Bug#912768: hplip-data: hp-toolbox fsck

2018-11-12 Thread Brian Potkin
On Mon 12 Nov 2018 at 20:05:14 +0100, Cristian Ionescu-Idbohrn wrote:

> # avahi-browse -rt _scanner._tcp
> +   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)_scanner._tcp 
>local=   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)
> _scanner._tcplocal
>hostname = [NPI495F80.local]
>address = [192.168.x.y]
>port = [8289]
>txt = ["note=" "flatbed=T" "feeder=T" "button=T" "usb_MDL=HP ColorLaserJet 
> MFP M278-M281" "usb_MFG=HP" "mdl=ColorLaserJet MFP M278-M281" "mfg=HP" "ty=HP 
> ColorLaserJet MFP M278-M281" "adminurl=http://NPI495F80.local.; "txtvers=1" 
> "UUID=564e424e-4b43-4e38-5944-ace2d3495f80"]
> 
> # avahi-browse -rt _uscan._tcp
> +   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)_uscan._tcp   
>local=   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)
> _uscan._tcp  local
>hostname = [NPI495F80.local]
>address = [192.168.x.y]
>port = [8080]
>txt = ["note=" "duplex=F" "is=platen,adf" "cs=color,grayscale" 
> "pdl=application/pdf,image/jpeg" "uuid=564e424e-4b43-4e38-5944-ace2d3495f80" 
> "rs=eSCL" "representation=http://NPI495F80.local./ipp/images/printer.png; 
> "vers=2.63" "usb_MDL=HP ColorLaserJet MFP M278-M281" "usb_MFG=HP" 
> "mdl=ColorLaserJet MFP M278-M281" "mfg=HP" "ty=HP ColorLaserJet MFP 
> M278-M281" "adminurl=http://NPI495F80.local.; "txtvers=1"]

And now to show why these are important for you as well as for other
users.

Because the device has a _uscan._tcp service libsane-hpaio skips the
_scanner._tcp information. The important bits of the TXT record here are

  mdl=ColorLaserJet MFP M278-M28

and

  ty=HP ColorLaserJet MFP M278-M281

mdl= is effectively skipped and only ty= looked at by HPLIP.

The "HP" plus the space is removed and models.dat is tested for having
[colorLaserJet_mfp_m278-m281]. If it doesn't have it, what *should*
happen is that it should test for [hp_colorLaserJet_mfp_m278-m281]. On
3.17.10 this does not happen and the scanner is not found. This is
likely the source of you previous problem.

On 3.18.10 the search for [hp_colorLaserJet_mfp_m278-m281] succeeds. I
have no idea what causes this.

(One thing I have learnt from your response is that _uscan can have an
mdl= key).

Thanks again.

-- 
Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-12 Thread Brian Potkin
On Mon 12 Nov 2018 at 20:05:14 +0100, Cristian Ionescu-Idbohrn wrote:

> On Mon, 12 Nov 2018, Brian Potkin wrote:
> > On Mon 12 Nov 2018 at 19:28:05 +0100, Cristian Ionescu-Idbohrn wrote:
> > > On Mon, 12 Nov 2018, Brian Potkin wrote:
> > > > 
> > > > A small favour, Cristian. Please do
> > > > 
> > > > avahi-browse -rt _scanner._tcp
> > > > 
> > > > and
> > > > 
> > > > avahi-browse -rt _uscan._tcp
> > > > 
> > > > and let us know the outputs. avahi-browse is in the avahi-utils 
> > > > package.
> > > 
> > > Package avahi-utils is not installed on my systems.  I can print, 
> > > scan and possibly fax, without it.
> > 
> > You could install it and respond in a positive fashion. It would 
> > help in diagnosing problems which other users might face.
> 
> Alright, although I don't want bloat ;)

I understand that attitude and take the same view for my systems. But
when triaging bugs it is necessary for me to be more relaxed. The logs
help to revert what is installed as an extra, but much of the time I
make my own record

> Installing avahi-utils brings:
> 
>   avahi-daemon avahi-utils libavahi-core7 libdaemon0

just as you did.

> Uninstalled that immediatly.
> 
> # avahi-browse -rt _scanner._tcp
> +   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)_scanner._tcp 
>local=   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)
> _scanner._tcplocal
>hostname = [NPI495F80.local]
>address = [192.168.x.y]
>port = [8289]
>txt = ["note=" "flatbed=T" "feeder=T" "button=T" "usb_MDL=HP ColorLaserJet 
> MFP M278-M281" "usb_MFG=HP" "mdl=ColorLaserJet MFP M278-M281" "mfg=HP" "ty=HP 
> ColorLaserJet MFP M278-M281" "adminurl=http://NPI495F80.local.; "txtvers=1" 
> "UUID=564e424e-4b43-4e38-5944-ace2d3495f80"]
> 
> # avahi-browse -rt _uscan._tcp
> +   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)_uscan._tcp   
>local=   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)
> _uscan._tcp  local
>hostname = [NPI495F80.local]
>address = [192.168.x.y]
>port = [8080]
>txt = ["note=" "duplex=F" "is=platen,adf" "cs=color,grayscale" 
> "pdl=application/pdf,image/jpeg" "uuid=564e424e-4b43-4e38-5944-ace2d3495f80" 
> "rs=eSCL" "representation=http://NPI495F80.local./ipp/images/printer.png; 
> "vers=2.63" "usb_MDL=HP ColorLaserJet MFP M278-M281" "usb_MFG=HP" 
> "mdl=ColorLaserJet MFP M278-M281" "mfg=HP" "ty=HP ColorLaserJet MFP 
> M278-M281" "adminurl=http://NPI495F80.local.; "txtvers=1"]

Very helpful. Thank you for the trouble you went to.

