Duplex print selections ignored by M283fdw
I am using a HP MFP M283fdw with hplip version 3.20.9. The current driver setting is "ColorLaserJet MFP M282-M285, driverless, cups-filters 1.27.4".
The printer's duplex capabilities are available both in hp-toolbox and in application print dialogs. But any selection done in hp-toolbox or in print dialogs is ignored and does not get propagated to the printer.
The only "side channel" way to control duplex printing is by using the printer's web server, in the System -> Paper-Setup page. I can confirm the device actually is capable of duplex printing, but this is not a satisfying workaround.
I have also tried other versions of hplip (3.20.3, 3.20.11) and other drivers ("ColorLaserJet MFP M282-M285 Postscript (recommended)"), to no avail. Rather, using the Postscript driver disables per-job duplex selection to applications entirely, but then honors settings done in hp-toolbox.
Question information
- Language:
- English Edit question
- Status:
- Solved
- For:
- HPLIP Edit question
- Assignee:
- No assignee Edit question
- Solved by:
- brian_p
- Solved:
- Last query:
- Last reply:
Revision history for this message
|
#1 |
You do not say what your OS is. We assume you have a wireless
connection.
Give the output of
driverless
--
Brian.
Revision history for this message
|
#2 |
I'm sorry: The OS is Ubuntu 20.04.
Yes, the network link is a WIFI connection. How does that, as opposed to an Ethernet link, relate to this issue?
$ driverless
ipp://HP%
Revision history for this message
|
#3 |
There isn't any difference between WIFI and ethernet connections.
The output of driverless is the URI. It is substituted in the next command.
Set up this print queue:
lpadmin -p m127test -v URI -E -o sides=two-
Print to it with
lp -d m127test /etc/services
Does it print in duplex?
remove the queue with
lpadmin -x m127test
Revision history for this message
|
#5 |
Substituting the output of 'driverless' as URI
$ lpadmin -p m127test -v "ipp://
created the queue, but the lp command doesn't print at all. The job is immediately stopped, reporting "Idle - 'Filter failed'".
Revision history for this message
|
#6 |
The filter issue above is apparently caused by using /etc/services as content. Created a two page PDF document and executed
$ lp -d m127test TwoPages.pdf
request id is m127test-268 (1 file(s))
The document is printed non-duplex on two sheets.
Revision history for this message
|
#7 |
> created the queue, but the lp command doesn't print at all.
> The job is immediately stopped, reporting "Idle - 'Filter failed'"
Would you repeat the queue setup with
-o Duplex=
and repeat printing with /etc/adduser.conf and TwoPages.pdf.
lpoptions -p m127test -l
should show DuplexNoTumble as the default.
Revision history for this message
|
#8 |
Printing /etc/adduser.conf results in a filter failure in either case, so I just could test print with my two page PDF.
$ lpadmin -p m127test -v "ipp://
$ lp -d m127test TwoPages.pdf
request id is m127test-271 (1 file(s))
$ lpoptions -p m127test -l
PageSize/Media Size: 100x150mm 184x260mm 195x270mm 4x6 5x8 *A4 A5 A6 B5 B6 DoublePostcardR
InputSlot/Media Source: *Auto Manual Tray1 Tray2
MediaType/Media Type: *Stationery Com.hp.
ColorModel/Output Mode: Gray AdobeRGB DeviceRGB DeviceGray *RGB
Duplex/Duplex: None *DuplexNoTumble DuplexTumble
cupsPrintQualit
DuplexNoTumble is indeed the default, but the print ends up non-duplex on two sheets. It doesn't make any difference whether I also include "-o sides=two-
Revision history for this message
|
#9 |
I am flummoxed!
