The officially official Devuan Forum!

You are not logged in.

#501 Re: Hardware & System Configuration » [SOLVED] CUPS and printer drivers » 2024-06-25 09:13:25

Hello:

... as I understand it.

Yes, that is one way.

But you can forego the hplip part by installing CUPS and adding the local printer via the UI.
If the printer is connected (parallel, serial, USB), the system will see it and as a result, CUPS will find it.

To finish you only need to provide it with the printer's *.ppd file, at least that is how I recall having done it at the time. 

As I mentioned in the OP, at some point in the future CUPS will no longer use *.ppd files.
It is not clear when that will happen but it has been talked about since 2021.
As a result of this change, it would seem that whatever printer you have installed will either work as usual or need some application from the OEM to work.

I do not expect Samsung/HP to provide any such thing.

It is still not clear to me if the M2020W will work or I will need to pin CUPS at the last working version which is why I asked.

Thanks for your input.

Best,

A.

#502 Re: Hardware & System Configuration » [SOLVED] CUPS and printer drivers » 2024-06-24 19:19:12

Hello:

aluma wrote:

... indestructible LaserJet ...

Yes.

I still shed a tear when I remember all those LaserJet 4/5 sent to the landfill in early 2000 where I worked.
It was a very different Hewlett Packard back then.

aluma wrote:

... run the install.sh script in the uld folder.

Indeed ...

After having to wrangle (at that time) with this printer's installation in Linux, I resorted to CUPS, which worked from the get-go.

As it did a good job, I want to continue with CUPS instead of using the Samsung/HP ware and its install routine.
Besides, doing so is more akin to the Unix/Linux philosophy.

Thanks for your input.

Best,

A.

#503 Hardware & System Configuration » [SOLVED] CUPS and printer drivers » 2024-06-24 13:19:14

Altoid
Replies: 6

Hello:

After quite a bit of hassle, I was able to get CUPS working properly in my upgraded Daedalus installation.
It was working perfectly well in Beowulf but something happened on the way to Daedalus, so to speak and eventually had to resort to a complete purge of anything and everything CUPS related to then reinstall the application.

Doing things 'the Windows way' is not my idea of fixing things like these in a proper manner, but I was at my wits' end.

I have no idea why all this happened but now it seems to work properly.
As I do not print much, time will tell if it holds or not.

My printer is an inexpensive+basic Samsung / HP Xpress M2020W released ca. 2014, good enough* for what I need at home.
* no ethernet port, just WiFi (which I do not use) but given the price I paid for it cannot complain.

I managed to get it working by using the *.ppd file Samsung_M2020_Series.ppd extracted from a driver file named uld_V1.00.39_01.17.tar.gz downloaded directly from HP, which seems to be the last available version.

Note: it would seem that a file named rastertospl, located the uld/x86_64/ folder is also needed to get the printer working properly.
Otherwise you get an error message in the log which reads:

File "/usr/lib/cups/filter/rastertospl" not available: No such file or directory

It does not seem to be available in any Linux package and the problem is solved by copying it to that directory as root.
Now, with CUPS working properly, at boot time my /var/log/cups/error_log reads:

~$ cat /var/log/cups/error_log
W [24/Jun/2024:07:37:19 -0300] Printer drivers are deprecated and will stop working in a future version of CUPS. See https://github.com/OpenPrinting/cups/issues/103
~$ 

The link in the terminal printout redirects to this webpage which started getting comments in 02/2021, over three years ago.

"Printer drivers are deprecated and will stop working" reads just as forboding as when Nvidia dropped their legacy 340.108 drivers.
Hopefully I am wrong.

The whole thread I linked to above is pretty much printer protocols/drivers specific and is way over my head.

So the question is:
Anyone here at Dev1 use one of these printers (or any other printer needing a *.ppd) with CUPS without resorting to specific printer drivers?

Thanks in advance.

Best,

A.

#504 Re: Hardware & System Configuration » [SOLVED] ThinkPad, Qualcomm Gobi 2000 and SIM card. » 2024-06-17 09:35:53

Hello:

aluma wrote:

Thank you ...

You're welcome.

... but what a process...

Indeed ...
But at least there is a solution to whatever is evidently wrong with that HW.

That said, I cannot imagine why it would be disabled in this manner, but then again ...
Cheaper this way for the OEM?