-- 
Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-12 Thread Cristian Ionescu-Idbohrn


On Mon, 12 Nov 2018, Brian Potkin wrote:
> On Mon 12 Nov 2018 at 19:28:05 +0100, Cristian Ionescu-Idbohrn wrote:
> > On Mon, 12 Nov 2018, Brian Potkin wrote:
> > > 
> > > A small favour, Cristian. Please do
> > > 
> > > avahi-browse -rt _scanner._tcp
> > > 
> > > and
> > > 
> > > avahi-browse -rt _uscan._tcp
> > > 
> > > and let us know the outputs. avahi-browse is in the avahi-utils 
> > > package.
> > 
> > Package avahi-utils is not installed on my systems.  I can print, 
> > scan and possibly fax, without it.
> 
> You could install it and respond in a positive fashion. It would 
> help in diagnosing problems which other users might face.

Alright, although I don't want bloat ;)
Installing avahi-utils brings:

avahi-daemon avahi-utils libavahi-core7 libdaemon0

Uninstalled that immediatly.

# avahi-browse -rt _scanner._tcp
+   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)_scanner._tcp   
 local=   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)_scanner._tcp 
   local
   hostname = [NPI495F80.local]
   address = [192.168.x.y]
   port = [8289]
   txt = ["note=" "flatbed=T" "feeder=T" "button=T" "usb_MDL=HP ColorLaserJet 
MFP M278-M281" "usb_MFG=HP" "mdl=ColorLaserJet MFP M278-M281" "mfg=HP" "ty=HP 
ColorLaserJet MFP M278-M281" "adminurl=http://NPI495F80.local.; "txtvers=1" 
"UUID=564e424e-4b43-4e38-5944-ace2d3495f80"]

# avahi-browse -rt _uscan._tcp
+   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)_uscan._tcp 
 local=   eth0 IPv6 HP Color LaserJet MFP M281fdw (495F80)_uscan._tcp   
   local
   hostname = [NPI495F80.local]
   address = [192.168.x.y]
   port = [8080]
   txt = ["note=" "duplex=F" "is=platen,adf" "cs=color,grayscale" 
"pdl=application/pdf,image/jpeg" "uuid=564e424e-4b43-4e38-5944-ace2d3495f80" 
"rs=eSCL" "representation=http://NPI495F80.local./ipp/images/printer.png; 
"vers=2.63" "usb_MDL=HP ColorLaserJet MFP M278-M281" "usb_MFG=HP" 
"mdl=ColorLaserJet MFP M278-M281" "mfg=HP" "ty=HP ColorLaserJet MFP M278-M281" 
"adminurl=http://NPI495F80.local.; "txtvers=1"]


-- 
Cristian



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-12 Thread Brian Potkin
On Mon 12 Nov 2018 at 19:28:05 +0100, Cristian Ionescu-Idbohrn wrote:

> On Mon, 12 Nov 2018, Brian Potkin wrote:
> > 
> > A small favour, Cristian. Please do
> > 
> > avahi-browse -rt _scanner._tcp
> > 
> > and
> > 
> > avahi-browse -rt _uscan._tcp
> > 
> > and let us know the outputs. avahi-browse is in the avahi-utils 
> > package.
> 
> Package avahi-utils is not installed on my systems.  I can print, 
> scan and possibly fax, without it.

You could install it and respond in a positive fashion. It would help
in diagnosing problems which other users might face.

-- 
Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-12 Thread Cristian Ionescu-Idbohrn
On Mon, 12 Nov 2018, Brian Potkin wrote:
> On Thu 08 Nov 2018 at 20:43:04 +0100, Cristian Ionescu-Idbohrn wrote:
> > On Tue, 6 Nov 2018, Till Kamppeter wrote:
> > > 
> > > After applying this change "scanimage -L" discovers also my 
> > > scanner when connected to the network and I can scan from any 
> > > SANE-based application without needing a print queue using the 
> > > "hp" CUPS backend of HPLIP.
> > 
> > Till and Brian,
> > 
> > I now managed to test both printing and scanning with 
> > 3.18.10+dfsg0-2 and both work out-of-the-box, no hax needed :)
> > 
> > Thanks.
> 
> A small favour, Cristian. Please do
> 
> avahi-browse -rt _scanner._tcp
> 
> and
> 
> avahi-browse -rt _uscan._tcp
> 
> and let us know the outputs. avahi-browse is in the avahi-utils 
> package.

Package avahi-utils is not installed on my systems.  I can print, 
scan and possibly fax, without it.

The problem I'm currently experiencing is a hp-toolbox with 
removed/buggy functionality.  I'd be very much interested in 
NEWS/README information/motivation about it.


Cheers,

-- 
Cristian



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-12 Thread Brian Potkin
On Thu 08 Nov 2018 at 20:43:04 +0100, Cristian Ionescu-Idbohrn wrote:

> On Tue, 6 Nov 2018, Till Kamppeter wrote:
> > 
> > After applying this change "scanimage -L" discovers also my scanner when
> > connected to the network and I can scan from any SANE-based application
> > without needing a print queue using the "hp" CUPS backend of HPLIP.
> 
> Till and Brian,
> 
> I now managed to test both printing and scanning with 3.18.10+dfsg0-2 
> and both work out-of-the-box, no hax needed :)
> 
> Thanks.

A small favour, Cristian. Please do

avahi-browse -rt _scanner._tcp

and

avahi-browse -rt _uscan._tcp

and let us know the outputs. avahi-browse is in the avahi-utils package.

Thanks,

Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-09 Thread Brian Potkin
On Wed 07 Nov 2018 at 13:19:41 +, Brian Potkin wrote:

> I will draw attention to LP #1797501 even though it does not appear to
> be the same issue.
> 
>   https://bugs.launchpad.net/hplip/+bug/1797501

Till, you report

 > 1. Printer on the network
 >
 > scanimage -L
 >
 > does not find the scanner in my MF device, but

My guess is that this is indeed LP #1797501, which now has a testing
procedure added to it (the line with "mdl" has a missing "=").

Assuming your device advertises a _uscan._tcp service (an HP OfficeJet
Pro 8710 does) then its TXT record is used in preference to the one for
_scanner._tcp.