Please give
avahi-browse -rt _ipp._tcp
and
avahi-browse -rt _uscan._tcp
Revision history for this message
|
#10 |
$ avahi-browse -rt _ipp._tcp
+ enp5s0 IPv6 HP Color LaserJet MFP M283fdw (EFBB7A) Internet Printer local
+ enp5s0 IPv4 HP Color LaserJet MFP M283fdw (EFBB7A) Internet Printer local
= enp5s0 IPv4 HP Color LaserJet MFP M283fdw (EFBB7A) Internet Printer local
hostname = [NPIEFBB7A.local]
address = [192.168.178.253]
port = [631]
txt = ["mopria-
= enp5s0 IPv6 HP Color LaserJet MFP M283fdw (EFBB7A) Internet Printer local
hostname = [NPIEFBB7A.local]
address = [192.168.178.253]
port = [631]
txt = ["mopria-
$ avahi-browse -rt _uscan._tcp
+ enp5s0 IPv4 HP Color LaserJet MFP M283fdw (EFBB7A) _uscan._tcp local
= enp5s0 IPv4 HP Color LaserJet MFP M283fdw (EFBB7A) _uscan._tcp local
hostname = [NPIEFBB7A.local]
address = [192.168.178.253]
port = [8080]
txt = ["mopria-
+ enp5s0 IPv6 HP Color LaserJet MFP M283fdw (EFBB7A) _uscan._tcp local
= enp5s0 IPv6 HP Color LaserJet MFP M283fdw (EFBB7A) _uscan._tcp local
hostname = [NPIEFBB7A.local]
address = [192.168.178.253]
port = [8080]
txt = ["mopria-
Revision history for this message
|
#11 |
From the text record:
> "Duplex=T"
The printer claims to do automatic duplex, but we knew that anyway.
I have just set up a queue for my ENVY with -o Duplex=
I get automatic duplex with /etc/adduser.conf on Debian unstable.
It is almost as though your printer is ignoring DuplexNoTumble, which
would be a firmware bug, or there a bug in your cups-filters.
More to come in 5/10 minutes.
Revision history for this message
|
#12 |
In your first post there is
> "ColorLaserJet MFP M282-M285, driverless, cups-filters 1.27.4".
This indicates the queue has been automatically set up by
cups-browsed from what is in the printer's Bonjour broadcasts.
In other words, HPLIP is not being used. The queue I asked you
to set up is not significantly different from the cups-filters one.
We'll try using HPLIP. Do
lpinfo -m | grep "M282-M285"
This gives you a PPD to substitute in the next command. It begins
"postscript" and ends "ppd". Now
lpadmin -p m127test -v "URI" -E -o sides=two-
Revision history for this message
|
#13 |
I'm afraid, no:
$ lpadmin -p m127test -v "postscript-
lpadmin: Printer drivers are deprecated and will stop working in a future version of CUPS.
lpadmin: Bad device-uri scheme "postscript-hp".
But, as mentioned in my initial request, I have also tried other versions of hplip (3.20.3, 3.20.11) and other drivers ("ColorLaserJet MFP M282-M285 Postscript (recommended)"), to no avail.
That is the driver configured by hp-setup, and tried it both with queue created by hp-setup as well as KDE's printer control page. It still can't propagate per-job duplex selections to the printer.
Revision history for this message
|
#14 |
> lpadmin -p m127test
> -v "postscript-
> -E -o sides=two-
Michael, you have the URI earlier in the thread! It is
"ipp:/
Also
postscript-
is substituted for PPD.
> But, as mentioned in my initial request, I have also tried other versions of
> hplip (3.20.3, 3.20.11) and other drivers ("ColorLaserJet MFP M282-M285
> Postscript (recommended)"), to no avail.
Indeed you did, but let's see this through.
Revision history for this message
|
#15 |
I forgot to mention: -o sides=two-
for the queue. Just try printing TwoPages.pdf from an application in duplex.
Revision history for this message
|
#16 |
I have used the printer config dialog to reconfigure my default queue to use the "ColorLaserJet MFP M282-M285 Postscript" driver instead. That should now make sure it's routing via hplip, doesn't it?
The strange observation: In the print job dialog, this queue now has duplex options disabled, just as the printer would no longer support it! In the avahi-browse output I can't see any difference (see below). Is it correct that these queries also bypass hplip and directly query print services on the network?
Another weird observation: While _ipp._tcp returns "Duplex=T", _uscan._tcp reports "duplex=F". Is that to describe that printing support duplex, while scanning doesn't? Or is this some contradiction?
