slacko-5.6 Final - plus testing next

[rcrsn51]

Jim, is that a WiFi printer? Cups won't find a WiFi printer with firewall on.

HP Photosmart C5280 All-in-One is USB.

I am not seeing this problem at all with any USB printer.

1. What install method are you using?

2. Go to the folder /dev/usb and check the properties of lp0.
What are its permissions?

3. Type the command: /usr/lib/cups/backend/usb
What do you get back?

4. Do you have the same problem on a clean "pfix=ram" boot off the CD?

[Jim1911]

Jim, is that a WiFi printer? Cups won't find a WiFi printer with firewall on.

HP Photosmart C5280 All-in-One is USB.

I am not seeing this problem at all with any USB printer.

1. What install method are you using?

2. Go to the folder /dev/usb and check the properties of lp0.
What are its permissions?

3. Type the command: /usr/lib/cups/backend/usb
What do you get back?

4. Do you have the same problem on a clean "pfix=ram" boot off the CD?

1. Installation method:
a. Installed your hplip_print-3.12.4 and hplip_scan-3.12.4 pets (Note: USB scan function works)
b. Used radky's pupcontrol to bring up the CUPS 1.4.8 installation program to install the printer.

2. See image-1

3. # /usr/lib/cups/backend/usb
direct usb://HP/Photosmart%20C5200%20series?serial=MY79BDC09Z04XQ "HP Photosmart C5200 series" "HP Photosmart C5200 series" "MFG:HP;MDL:Photosmart C5200 series;CMD:MLC,PCL,PML,DW-PCL,DESKJET,DYN;1284.4DL:4d,4e,1;CLS:PRINTER;DES:Q8330A;SN:MY79BDC09Z04XQ;S:038008C484001021002c1800062c288003f;J: ;Z:0102,0503b929014dc9,0600,0c100,0e00,0f00;" ""

4. I use a clean frugal installation on a hard drive, I was getting too many CD coasters.

Thank you for keeping your specialized pets up-to-date and continuing to share your printer knowlege with us. Your pets have worked great on numerous pup installations, without which it would have been a chore to get the printer to work.

Thanks again,
Jim

[rcrsn51]

@Jim1911:
I have tested this on several computers with several printers. I ran:

PupControl > Hardware > Printer > Install > "No" > Administration tab > Add Printer. Or look under Find New Printers.

In each case, my USB printer was listed under Local Printer.

I don't know why your experience is different. Remember that some HP all-in-ones must be rebooted after a scan session before they become visible as printers.

Try this:

Code: Select all

chmod 666 /dev/usb/lp0

Then bring up CUPS.

I use a clean frugal installation on a hard drive

But you have also installed some other content. You need to boot your frugal install using "pfix=ram".

[Jim1911]

@Jim1911:
I have tested this on several computers with several printers. I ran:

PupControl > Hardware > Printer > Install > "No" > Administration tab > Add Printer. Or look under Find New Printers.

In each case, my USB printer was listed under Local Printer.

I don't know why your experience is different. Remember that some HP all-in-ones must be rebooted after a scan session before they become visible as printers.

Try this:

Code: Select all

chmod 666 /dev/usb/lp0

Then bring up CUPS.

I use a clean frugal installation on a hard drive

But you have also installed some other content. You need to boot your frugal install using "pfix=ram".

I use the same procedure. In this try I did boot into ram, installed your pets, then called up Add Printer (results below). Then I closed CUPS, executed above code, and called up CUPS again, no change, the printer still doesn't show up. It's also not accessed by find New Printers. I've also shut down and restarted the printer several times. Xsane image scanner still has access.

[rcrsn51]

One of the things that has changed in this Slacko is USB detection through udev. I can only assume that your problem is somehow connected.

[Jim1911]

rcrsn51

You may have hit on the problem because my printer works well on 5.5 . More food for thought for Mick.

Thanks,
Jim

[SFR]

(...continuation of this...)
I'm still trying to investigate what's going on in /usr/local/pup_event/ and things become a tiny bit more clear...

When Gparted is starting formatting a partition (eg. sdb1, a pendrive), /usr/local/pup_event/frontend_change receives (from .../pup_event/pup_event_frondend_d, I suppose) this string:
rem:sdb1 add:sdb1
but does nothing with it.
Whereas insertion/removal of a usb drive produces:
add:sdb add:sdb1
or
rem:sdb rem:sdb1
and frontend_change properly creates/removes an icon on/from the pinboard.

@Mick: did you mention to Barry about this yet (or should I do it?)? Maybe it's an easy fix for him..?
Well, in the end it's not a major issue, but could be confusing for a user who formats a partition to another filesystem and ends up with useless icon on the pinborard (unless X will be restarted).
Thank you.
_____________

@Zigbert: could you please add support for f2fs to your pmount version (see original pmount)?
Would be nice.

Thanks &
Greetings!

[01micko]

I think the problem lies @line 398 in /etc/rc.d/rc.sysinit. You'll see,

Now, there is a bug in woof where the variable DEVTMPFSFLG gets set. In out situation, it should get set to 0. As you will see it's set to 2. That will make those entries irrelevant to us, but still should probably be fixed.

More later..

[sszindian]

rcrsn51 wrote:

As you know, there are two steps to installing a printer in CUPS.

1. Detecting the physical device through Add Printer / Find New Printers.
2. Providing the correct printer driver software.

Which one are we talking about?
-----------------------------------------------------------------------------------------------------------

I only ever use Find New Printers!

I can't understand why it doesn't find the HP 1050? I'm aware nothing has changed since 5.5?

I believe I've tried all the little 'tricks' you've taught me over several versions of puppy's!