Regards,

Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-08 Thread Cristian Ionescu-Idbohrn
On Tue, 6 Nov 2018, Till Kamppeter wrote:
> 
> After applying this change "scanimage -L" discovers also my scanner when
> connected to the network and I can scan from any SANE-based application
> without needing a print queue using the "hp" CUPS backend of HPLIP.

Till and Brian,

I now managed to test both printing and scanning with 3.18.10+dfsg0-2 
and both work out-of-the-box, no hax needed :)

Thanks.

Though, I experience some oddoties with hp-toolbox.  On both stretch 
and buster, the "Actions" tab looks like this upon startup:

https://bugs.debian.org/cgi-bin/bugreport.cgi?att=1;bug=912768;filename=ok.jpg;msg=27

and _all_ the other tabs show reasonable contents.

On sid, the "Actions" tab is completly empty, on startup.

The "Status" tab shows "No status history available.", "Status 
information not available for this device.".

The "Supplies" tab shows "Supplies information not available for this 
device.".

The "Print Settings" tab shows reasonable contents.  "Printer 
Control", the same.

Going back to "Actions", I see this now:

https://bugs.debian.org/cgi-bin/bugreport.cgi?att=2;bug=912768;filename=fsck.jpg;msg=27

This looks like a bug to me.  I can workaround the missing "Scan" 
action (by manually starting the `simple-scan' application), but that 
won't make my wife happy :)


Cheers,

-- 
Cristian



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-07 Thread Brian Potkin
On Tue 06 Nov 2018 at 20:47:25 +0100, Till Kamppeter wrote:

> Concerning the scanning, I have done the following observation:
> 
> I have the HP OfficeJet Pro 8730.
> 
> I have removed all print queues using the "hp" CUPS backend (I print with a
> driverless queue). With an HPLIP-based print queue my scanner is always
> found.

I too can set up an HPLIP-based print queue for the 8730:

  lpadmin -p 8730 -v hp:/net/HP_OfficeJet_Pro_8730?ip=192.168.7.238 -E -m 
drv:///hpcups.drv/hp_officejet_pro_8730.ppd

'scanimage -L' gives

  device `hpaio:/net/HP_OfficeJet_Pro_8730?ip=192.168.7.238' is a 
Hewlett-Packard Officejet_Pro_8730 all-in-one

I would contend that no discovery over the network is taking place.
HPLIP knows the URI of the printer and simply substitutes "hpaio" for
"hp" to get the scanner URI. It is bound to work unless cupsd is
stopped.

> 1. Printer on the network
> 
> scanimage -L
> 
> does not find the scanner in my MF device, but

I think something similar happens in #879719.

> scanimage -d hpaio:/net/HP_OfficeJet_Pro_8730?ip=w.x.y.z > x.pnm
> 
> scans correctly.
> 
> hp-probe finds the URI:
> 
> hp:/net/HP_OfficeJet_Pro_8730?ip=w.x.y.z

By default hp-probe uses SLP for discovery. What is a little puzzling is
that the specs for the 8730 do not have it as a network protocol. A way
to check for the service is at

  https://wiki.debian.org/PrintQueuesCUPS#hplip.
 
> 2. Printer on USB
> 
> scanimage -L
> 
> finds the scanner with the following URI:
> 
> hpaio:/usb/HP_OfficeJet_Pro_8730?serial=xx
> 
> and it scans also when specifying this URI.
> 
> hp-probe finds the URI:
> 
> hp:/usb/HP_OfficeJet_Pro_8730?serial=CN783F60W1
> 
> 
> Now I did some debugging and found out that "scanimage -l" discovers the
> scanner with the URI (device on network, note the "hp_" missing in the model
> name, upper/lower case is ignored by HPLIP):

Is "-l" a typo?
 
> hpaio:/net/officejet_pro_8730?ip=w.x.y.z
> 
> This does not match the model name in
> /usr/share/hplip/data/models/models.dat which is "[hp_officejet_pro_8730]".
> 
> So I did the following change:
> 
> --
> --- io/hpmud/model.c~   2018-08-21 17:42:16.0 +0200
> +++ io/hpmud/model.c2018-11-06 17:14:04.302446688 +0100
> @@ -420,7 +420,10 @@
>   strncpy(section, rcbuf+1, sizeof(section)); /* found new section
> */
>   n = strlen(section);
>   section[n-2]=0; /* remove ']' and CR */
> - if (strcasecmp(model, section) == 0)
> + if (strcasecmp(model, section) == 0 ||
> +(section[0] == 'h' && section[1] == 'p' &&
> + section[2] == '_' &&
> + strcasecmp(model, section + 3) == 0))
>   {
>  /* Found model match. */
>  *bytes_read = ResolveAttributes(fp, attr, attrSize);
> --
> 
> This matches the model names both with and without "hp_" in the beginning.
> 
> Note that I did not yet upload an Ubuntu package of HPLIP with this patch as
> the archive is not yet opened for the disco (19.04) development cycle.
> 
> After applying this change "scanimage -L" discovers also my scanner when
> connected to the network and I can scan from any SANE-based application
> without needing a print queue using the "hp" CUPS backend of HPLIP.
> 
> Cristian, this could solve your scanner problem if the other fixes did not
> solve it yet, please try and report back.
> 
> In general this patch helps for using driverless printing on HP devices as
> long as they are connected by the network. If you do driverless printing via
> USB, with ippusbxd (IPP over USB) scanning does not work while ippusbxd is
> connected to the printer (it is really time that manufacturers start with
> driverless IPP scanning).
> 
> So for USB connection you will still need to print with HPLIP if you want to
> be able to scan.

I will draw attention to LP #1797501 even though it does not appear to
be the same issue.

  https://bugs.launchpad.net/hplip/+bug/1797501

-- 
Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-07 Thread Cristian Ionescu-Idbohrn
On Tue, 6 Nov 2018, Till Kamppeter wrote:
> 

[...]