$ avahi-browse -rt _ipp._tcp
+ enp5s0 IPv6 HP Color LaserJet MFP M283fdw (EFBB7A) Internet Printer local
+ enp5s0 IPv4 HP Color LaserJet MFP M283fdw (EFBB7A) Internet Printer local
= enp5s0 IPv6 HP Color LaserJet MFP M283fdw (EFBB7A) Internet Printer local
hostname = [NPIEFBB7A.local]
address = [fe80::
port = [631]
txt = ["mopria-
= enp5s0 IPv4 HP Color LaserJet MFP M283fdw (EFBB7A) Internet Printer local
hostname = [NPIEFBB7A.local]
address = [192.168.178.253]
port = [631]
txt = ["mopria-
$ avahi-browse -rt _uscan._tcp
+ enp5s0 IPv4 HP Color LaserJet MFP M283fdw (EFBB7A) _uscan._tcp local
+ enp5s0 IPv6 HP Color LaserJet MFP M283fdw (EFBB7A) _uscan._tcp local
= enp5s0 IPv4 HP Color LaserJet MFP M283fdw (EFBB7A) _uscan._tcp local
hostname = [NPIEFBB7A.local]
address = [192.168.178.253]
port = [8080]
txt = ["mopria-
= enp5s0 IPv6 HP Color LaserJet MFP M283fdw (EFBB7A) _uscan._tcp local
hostname = [NPIEFBB7A.local]
address = [192.168.178.253]
port = [8080]
txt = ["mopria-
Revision history for this message
|
#17 |
Bingo! The command
$ lpadmin -p m127test -v "ipp://
created a queue where "lp -d m127test TwoPages.pdf" does print duplex. But as Said in my previous post, the application now consideres duplex not to be supported for m127test.
Revision history for this message
|
#18 |
It seems we are getting closer. Thanks to all the tools you provided me with :)
The "lpoptions -p m127test -l" correctly reports the available options and default as seen by CUPS. The Dialog doesn't because there is an extra menu for "Installable options" in the printer setup form, which has Duplex Unit set to "Off". If I change that to "On", I can set the option and it's honored by the printer.
When using the driverless "driver", there is a "Two-Sided Printing" option, which can be enabled, and does propagate the print dialog default setting, but is then ignored on printing.
It's confusing: What is the concept of an "Installable option"? Is that a CUPS thing, or something in HPLIP?
Revision history for this message
|
#19 |
> That should now make sure it's routing via hplip, doesn't it?
Yes.
> Is it correct that these queries also bypass hplip and directly
> query print services on the network?
Correct. The driverless setup sets up a PDD by *directly* querying
the printer The Postscript PPD is provided by HP from what *they*
know about the printer's capabilities.
> Another weird observation: While _ipp._tcp returns "Duplex=T",
> _uscan._tcp reports "duplex=F". Is that to describe that printing
> support duplex, while scanning doesn't? Or is this some
> contradiction?
No. Printing and scanning are separate processes. "duplex=F" means
that the ADF will not do duplex.
Revision history for this message
|
#20 |
'lpoptions -p m127test -l' has this:
HPOption Duplexer/Duplex Unit: True/*False
This is non-standard. It appears to have to be set to "True" for
automatic duplexing to take place.
Would you confirm that you now have double-sided printing
via HPLIP?
Revision history for this message
|
#21 |
Michael,
I am still very bothered that the cups-browsed queue set up for you by the
the system does not work. Would you help in debugging this by giving the
output of
ipptool -tv "URI" get-printer-
Revision history for this message
|
#22 |
> When using the driverless "driver", there is a "Two-Sided Printing"
> option, which can be enabled, and does propagate the print dialog
> default setting, but is then ignored on printing.
Which application was used for printing?
> It's confusing: What is the concept of an "Installable option"? Is that a
> CUPS thing, or something in HPLIP?
I am unsure about this but the actual option is an HP-specific one.
Revision history for this message
|
#23 |
Responding the last three post:
> HPOption Duplexer/Duplex Unit: True/*False
>
>This is non-standard. It appears to have to be set to "True" for
>automatic duplexing to take place.
>
>Would you confirm that you now have double-sided printing
>via HPLIP?
Yes, by using the lp command on m127test queue. And also from print dialog after enabling the duplexer installable option. The lpoptions output then changes to "HPOption_
>I am still very bothered that the cups-browsed queue set up for you by the
>the system does not work. Would you help in debugging this by giving the
>output of
>
> ipptool -tv "URI" get-printer-
That fails. Do I need to use a different URI than the one user for queue creation?