>>>---Indian------>

[rcrsn51]

@jim1911 and sszindian:

Here is something to try:

1. Install the attached PET.

2. Run:
rmmod usblp
/etc/init.d/cups restart

3. Run CUPS.

[sszindian]

rcrsn51 wrote:

1. Install the attached PET.

2. Run:
rmmod usblp
/etc/init.d/cups restart

3. Run CUPS.
-------------------------------------------------
Yep... that patch was needed. the 1050 printer found, setup and printed a test-page in about a minute!

'As always, Thanks rcrsn51' and hope this helps mic, looks like he has a nice build going here!

>>>---Indian------>

[rcrsn51]

sszindian: Excellent. However, you will need to permanently blacklist the usblp module for this to keep working.

Could you try one more thing for me:

1. Do a clean install.
2. Run: chmod 700 /usr/lib/cups/backend/usb
3. Run CUPS.

@Mick: This patch comes from compiling CUPS without the --disable-libusb option. So CUPS detects USB printers through libusb (like SANE) instead of through the device node.

However, a few drivers need to work through /dev/usb/lp0, so this is not a 100% solution.

I would like to hear from other users with USB printers - not just HP - about how they work.

[01micko]

Thanks rcrsn51,

I did notice that the 'usb' file in your pet was -rwx------ (0700) and is 2/3 the size of the original.

I'm just wondering though if when I revert the error in woof that causes the kernel to take over creation of device nodes whether this will be a waste of effort.

I'll run the current slacko alpha on my file/print server (cannon mp490) and see what happens, also with my dev version.

[rcrsn51]

Even though my USB printers work, there is something else going on with udev. In other Puppies, /etc/udev/rules.d/55-cups.rule is responsible for giving the printer device nodes 666 permissions.

In this Slacko, it isn't working - /dev/usb/lp0 always has 660. But this is good enough to allow printing.

[01micko]

Even though my USB printers work, there is something else going on with udev. In other Puppies, /etc/udev/rules.d/55-cups.rule is responsible for giving the printer device nodes 666 permissions.

In this Slacko, it isn't working - /dev/usb/lp0 always has 660. But this is good enough to allow printing.

And the permissions is I believe the problem. SFR touched on it.

Now, the bug in woof is setting a new variable, DEVTMPFSFLG to 2 in the initrd and /etc/rc.d/rc.sysinit. It shouldn't becuase the condition is if udev >= 176. We are using 167. Our DEVTMPFSFLG var should be 0. However it is working anyway.

Here are the relevant lines from rc.sysinit (@line 395)

Code: Select all

DEVTMPFSFLG=2
#130618 devtmpfs-enabled kernel, initrd may have loaded it on /dev, if DEVTMPFSFLG<>0.
if [ $DEVTMPFSFLG -ne 0 ];then
 #device nodes created by kernel in initrd are all 'root' group. fix (before starting udevd)...
 chmod 660 /dev/* 2>/dev/null
 chmod 666 /dev/null
 chmod 666 /dev/zero
 chgrp tty /dev/[pt]ty*
 chgrp  /dev/console
 chgrp video /dev/fb0
 chgrp floppy /dev/fd[0-9]
 chgrp disk /dev/[hs]d[a-z]*
 chgrp disk /dev/fuse
 chgrp disk /dev/mmcblk*
 chgrp cdrom /dev/sr[0-9]
fi

Probably should be added to that:

Code: Select all

chmod 666 /dev/usb/* #rcrsn51
chmod -R 1777 /dev/shm #SFR
chmod 666 /dev/urandom

That *should* be superfluous in my next release because I'll fix the woof bug and we should then be back to the old method of udev creating device nodes.

EDIT: one other thing.

Pemasu probably hasn't got the variable in his rc.sysinit. That block of code will error with "unary operator expected" or something because the variable isn't in quotes. Therefore that block fails to run (which is a good thing for pemasu).

[Jim1911]

@jim1911 and sszindian:

Here is something to try:

1. Install the attached PET.

2. Run:
rmmod usblp
/etc/init.d/cups restart

3. Run CUPS.

Do you need me to try this in view of 01micko's comments? If so, it will have to be tomorrow.
Jim

[01micko]

Do you need me to try this in view of 01micko's comments? If so, it will have to be tomorrow.
Jim

Up to you Jim, but I just did a test build with 3.4.52 (patches for f2fs worked ) and the perms problems mentioned by SFR with using Firefox as Spot have disappeared after fixing the woof bug .

I'll post the info on to BK.

[zigbert]

@Zigbert: could you please add support for f2fs to your pmount version (see original pmount)?
Would be nice.

I have missed this one... When did it happened

If the support came after Wary 5.5, I can diff that pmount with the latest.
I am the one to blame for not keeping the original pmount that I have based my version of


Sigmund

[01micko]

Sigmund, I just plucked this one from the recent woof tree..(attached)


---------------------------------------

@Mick: did you mention to Barry about this yet (or should I do it?)? Maybe it's an easy fix for him..?
Well, in the end it's not a major issue, but could be confusing for a user who formats a partition to another filesystem and ends up with useless icon on the pinborard (unless X will be restarted).

Reported

[SFR]

chmod 660 /dev/* 2>/dev/null

It makes sense now.

Reported

Thanks. And yes, a blank CD/DVD icon indeed doesn't show up on the pinboard and I have vague memory that it used to do this before, somewhere in 5.3.3 or 5.4...

@Zigbert:
Hopefully, it's just a matter of a few additional bytes in:
VALIDPARTS="`echo "$PARTSINFO" | grep -E 'f2fs|vfat|msdos...
It seems to work fine with this, but I might be wrong...
EDIT: almost forgot again - Pburn creates a debug file in /root named "test" (looks familiar ).

Greetings!