> Now I did some debugging and found out that "scanimage -l" discovers 
> the scanner with the URI (device on network, note the "hp_" missing 
> in the model name, upper/lower case is ignored by HPLIP):
> 
> hpaio:/net/officejet_pro_8730?ip=w.x.y.z
> 
> This does not match the model name in /usr/share/hplip/data/models/models.dat
> which is "[hp_officejet_pro_8730]".
> 
> So I did the following change:
> 
> --
> --- io/hpmud/model.c~   2018-08-21 17:42:16.0 +0200
> +++ io/hpmud/model.c2018-11-06 17:14:04.302446688 +0100
> @@ -420,7 +420,10 @@
>   strncpy(section, rcbuf+1, sizeof(section)); /* found new section */
>   n = strlen(section);
>   section[n-2]=0; /* remove ']' and CR */
> - if (strcasecmp(model, section) == 0)
> + if (strcasecmp(model, section) == 0 ||
> +(section[0] == 'h' && section[1] == 'p' &&
> + section[2] == '_' &&
> + strcasecmp(model, section + 3) == 0))
>   {
>  /* Found model match. */
>  *bytes_read = ResolveAttributes(fp, attr, attrSize);
> --
> 
> This matches the model names both with and without "hp_" in the beginning.
> 
> Note that I did not yet upload an Ubuntu package of HPLIP with this 
> patch as the archive is not yet opened for the disco (19.04) 
> development cycle.
> 
> After applying this change "scanimage -L" discovers also my scanner 
> when connected to the network and I can scan from any SANE-based 
> application without needing a print queue using the "hp" CUPS 
> backend of HPLIP.
> 
> Cristian, this could solve your scanner problem if the other fixes 
> did not solve it yet, please try and report back.

Thanks Till.  I'll do some testing soon.


Cheers,

-- 
Cristian



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-06 Thread Till Kamppeter

Concerning the scanning, I have done the following observation:

I have the HP OfficeJet Pro 8730.

I have removed all print queues using the "hp" CUPS backend (I print 
with a driverless queue). With an HPLIP-based print queue my scanner is 
always found.


1. Printer on the network

scanimage -L

does not find the scanner in my MF device, but

scanimage -d hpaio:/net/HP_OfficeJet_Pro_8730?ip=w.x.y.z > x.pnm

scans correctly.

hp-probe finds the URI:

hp:/net/HP_OfficeJet_Pro_8730?ip=w.x.y.z

2. Printer on USB

scanimage -L

finds the scanner with the following URI:

hpaio:/usb/HP_OfficeJet_Pro_8730?serial=xx

and it scans also when specifying this URI.

hp-probe finds the URI:

hp:/usb/HP_OfficeJet_Pro_8730?serial=CN783F60W1


Now I did some debugging and found out that "scanimage -l" discovers the 
scanner with the URI (device on network, note the "hp_" missing in the 
model name, upper/lower case is ignored by HPLIP):


hpaio:/net/officejet_pro_8730?ip=w.x.y.z

This does not match the model name in 
/usr/share/hplip/data/models/models.dat which is "[hp_officejet_pro_8730]".


So I did the following change:

--
--- io/hpmud/model.c~   2018-08-21 17:42:16.0 +0200
+++ io/hpmud/model.c2018-11-06 17:14:04.302446688 +0100
@@ -420,7 +420,10 @@
  strncpy(section, rcbuf+1, sizeof(section)); /* found new 
section */

  n = strlen(section);
  section[n-2]=0; /* remove ']' and CR */
- if (strcasecmp(model, section) == 0)
+ if (strcasecmp(model, section) == 0 ||
+(section[0] == 'h' && section[1] == 'p' &&
+ section[2] == '_' &&
+ strcasecmp(model, section + 3) == 0))
  {
 /* Found model match. */
 *bytes_read = ResolveAttributes(fp, attr, attrSize);
--

This matches the model names both with and without "hp_" in the beginning.

Note that I did not yet upload an Ubuntu package of HPLIP with this 
patch as the archive is not yet opened for the disco (19.04) development 
cycle.


After applying this change "scanimage -L" discovers also my scanner when 
connected to the network and I can scan from any SANE-based application 
without needing a print queue using the "hp" CUPS backend of HPLIP.


Cristian, this could solve your scanner problem if the other fixes did 
not solve it yet, please try and report back.


In general this patch helps for using driverless printing on HP devices 
as long as they are connected by the network. If you do driverless 
printing via USB, with ippusbxd (IPP over USB) scanning does not work 
while ippusbxd is connected to the printer (it is really time that 
manufacturers start with driverless IPP scanning).


So for USB connection you will still need to print with HPLIP if you 
want to be able to scan.


   Till



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-06 Thread Cristian Ionescu-Idbohrn
On Mon, 5 Nov 2018, Brian Potkin wrote:
> 
> It would be useful to have what happens with
> 
>  scanimage -d "hpaio:/net/HP_ColorLaserJet_MFP_M278-M281?ip=192.168.x.y" > 
> out.pnm
> 
> for 3.18.10+dfsg0-1.
> 
> Does the scanner work? Do you get out.pnm?

Will do.  I'll try to find the time within the next few days.


-- 
Cristian



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-05 Thread Brian Potkin
On Mon 05 Nov 2018 at 21:41:58 +0100, Cristian Ionescu-Idbohrn wrote:

> On Mon, 5 Nov 2018, Brian Potkin wrote:
> > On Sun 04 Nov 2018 at 23:21:14 +, Brian Potkin wrote:
> > 
> > > avahi-deamon is essential; there were no instructions to uninstall it.
> > > Without it you will be unable to detect the DNS-SD broadcasts of the
> > > printer/scanner device. Put it back.
> > 
> > Please forget about this, it is nonsense.
> 
> Right.  I'll say that much.  I don't intend to follow the avahi track.
> 
> I just wish the bugs are fixed in 3.18.10+dfsg0-1, specially when 
> patches are readily available.  Just like other distributions (ubuntu, 
> mint, arch, fedora, and so on do already) do.  So, please patch.  The 
> little I digged in upstream bug DB, reported bugs (those that concern 
> me) seem to be buried, although solutions exist :(
> 
> 3.17.10+repack0-5 (in testing) is working for me as expected (with a 
> minimally patched models.dat). The downgraded to 3.17.10+repack0-5 in 
> unstable (with the same patch) is also usable, but not working as 
> well.  I see this kind of messages in the syslog (in unstable):
> 
> hp-toolbox(UI)[1127]: warning: Supplies information not available for this 
> device.
> /hpfax: [1292]: error: Failed to create/var/spool/cups/tmp/.hplip
> xpraforwarder: '/usr/lib/cups/backend/xpraforwarder'
> python3: io/hpmud/hpmud.c 246: invalid channel_open state, current 
> io_mode=raw/uni service=HP-MESSAGE 
> hp:/net/HP_ColorLaserJet_MFP_M278-M281?ip=192.168.x.y
> hpcups[3963]: prnt/hpcups/Hbpl1.cpp 52: Hbpl1 constructor : m_szLanguage = 
> HBPL1
> hp[7978]: prnt/backend/hp.c 919: ERROR: null print job total=0
> 
> and others.
> 
> The hplip* packages are highly rated in popcon (~40%).  So, that shows 
> the importance of stabilizing the hplip* packages.

It would be useful to have what happens with

 scanimage -d "hpaio:/net/HP_ColorLaserJet_MFP_M278-M281?ip=192.168.x.y" > 
out.pnm

for 3.18.10+dfsg0-1.

Does the scanner work? Do you get out.pnm?

Cheers,

Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-05 Thread Cristian Ionescu-Idbohrn
On Mon, 5 Nov 2018, Brian Potkin wrote:
> On Sun 04 Nov 2018 at 23:21:14 +, Brian Potkin wrote:
> 
> > avahi-deamon is essential; there were no instructions to uninstall it.
> > Without it you will be unable to detect the DNS-SD broadcasts of the
> > printer/scanner device. Put it back.
> 
> Please forget about this, it is nonsense.

Right.  I'll say that much.  I don't intend to follow the avahi track.

I just wish the bugs are fixed in 3.18.10+dfsg0-1, specially when 
patches are readily available.  Just like other distributions (ubuntu, 
mint, arch, fedora, and so on do already) do.  So, please patch.  The 
little I digged in upstream bug DB, reported bugs (those that concern 
me) seem to be buried, although solutions exist :(

3.17.10+repack0-5 (in testing) is working for me as expected (with a 
minimally patched models.dat). The downgraded to 3.17.10+repack0-5 in 
unstable (with the same patch) is also usable, but not working as 
well.  I see this kind of messages in the syslog (in unstable):

hp-toolbox(UI)[1127]: warning: Supplies information not available for this 
device.
/hpfax: [1292]: error: Failed to create/var/spool/cups/tmp/.hplip
xpraforwarder: '/usr/lib/cups/backend/xpraforwarder'
python3: io/hpmud/hpmud.c 246: invalid channel_open state, current 
io_mode=raw/uni service=HP-MESSAGE 
hp:/net/HP_ColorLaserJet_MFP_M278-M281?ip=192.168.x.y
hpcups[3963]: prnt/hpcups/Hbpl1.cpp 52: Hbpl1 constructor : m_szLanguage = HBPL1
hp[7978]: prnt/backend/hp.c 919: ERROR: null print job total=0

and others.

The hplip* packages are highly rated in popcon (~40%).  So, that shows 
the importance of stabilizing the hplip* packages.


-- 
cii



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-05 Thread Brian Potkin
On Sun 04 Nov 2018 at 23:21:14 +, Brian Potkin wrote:

> avahi-deamon is essential; there were no instructions to uninstall it.
> Without it you will be unable to detect the DNS-SD broadcasts of the
> printer/scanner device. Put it back.

Please forget about this, it is nonsense.

Cheers,

Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-04 Thread Brian Potkin
On Sun 04 Nov 2018 at 22:44:47 +0100, Cristian Ionescu-Idbohrn wrote:

> On Sun, 4 Nov 2018, Brian Potkin wrote:
> > 
> > Can we just forget about the bug you reported for a while and, to begin
> > with, get you printing satisfactorily?
> 
> But than you forget about the scanning, don't you, which is the major 
> part of the problem.

Nothing was forgotten; we were going step by step and dealing with
printing first. In any case, scanning was mentioned in another post.
Please see below for the detail.

You seem to be happy with printing, so we will leave that aspect of
the issue.

> > Why tie yourself into the tyranny of vendor supplied drivers? Get your
> > cups back to 2.2.8-5
> 
> But that's what it is at.  Both testing _and_ unstable:

That's fine.

> $ apt-cache policy cups
> cups:
>   Installed: 2.2.8-5
>   Candidate: 2.2.8-5
>   Version table:
>  2.3~b5-2 59
>  59 http://deb.debian.org/debian experimental/main amd64 
> Packages
>  2.2.8-5 500
> 500 http://deb.debian.org/debian testing/main amd64 Packages
>  99 http://deb.debian.org/debian unstable/main amd64 Packages
> 
> > and purge hplip and everything it pulled in. Have
> > the printer on the network and doing DNS-SD (Bonjour) broadcasting.
> > Purge cups-browsed. Now do
> > 
> >   lpstat -e
> > 
> > What do you get?
> 
> $ lpstat -e
> HP-ColorLaserJet-MFP-M278-M281
> 
> already gives me that (with the tyranny of vendor supplied drivers, 
> duplex too) and cups-browsed is already purged and avahi-daemon is not 
> installed.

avahi-deamon is essential; there were no instructions to uninstall it.
Without it you will be unable to detect the DNS-SD broadcasts of the
printer/scanner device. Put it back.

> AFAICT, m281fdw should be supported since some time (before 3.18.10), 
> but came out with new bug(s) that make it unusable :(  Is there a way 
> to find out narrow down the details?  Other distributions find 
> appropriate to identify, patch and feed back on bugs.  Is debian doing 
> that too?