Please do let the forum know how you fared with this endeavour.
It will surely be useful to others in the same spot.

Best,

A.

#505 Re: Hardware & System Configuration » [SOLVED] ThinkPad, Qualcomm Gobi 2000 and SIM card. » 2024-06-16 20:50:15

Hello:

aluma wrote:

... anyone know what is missing?

Quite possibly (?) the correct firmware is missing/not being loaded.
Seems that hardware has a few issues:

pdc_2 wrote:

... if you want to slowly drive yourself mad; and quietly pull all your hair out at home; all by yourself ...

See: https://forums.opensuse.org/t/gobi2000- … swer/71270

stakanov wrote:

For the record:
Laptop is an Lenovo X201 12’ Notebook. It was sold with WWAN (generical, not bound to provider - at least sold like this) and sold without any O.S. (which made me avoid the MS tax. The unpacking of the latest driver was done with wine, The option choosen was to combine the firmware solution “6” and “UMTS”. I copied all 3 files in /lib/firware/gobi (after having compiled successfully Gobiloader 0.7 from the wesite of the project beforehand).
Watch out to position the sim with the “cut” side versus the outer border (direction of battery) of the notebook and contact plate vs the bottom. This seems to be a bit “counterintuitive” but gives the desired result.
The device mounts under /dev/ttsy/usb1 (and not /dev//usb0 or 2). The other two positons are also claimed, maybe that part is the GPS I guess.

If that does not work, check these other posts:

https://forums.linuxmint.com/viewtopic.php?t=202158
https://forums.centos.org/viewtopic.php?t=50599

One of them may be a lead to a solution, HTH.

Best,

A.

#506 Re: Installation » xserver-xorg-video-* unused packages » 2024-06-11 17:24:35

Hello:

... xserver-xorg-video-all: "It does not provide any drivers itself, an may be removed ...

Interesting ...

Maybe installing/reinstalling what I know I need ie: xserver-xorg-video-nouveau, xserver-xorg-video-vmware and maybe some other one/s (have to figure out which) will keep them in the system (as manually installed) when I get rid of the ones I don't need.

I'll see what gives in a few days' time.

Thank you for your input.

Best,

A.

#507 Re: Installation » xserver-xorg-video-* unused packages » 2024-06-10 17:00:25

Hello:

aluma wrote:

... have a look.

But I did, I did !!!  8^)

This page has a section for the Debian 12 "Bookworm" Nvidia driver Version 525.105.17.

Consequently, that same section has a link to an Nvidia webpage labelled Appendix A. Supported NVIDIA GPU Products, this list being basically divided in two important parts:

The first part is the list of current NVIDIA GPUs supported in the unified driver and the second part is the list of all the legacy NVIDIA GPUs which are no longer supported in the unified driver.

The second part is in turn divided into separate sections, according the the different unified driver versions and their supported GPUs.

If you open that page in your browser and search for Quadro FX 580 or for the card's PCI_ID, 0659, you will only find it under the heading that reads "The 340.xx driver supports the following set of GPUs:"

The 340.xx driver supports the following set of GPUs:

NVIDIA GPU product    Device PCI ID* VDPAU features
GeForce 8800 GTX 	  0191 	           -
--- snip ---
Quadro FX 580 	          0659 	           A      ###
--- snip ---
NVS 300 	          10D8 	           C
aluma wrote:

Wiki may also be wrong ...

Of course, shit happens all the time.

But the link is to an official Nvidia web page.
They most probably got it right.

Thanks for your efforts to help.

Best,

A.

#508 Re: Installation » xserver-xorg-video-* unused packages » 2024-06-10 12:45:22

Hello:

aluma wrote:

Debian 12 "Bookworm", Version 525.105.17, supported devices Quadro FX 580.
https://wiki.debian.org/NvidiaGraphicsDrivers
In repo Devuan v.525.147...

Hmm ...
Me thinks not.

This is what I get in my up-to-date box ...