$ ipptool -tv "ipp://
ipptool: Unable to connect to "HP Color LaserJet MFP M283fdw (EFBB7A)
I could recreate the queue with exactly that URI, but ipptool still fails.
>Which application was used for printing?
The okular PDF viewer.
Revision history for this message
|
#24 |
Try
ipptool -tv "ipp://
Yes, I guessed you were using Okular; thanks for confirmation. Any chance that
you could try double-sided printing with Firefox or Evince? They use a different
print dialog from Okular.
Revision history for this message
|
#25 |
$ ipptool -tv "ipp://
"/usr/share/
Get-
printer-uri (uri) = ipp://192.
Get printer attributes using get-printer-
RECEIVED: 948609 bytes in response
status-code = successful-ok (successful-ok)
Revision history for this message
|
#26 |
I forgot again!
The queue for testing Firefox or Evince should be
lpadmin -p m127test -v "URI" -E -m everywhere
"URI" can be what we had before or
"ipp:
Revision history for this message
|
#27 |
With the installable option Duplexer *Off*
- Firefox Page Setup offers Two-Sided -> One Sided|Long Edge|Short Edge. When selecting "Long Edge", it immediately prompts
"Some of the settings in the dialog conflict"
- The KDE okular print options tab displays duplex printing radio button list None, Long side, Short side, with the latter two disabled and the disable "Long side" preselected. Accepting this selection a popup occurs
"There are conflicts in duplex section. Do you want to fix them?"
With the installable option Duplexer *On* both Firefox and KDE okular offer duplex and print duplex.
With driverless bypass of HPLIP, both Firefox and KDE okular offer duplex, but print non-duplex on two sheets .
Revision history for this message
|
#28 |
Thank you for your efforts, Michael. I am disappointed that the
driverless bypass does not work for you. However, you are now
printing in duplex with HPLIP and that is good. You could now
consider marking this issue as solved. I do not think I have more
to contribute to it.
How did you go on with the ipptool command?
Revision history for this message
|
#29 |
And with the queue created with "-m everywhere" as you described, there is no installable option. CUPS reports no HPOption, either
$ lpoptions -p m127test -l
PageSize/Media Size: 100x150mm 184x260mm 195x270mm 4x6 5x8 *A4 A5 A6 B5 B6 DoublePostcardR
InputSlot/Media Source: *Auto Manual Tray1 Tray2
MediaType/Media Type: *Stationery Com.hp.
ColorModel/Output Mode: Gray AdobeRGB DeviceRGB DeviceGray *RGB
Duplex/Duplex: *None DuplexNoTumble DuplexTumble
cupsPrintQualit
Both Firefox and KDE okular offer duplex printing and print duplex as requested. The driver is reported to be "ColorLaserJet MFP M282-M285 - IPP Everywhere". Does that one route through HPLIP?
Revision history for this message
|
#30 |
Agreed. The output of ipptool is in comment #25. I'm now having two options working fine: The postscript PPD driver with Duplexer installable option enabled, and IPP Everywhere. The latter is not offered in HP vendor driver list of the KDE printer setup dialog. What is the difference between these drivers and which of them is preffered?
Revision history for this message
|
#31 |
> Both Firefox and KDE okular offer duplex printing and print duplex as requested.
> The driver is reported to be "ColorLaserJet MFP M282-M285 - IPP Everywhere".
> Does that one route through HPLIP?
No. Printing through IPP Everywhere and HPLIP uses completely independent
techniques, IPP Everywhere is the more modern method and works with recent
printers like yours. I would use it myself, but there is nothing wrong with using
HPLIP.
Revision history for this message
|
#32 |
Thank a lot Brian!
Revision history for this message
|
#33 |
> I'm now having two options working fine: The postscript PPD driver
> with Duplexer installable option enabled, and IPP Everywhere. The
> latter is not offered in HP vendor driver list of the KDE printer setup
> dialog.
I am not familiar with the KDE printer setup dialog but IPP Everywhere
will not be offered in the HP vendor driver list because it has nothing to
do with HP.
> What is the difference between these drivers and which of them is
> preffered?
IPP Everywhere.is built into the printing system; It is an integral part of it
and works well with *all* modern printers like yours. HPLIP is an addon
package specifically for HP printers. I prefer IPP Everywhere because it is
more versatile and futureproof.
Revision history for this message
|
#34 |
Thanks brian_p, that solved my question.