For scanning do

  apt install libsane-hpaio --no-install-recommends

Then post the outputs of

  lpstat -l -e

and

  scanimage -L

Cheers,

Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-04 Thread Cristian Ionescu-Idbohrn
On Sun, 4 Nov 2018, Brian Potkin wrote:
> 
> Can we just forget about the bug you reported for a while and, to begin
> with, get you printing satisfactorily?

But than you forget about the scanning, don't you, which is the major 
part of the problem.

> Why tie yourself into the tyranny of vendor supplied drivers? Get your
> cups back to 2.2.8-5

But that's what it is at.  Both testing _and_ unstable:

$ apt-cache policy cups
cups:
  Installed: 2.2.8-5
  Candidate: 2.2.8-5
  Version table:
 2.3~b5-2 59
 59 http://deb.debian.org/debian experimental/main amd64 
Packages
 2.2.8-5 500
500 http://deb.debian.org/debian testing/main amd64 Packages
 99 http://deb.debian.org/debian unstable/main amd64 Packages

> and purge hplip and everything it pulled in. Have
> the printer on the network and doing DNS-SD (Bonjour) broadcasting.
> Purge cups-browsed. Now do
> 
>   lpstat -e
> 
> What do you get?

$ lpstat -e
HP-ColorLaserJet-MFP-M278-M281

already gives me that (with the tyranny of vendor supplied drivers, 
duplex too) and cups-browsed is already purged and avahi-daemon is not 
installed.

AFAICT, m281fdw should be supported since some time (before 3.18.10), 
but came out with new bug(s) that make it unusable :(  Is there a way 
to find out narrow down the details?  Other distributions find 
appropriate to identify, patch and feed back on bugs.  Is debian doing 
that too?


Cheers,

-- 
Cristian



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-04 Thread Brian Potkin
On Sun 04 Nov 2018 at 20:03:52 +0100, Till Kamppeter wrote:

> Brian, if the user only wants to print with his printer (is it a print-only
> device or a multi-function device with scanner) driverless IPP printing
> works indeed, especially with HP devices. Then this is the recommended
> solution.

Till, as I said, I know I am not addressing the bug report directly, but
my skills do not extend to code writing; you have indicated a way forward
in this regard. Roll on driverless printing and raising users' awareness
it exists.

> Only for scanning one still needs drivers and in case of HP's multi-function
> devices HPLIP (a driverless IPP scanning standard is already there, but not
> yet adopted in actual hardware).

I was prepared to come to the question of scanning later. This does need
stuff from HP (which has done well by me) but not as much as provided by
the hplip package.

  apt install libsane-hpaio --no-install-recommends

gets everything needed for scanning.

Regards,

Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-04 Thread Till Kamppeter
In general, it looks like that the HPLIP guys at HP are not testing well 
the GUI part. This caused the following bugs, all forwarded upstream but 
no fixes from upstream yet, only distro patches in the Ubuntu Cosmic 
package of HPLIP (3.18.7+dfsg1-2ubuntu2):


https://launchpad.net/bugs/1688684
  hp-check does not show distro names correctly but internal
  index numbers for them

https://launchpad.net/bugs/1745383
  QMessageBox() is called incorrectly

https://launchpad.net/bugs/1789184
  hp-toolbox does not start at all

The drivers themselves (CUPS filters/PPD files and SANE module) seem to 
work reasonably well, but for some small part of the device range a 
proprietary plugin is needed, and some of these devices probably also 
offer driverless IPP printing and this way one could work around the 
proprietary plugin at least if it concerns only printing.


For diagnostic purposes you can perhaps also use hp-check instead of 
hp-doctor as it has no GUI.


   Till



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-04 Thread Till Kamppeter
Brian, if the user only wants to print with his printer (is it a 
print-only device or a multi-function device with scanner) driverless 
IPP printing works indeed, especially with HP devices. Then this is the 
recommended solution.


Only for scanning one still needs drivers and in case of HP's 
multi-function devices HPLIP (a driverless IPP scanning standard is 
already there, but not yet adopted in actual hardware).


   Till



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-04 Thread Brian Potkin
On Sun 04 Nov 2018 at 15:45:41 +0100, Cristian Ionescu-Idbohrn wrote:

> On Sun, 4 Nov 2018, Cristian Ionescu-Idbohrn wrote:
> > On Sat, 3 Nov 2018, Brian Potkin wrote:
> > > On Sat 03 Nov 2018 at 19:20:42 +0100, Cristian Ionescu-Idbohrn wrote:
> > > 
> > > > Package: hplip-data
> > > > Version: 3.18.10+dfsg0-1
> > > > Severity: grave
> > > > Justification: renders package unusable
> > > 
> > > 
> > > If you would explain why this is grave, we could adjust the severity.
> > 
> > I can no longer use primarily the scanner.
> > I think it got broken in 3.17.10 already.
> 
> So, what did I do?  Downgraded to 3.17.10+repack0-5.  Downgraded cups 
> too, to 2.2.8-5 (although I don't know if I had to; please advice, if 
> you can).  Got a laptop running testing where hplip/cups works as 
> expected, almost.  And the hack to make it work is here:
> 
>   https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=889468
> 
> This is just _one_ of the examples of how it it might look like:
> 
>   https://bugzilla.redhat.com/show_bug.cgi?id=1544912
> 
> And I experience similar problems.  So, this a list of what I 
> downgraded:
> 
> cups-bsd_2.2.8-5_amd64.deb
> cups-client_2.2.8-5_amd64.deb
> cups-common_2.2.8-5_all.deb
> cups-core-drivers_2.2.8-5_amd64.deb
> cups-daemon_2.2.8-5_amd64.deb
> cups-filters-core-drivers_1.21.3-2_amd64.deb
> cups-filters_1.21.3-2_amd64.deb
> cups-ipp-utils_2.2.8-5_amd64.deb
> cups-pk-helper_0.2.6-1+b1_amd64.deb
> cups-ppdc_2.2.8-5_amd64.deb
> cups-server-common_2.2.8-5_all.deb
> cups_2.2.8-5_amd64.deb
> hplip-data_3.17.10+repack0-5_all.deb
> hplip-gui_3.17.10+repack0-5_all.deb
> hplip_3.17.10+repack0-5_amd64.deb
> libcups2_2.2.8-5_amd64.deb
> libcupscgi1_2.2.8-5_amd64.deb
> libcupsfilters1_1.21.3-2_amd64.deb
> libcupsimage2_2.2.8-5_amd64.deb
> libcupsmime1_2.2.8-5_amd64.deb
> libcupsppdc1_2.2.8-5_amd64.deb
> libhpmud0_3.17.10+repack0-5_amd64.deb
> libsane-hpaio_3.17.10+repack0-5_amd64.deb
> printer-driver-hpcups_3.17.10+repack0-5_amd64.deb
> python3-cups_1.9.73-2+b1_amd64.deb
> python3-cupshelpers_1.5.11-3_all.deb
> 
> That's put on hold for now.
> 
> I got back printing and scanning functionality, but not all :(
> 
> The main problem with the hplip 3.18.10 is that hp-setup associates 
> the wrong:
> 
>   Found PPD file: hplip:0/ppd/hplip/HP/hp-laserjet_mfp_m28-m31.ppd
> 
> (which is not a _color_ printer, does not do _duplex_ and has only 
> _one_ tray, among other things)
> 
> with the printer:
> 
>   HP Color LaserJet MFP M281fdw
> 
> It should be:
> 
>   Found PPD file: 
> hplip:0/ppd/hplip/HP/hp-color_laserjet_pro_mfp_m277-ps.ppd  
> 
> instead.  This post:
> 
>   
> https://forums.fedoraforum.org/showthread.php?317915-Print-amp-scan-with-an-HP-Color-LaserJet-MFP-M281-(fdw)=1806395
> 
> could be another way to do it, but I don't know if that works with 
> hplip 3.18.10.
> 
> I wish hp-toolbox displayed like in attachment 1, not like in 
> attachment 2, but I'll live with it ;)
> 
> Any advise on ways to cleanup this mess is highly appreciated.

Can we just forget about the bug you reported for a while and, to begin
with, get you printing satisfactorily?

Why tie yourself into the tyranny of vendor supplied drivers? Get your
cups back to 2.2.8-5 and purge hplip and everything it pulled in. Have
the printer on the network and doing DNS-SD (Bonjour) broadcasting.
Purge cups-browsed. Now do

  lpstat -e

What do you get?

Cheers,

Brian



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-04 Thread Cristian Ionescu-Idbohrn
On Sun, 4 Nov 2018, Cristian Ionescu-Idbohrn wrote:
> On Sat, 3 Nov 2018, Brian Potkin wrote:
> > On Sat 03 Nov 2018 at 19:20:42 +0100, Cristian Ionescu-Idbohrn wrote:
> > 
> > > Package: hplip-data
> > > Version: 3.18.10+dfsg0-1
> > > Severity: grave
> > > Justification: renders package unusable
> > 
> > 
> > If you would explain why this is grave, we could adjust the severity.
> 
> I can no longer use primarily the scanner.
> I think it got broken in 3.17.10 already.

So, what did I do?  Downgraded to 3.17.10+repack0-5.  Downgraded cups 
too, to 2.2.8-5 (although I don't know if I had to; please advice, if 
you can).  Got a laptop running testing where hplip/cups works as 
expected, almost.  And the hack to make it work is here:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=889468

This is just _one_ of the examples of how it it might look like:

https://bugzilla.redhat.com/show_bug.cgi?id=1544912

And I experience similar problems.  So, this a list of what I 
downgraded:

cups-bsd_2.2.8-5_amd64.deb
cups-client_2.2.8-5_amd64.deb
cups-common_2.2.8-5_all.deb
cups-core-drivers_2.2.8-5_amd64.deb
cups-daemon_2.2.8-5_amd64.deb
cups-filters-core-drivers_1.21.3-2_amd64.deb
cups-filters_1.21.3-2_amd64.deb
cups-ipp-utils_2.2.8-5_amd64.deb
cups-pk-helper_0.2.6-1+b1_amd64.deb
cups-ppdc_2.2.8-5_amd64.deb
cups-server-common_2.2.8-5_all.deb
cups_2.2.8-5_amd64.deb
hplip-data_3.17.10+repack0-5_all.deb
hplip-gui_3.17.10+repack0-5_all.deb
hplip_3.17.10+repack0-5_amd64.deb
libcups2_2.2.8-5_amd64.deb
libcupscgi1_2.2.8-5_amd64.deb
libcupsfilters1_1.21.3-2_amd64.deb
libcupsimage2_2.2.8-5_amd64.deb
libcupsmime1_2.2.8-5_amd64.deb
libcupsppdc1_2.2.8-5_amd64.deb
libhpmud0_3.17.10+repack0-5_amd64.deb
libsane-hpaio_3.17.10+repack0-5_amd64.deb
printer-driver-hpcups_3.17.10+repack0-5_amd64.deb
python3-cups_1.9.73-2+b1_amd64.deb
python3-cupshelpers_1.5.11-3_all.deb

That's put on hold for now.

I got back printing and scanning functionality, but not all :(

The main problem with the hplip 3.18.10 is that hp-setup associates 
the wrong:

Found PPD file: hplip:0/ppd/hplip/HP/hp-laserjet_mfp_m28-m31.ppd

(which is not a _color_ printer, does not do _duplex_ and has only 
_one_ tray, among other things)

with the printer:

HP Color LaserJet MFP M281fdw

It should be:

Found PPD file: 
hplip:0/ppd/hplip/HP/hp-color_laserjet_pro_mfp_m277-ps.ppd  

instead.  This post:


https://forums.fedoraforum.org/showthread.php?317915-Print-amp-scan-with-an-HP-Color-LaserJet-MFP-M281-(fdw)=1806395

could be another way to do it, but I don't know if that works with 
hplip 3.18.10.

I wish hp-toolbox displayed like in attachment 1, not like in 
attachment 2, but I'll live with it ;)

Any advise on ways to cleanup this mess is highly appreciated.