~$ uname -a
Linux devuan 6.1.0-21-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.90-1 (2024-05-03) x86_64 GNU/Linux
~$ 

... with nvidia-detect:

~$ nvidia-detect
Detected NVIDIA GPUs:
01:00.0 VGA compatible controller [0300]: NVIDIA Corporation G96CGL [Quadro FX 580] [10de:0659] (rev a1)
02:00.0 VGA compatible controller [0300]: NVIDIA Corporation G96CGL [Quadro FX 580] [10de:0659] (rev a1)

Checking card:  NVIDIA Corporation G96CGL [Quadro FX 580] (rev a1)
Your card is only supported by the 340 legacy drivers series, which is only available up to buster.

Checking card:  NVIDIA Corporation G96CGL [Quadro FX 580] (rev a1)
Your card is only supported by the 340 legacy drivers series, which is only available up to buster.

Unless I have missed something, the G96CGL/[Quadro FX 580 card (PCI ID:0659) needs the 390.xx driver but it is not available in the Bookworm/Deadalus repository.

~$ sudo lspci -nn | grep -e VGA
[sudo] password for groucho: 
01:00.0 VGA compatible controller [0300]: NVIDIA Corporation G96CGL [Quadro FX 580] [10de:0659] (rev a1)   # <-
02:00.0 VGA compatible controller [0300]: NVIDIA Corporation G96CGL [Quadro FX 580] [10de:0659] (rev a1)   # <-
~$

Thanks for your input.

Best,

A.

#509 Re: Desktop and Multimedia » [SOLVED] Xfce panel transparency not working » 2024-06-10 11:18:48

Hello:

I wrote:

From what I have read, a working compton.conf ...

Not really ...

There are a set of deprecated entries which will prevent picom from starting when you test it from the terminal, but there is a warning to that effect with instructions.

Installing picom gets you a /usr/share/doc/picom/examples/picom.sample.conf file that has to be copied to ~/.config/picom/picom.conf after creating the destination directory.

I have purged compton and installed picom with no ill effects.
And after editing the *.conf file to get rid of all the shadows, fading, partial opacity, bluring and such I now have a snappy Linux desktop.

There is still the matter of some slight tearing which did not ocurr with the Nvidia drivers, but they are in the past now.
I'll have to make do with what is available, there may be some xorg.conf magic to be done as nouveau does not seem to use one by default.

Best,

A.

#510 Installation » xserver-xorg-video-* unused packages » 2024-06-10 10:39:04

Altoid
Replies: 6

Hello:

After (more or less) getting my nouveau+compositor setup working, I set out to see about a bit of screen tearing I am still getting.
Not terrible ...

But it was not there when I used the (now unsupported) Nvidia drivers for my twin G96CGL / Quadro FX580 cards and would like to see if I can improve things.

While at that, I saw that the task-desktop package installs, besides the nouveau package, quite a few other xserver-xorg-video-* packages:

~$ apt list | grep installed | grep xorg-video
--- snip ---
xserver-xorg-video-all/stable,now 1:7.7+23 amd64 [installed,automatic]
xserver-xorg-video-amdgpu/stable,now 23.0.0-1 amd64 [installed,automatic]
xserver-xorg-video-ati/stable,now 1:19.1.0-3 amd64 [installed,automatic]
xserver-xorg-video-cirrus/stable,now 1:1.5.3-1+b5 amd64 [installed,automatic]
xserver-xorg-video-fbdev/stable,now 1:0.5.0-2 amd64 [installed,automatic]
xserver-xorg-video-nouveau/stable,now 1:1.0.17-2 amd64 [installed,automatic]
xserver-xorg-video-radeon/stable,now 1:19.1.0-3 amd64 [installed,automatic]
xserver-xorg-video-vesa/stable,now 1:2.5.0-1+b1 amd64 [installed,automatic]
xserver-xorg-video-vmware/stable,now 1:13.3.0-3.1+b1 amd64 [installed,automatic]
~$ 

All automatically installed.

Save some severe hardware problem or coming across a more interesting ie: better+free/very cheap video card alternative, I have no plans to make any changes, so the Nvidia hardware will stay where it is, works perfectly well.

That being the case, I was wondering how to safely get rid of the driver packages I am loading and will not ever load.
After all, as the sole user of this box, if I need drivers for a different video card, I will be quite aware of that.
There will be no surprises.

