LJ is installed and working in Suse Linux. I think this is related to an "allowed users" problem within the cups.conf on the cups print server.

[Juanito]

but only the print out of the test page worked.

Are you able to print a test page on the local machine?

[beerstein]

no - I was not able to print anything to the local Samsung printer.
The error messages when printing a test page showed the ownership problems
But when the ownership problem was fixed (manually) and I tried to print from Leafpad
I got a longer message output where everything looked ok exept a few lines where it said
something like "empty file printed"

here is a print out:
==========================================================================
Closing foomatic-rip.
D [28/Aug/2011:15:26:57 +0000] [Job 19] End of messages
D [28/Aug/2011:15:26:57 +0000] [Job 19] printer-state=3(idle)
D [28/Aug/2011:15:26:57 +0000] [Job 19] printer-state-message="Empty print file!"
D [28/Aug/2011:15:26:57 +0000] [Job 19] printer-state-reasons=none
E [28/Aug/2011:15:33:30 +0000] Filter "/usr/local/lib/cups/filter/rastertogutenprint.5.2" for printer "samsung2" not owned by root
E [28/Aug/2011:15:33:38 +0000] Filter "/usr/local/lib/cups/filter/rastertogutenprint.5.2" for printer "samsung2" not owned by root
E [28/Aug/2011:15:33:48 +0000] Unable to execute /usr/local/lib/cups/filter/rastertogutenprint.5.2: insecure file permissions (0100755)
E [28/Aug/2011:15:33:48 +0000] [Job 20] Unable to start filter "rastertogutenprint.5.2" - Operation not permitted.
E [28/Aug/2011:15:33:48 +0000] [Job 20] Stopping job because the scheduler could not execute a filter.
E [28/Aug/2011:15:35:33 +0000] Filter "/usr/local/lib/cups/filter/rastertogutenprint.5.2" for printer "samsung2" not owned by root
E [28/Aug/2011:15:35:39 +0000] Unable to execute /usr/local/lib/cups/filter/rastertogutenprint.5.2: insecure file permissions (0100755)
E [28/Aug/2011:15:35:39 +0000] [Job 21] Unable to start filter "rastertogutenprint.5.2" - Operation not permitted.
E [28/Aug/2011:15:35:39 +0000] [Job 21] Stopping job because the scheduler could not execute a filter.

=======================================================================

[Juanito]

I just sent you a pm with a foomatic-rip extension with the correct permissions - note that it has libgs as a dep. not libgs_ps

[beerstein]

Hi Juanito: thanks for the new foomatic-rip. I replaced the old one and the ownership problem
seems to be solved. I also loaded libgs instead of libgs_ps.
Now the only error message I am getting, when I try to print a test page:

E [06/Sep/2011:14:24:19 +0000] Unable to open listen socket for address ::1:631 - Address family not supported by protocol.
E [06/Sep/2011:14:25:20 +0000] [cups-driverd] Unable to open driver directory "/usr/local/lib/cups/driver": No such file or directory
E [06/Sep/2011:14:25:20 +0000] [cups-driverd] Unable to open driver directory "/usr/local/lib/cups/driver": No such file or directory

I did not know that the system expects a driver in addition to the PPD file. I always thought that the PPD file is the driver.. There is a driver on the CD which came with the printer, but this is an .rpm
with an install script which does not completely install on my TC.
I downloaded a linux driver for this model from the samsung website. This is a binary tgz file and I need to install that first. I have not done it before but I will try.

What do you think I should do next?

Thank you

[Rich]

Hi beerstein

::1:631

I think that notation is for IP6. I don't know if it doesn't support IP6, or if it just doesn't like that address.
At any rate, unless you are running billions of computers locally, you probably should be using IP4.