-- 
cii

Bug#912768: hplip-data: hp-toolbox fsck

2018-11-04 Thread Cristian Ionescu-Idbohrn
On Sat, 3 Nov 2018, Brian Potkin wrote:
> On Sat 03 Nov 2018 at 19:20:42 +0100, Cristian Ionescu-Idbohrn wrote:
> 
> > Package: hplip-data
> > Version: 3.18.10+dfsg0-1
> > Severity: grave
> > Justification: renders package unusable
> 
> 
> If you would explain why this is grave, we could adjust the severity.

I can no longer use primarily the scanner.
I think it got broken in 3.17.10 already.


-- 
Cristian



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-03 Thread Brian Potkin
severity 912768 normal
thanks



On Sat 03 Nov 2018 at 19:20:42 +0100, Cristian Ionescu-Idbohrn wrote:

> Package: hplip-data
> Version: 3.18.10+dfsg0-1
> Severity: grave
> Justification: renders package unusable


If you would explain why this is grave, we could adjust the severity.

Regards,

Brian.



Bug#912768: hplip-data: hp-toolbox fsck

2018-11-03 Thread Till Kamppeter
I have fixed this bug and two others one in the HPLIP package for Ubuntu 
Cosmic (18.10). Simply overtake the two patches which I have added.


https://launchpad.net/ubuntu/+source/hplip/+changelog
https://launchpad.net/ubuntu/+source/hplip/3.18.7+dfsg1-2ubuntu2
https://launchpad.net/ubuntu/+source/hplip/3.18.7+dfsg1-2ubuntu1

https://launchpad.net/bugs/1789184

   Till


On 03/11/2018 19:20, Cristian Ionescu-Idbohrn wrote:

Package: hplip-data
Version: 3.18.10+dfsg0-1
Severity: grave
Justification: renders package unusable

$ hp-toolbox

HP Linux Imaging and Printing System (ver. 3.18.10)
HP Device Manager ver. 15.0

Copyright (c) 2001-15 HP Development Company, LP
This software comes with ABSOLUTELY NO WARRANTY.
This is free software, and you are welcome to distribute it
under certain conditions. See COPYING file for more details.

QStandardPaths: XDG_RUNTIME_DIR not set, defaulting to '/tmp/runtime-cii'   
   \Traceback (most recent call last):
   File "/usr/bin/hp-toolbox", line 280, in 
 toolbox = ui.DevMgr5(__version__, device_uri,  None)
   File "/usr/share/hplip/ui5/devmgr5.py", line 253, in __init__
 self.initUI()
   File "/usr/share/hplip/ui5/devmgr5.py", line 324, in initUI
 self.DiagnoseQueueAction.setIcon(QIcon(load_pixmap('warning', '16x16')))
AttributeError: 'DevMgr5' object has no attribute 'DiagnoseQueueAction'

Patch here:

https://bugs.launchpad.net/ubuntu/+source/hplip/+bug/1789184/comments/7

seems to help.

-- System Information:
Debian Release: buster/sid
   APT prefers unstable
   APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 4.18.0-1-amd64 (SMP w/8 CPU cores)
Locale: LANG=C, LC_CTYPE=C (charmap=ANSI_X3.4-1968), LANGUAGE=en_US:en 
(charmap=ANSI_X3.4-1968)
Shell: /bin/sh linked to /bin/dash
Init: sysvinit (via /sbin/init)

Versions of packages hplip-data depends on:
ii  python3   3.6.7-1
ii  xz-utils  5.2.2-1.3

hplip-data recommends no packages.

Versions of packages hplip-data suggests:
ii  hplip  3.18.10+dfsg0-1

-- no debconf information

-- debsums errors found:
debsums: changed file /usr/share/hplip/ui5/devmgr5.py (from hplip-data package)


Cheers,





Bug#912768: hplip-data: hp-toolbox fsck

2018-11-03 Thread Cristian Ionescu-Idbohrn
Package: hplip-data
Version: 3.18.10+dfsg0-1
Severity: grave
Justification: renders package unusable

$ hp-toolbox

HP Linux Imaging and Printing System (ver. 3.18.10)
HP Device Manager ver. 15.0

Copyright (c) 2001-15 HP Development Company, LP
This software comes with ABSOLUTELY NO WARRANTY.
This is free software, and you are welcome to distribute it
under certain conditions. See COPYING file for more details.

QStandardPaths: XDG_RUNTIME_DIR not set, defaulting to '/tmp/runtime-cii'   
   \Traceback (most recent call last):
  File "/usr/bin/hp-toolbox", line 280, in 
toolbox = ui.DevMgr5(__version__, device_uri,  None)
  File "/usr/share/hplip/ui5/devmgr5.py", line 253, in __init__
self.initUI()
  File "/usr/share/hplip/ui5/devmgr5.py", line 324, in initUI
self.DiagnoseQueueAction.setIcon(QIcon(load_pixmap('warning', '16x16')))
AttributeError: 'DevMgr5' object has no attribute 'DiagnoseQueueAction'

Patch here:

https://bugs.launchpad.net/ubuntu/+source/hplip/+bug/1789184/comments/7

seems to help.

-- System Information:
Debian Release: buster/sid
  APT prefers unstable
  APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 4.18.0-1-amd64 (SMP w/8 CPU cores)
Locale: LANG=C, LC_CTYPE=C (charmap=ANSI_X3.4-1968), LANGUAGE=en_US:en 
(charmap=ANSI_X3.4-1968)
Shell: /bin/sh linked to /bin/dash
Init: sysvinit (via /sbin/init)

Versions of packages hplip-data depends on:
ii  python3   3.6.7-1
ii  xz-utils  5.2.2-1.3

hplip-data recommends no packages.

Versions of packages hplip-data suggests:
ii  hplip  3.18.10+dfsg0-1

-- no debconf information

-- debsums errors found:
debsums: changed file /usr/share/hplip/ui5/devmgr5.py (from hplip-data package)


Cheers,

-- 
Cristian