I asked aptitude about them all:

~$ aptitude why xserver-xorg-video-all
i   task-desktop Depends xserver-xorg-video-all
~$ 
~$ aptitude why xserver-xorg-video-amdgpu
i   task-desktop           Depends xserver-xorg-video-all   
i A xserver-xorg-video-all Depends xserver-xorg-video-amdgpu
~$ 
~$ aptitude why xserver-xorg-video-ati
i   task-desktop           Depends xserver-xorg-video-all
i A xserver-xorg-video-all Depends xserver-xorg-video-ati
~$ 
~$ aptitude why xserver-xorg-video-cirrus
i   task-desktop              Depends  xorg                                      
i A xorg                      Depends  xserver-xorg (>= 1:7.7+23)                
i A xserver-xorg              Depends  xserver-xorg-video-all | xorg-driver-video
i A xserver-xorg-video-cirrus Provides xorg-driver-video                         
~$ 
~$ aptitude why xserver-xorg-video-fbdev
i   task-desktop           Depends xserver-xorg-video-all  
i A xserver-xorg-video-all Depends xserver-xorg-video-fbdev
~$ 
~$ aptitude why xserver-xorg-video-nouveau
i   task-desktop           Depends xserver-xorg-video-all    
i A xserver-xorg-video-all Depends xserver-xorg-video-nouveau
~$ 
~$ aptitude why xserver-xorg-video-radeon
i   task-desktop           Depends xserver-xorg-video-all   
i A xserver-xorg-video-all Depends xserver-xorg-video-ati   
i A xserver-xorg-video-ati Depends xserver-xorg-video-radeon
~$ 
~$ aptitude why xserver-xorg-video-vesa
i   task-desktop           Depends xserver-xorg-video-all 
i A xserver-xorg-video-all Depends xserver-xorg-video-vesa
~$ 
~$ aptitude why xserver-xorg-video-vmware
i   task-desktop           Depends xserver-xorg-video-all   
i A xserver-xorg-video-all Depends xserver-xorg-video-vmware
~$ 

From what I can make out, xserver-xorg-video-all drags in the amdgpu, ati, fbdev, nouveau, vesa and vmware driver packages while the cirrus and radeon packages have different dependencies.

I guess that this is (?) because they are fallback/basic drivers required by xserver-xorg to make sure there is always a screen at boot time if X is being used.

Unless I have missed something and some other package needs to remain installed ...
What would the proper/safe method for removing the unused packages and keep only the nouveau package?

Thanks in advance.

Best,

A.

#511 Re: Desktop and Multimedia » [SOLVED] Xfce panel transparency not working » 2024-06-09 21:03:34

Hello:

A necessary heads up.

I wrote:

... because the repository was archived on 08/2023 and is considered to be deprecated.

It would seem that compton itself, although it has not been explicitly archived by the maintainer, would be at least in a pre-archived state as (apparently) they have had no activity there for the past five years.
ie: there are unassigned issues from as far back as 2019 with no input from them.

[rant]
Would it be too hard to at least say: "ain't doing no mo' maintaining, bitch!"
But in a timely fashion ...
[/rant]

An unanswered post post from 01/2022:

skyyysi wrote:

Since the development on compton has halted, this repository should advise to switch to picom (https://github.com/yshui/picom) at the top of the README. In addition, it should be archived to further make it clear that asking for support here is a waste of time.

This would clarify the status of the compton package with picom packages available for Ceres, Daedalus and Chimaera.

From what I have read, a working compton.conf file would work also with picom, but we'll have to see about that.

Best,

A.

#512 Re: Desktop and Multimedia » [SOLVED] Xfce panel transparency not working » 2024-06-09 12:27:08

Hello gl:

golinux wrote:

Thanks ...

You're welcome.

golinux wrote:

When I finally get to ...

Ahh ...
Good to see I was not the only one deferring a system upgrade.  8^)
Not looking forward to doing it again, I see that /usr merge looms in the horizon.

Who knows what else awaits my (ca. 2007 and perfectly working) non-UEFI box.

That said, it may be worth considering ie: packagers taking note, that the native Xfce compositor does not work properly while compton 1-1+deb12u1 does.

It is a sign of the times we live in that Xfce maintainers never bothered to fix the long standing (pre 2018) compositor issues they had and (surprise!) added native compositing. I wonder what else they will add in future versions, seeing that the 2020 row generated by pushing CSDs was not duly noted.

I did at one point try out compton-conf but did not manage to get things properly configured.
Probably because the repository was archived on 08/2023 and is considered to be deprecated.

Fortunately, I got a good result following the instructions in the link I posted, although there may still be some room for improvement.

Best,

A.

#513 Re: Desktop and Multimedia » [SOLVED] Xfce panel transparency not working » 2024-06-08 10:40:05

Hello:

Andre4freedom wrote:

...  encountered the same problem ...
... could do nothing about it.

Well ...
There seems to be (and evidently still is) some sort of problem with Xfce+compton.

Previous versions of Xfce (4.10-4.16) had a problem if compton was installed/enabled: Firefox, Thunar (and other desktop applications) would crash if you tried to drag something with the mouse pointer. eg: move around shortcuts within menus in FF, drag folders in Thunar.

Being a common issue to both applications, it was obviously not a FF or Thunar issue so I posted a question to the forum in 2018 and again in 2021.

I also posted a bug.

The questions to the forum never got a reply and the bug report got one stating that it was an Nvidia/compton/unsupported Xfce 4.10 problem, to try with  4.12 ...

The same problem persisted till Xfce 4.18 when it got "solved" by adding a native compositor.
A compositor which does not work properly.

So ...
Where is the problem?

Moot point now.
Disabling the Xfce native compositor and installing compton seems to be what is needed.

Best,

A.

#514 Re: Desktop and Multimedia » [SOLVED] Xfce panel transparency not working » 2024-06-07 23:34:25

Hello:

I wrote:

... defies the idea of panel transparency ...

Right after posting my question I remembered a package called compton that had to do with transparency.

I saw that it in was the repository but was not installed (it was installed in ASCII and Beowulf), so I disabled compositing in Xfce and installed / enabled compton.

And that was  that, problem solved.
The panel transparency works as before.

I am still getting a very faint diagonal tear I had across (upper left to lower right corners) in my centre monitor when scrolling and in every monitor when resizing windows by dragging with the mouse pointer.

I'll look into that in another thread.

Best,

A.

#515 Desktop and Multimedia » [SOLVED] Xfce panel transparency not working » 2024-06-07 23:16:25

Altoid
Replies: 7

Hello:

After finally upgrading my box from Beowulf to Chimaera and then to Daedalus and having solved problems that arose with Wine and CUPS, I am now in the process of getting the desktop working as it did before the upgrades.

My original Xfce setup took me some time to get right but it worked: from Jesse to ASCII to Beowulf and I got used to it.

But now, with Xfce 4.18 I cannot get the panel transparency working properly.
ie: as it worked in the previous version of Xfce, with the panel itself totally transparent and the launcher icons / minimised windows completely opaque.

I have followed the instructions but have not been able to get it to work as before: now both the panel background and the launcher icons / minimised windows become transparent at the same time.

Sort of defies the idea of panel transparency ... 8^/

Does anyone have the same problem?

Thanks in advance.

Best,

A.

#516 Re: Hardware & System Configuration » GPM and Synaptics touchpad features » 2024-06-06 14:52:25

Hello:

semil wrote:

... have to do any GPM configuration ...

Sorry, I misunderstood your question. ie: I thought the trackpad was not working at all.

I don't use GPM.

But from what I gather, it does need a bit of configuration to get it to work, albeit with varied results.

HTH.

Best,

A.

#517 Re: Hardware & System Configuration » GPM and Synaptics touchpad features » 2024-06-06 12:26:02

Hello:

semil wrote:

... anyone have this working?

Yes, I don't like it so I don't use it much, but it works.

This on an Asus 1000HE running Devuan Beowulf on a backported kernel:

~$ uname -a
Linux eee-dev 5.10.0-0.deb10.16-686-pae #1 SMP Debian 5.10.127-2-bpo10+1 (2022-07-28) i686 GNU/Linux

My dmesg printout says this:

~$ sudo dmesg | grep -i serio1
3 --- snip ---
4 [    6.186975] psmouse serio1: elantech: assuming hardware version 2 (with firmware version 0x020030)
5 [    6.285227] psmouse serio1: elantech: Synaptics capabilities query result 0x00, 0x02, 0x64.
6 [    6.371781] psmouse serio1: elantech: Elan sample query result 00, 03, 64
7 [    6.595103] input: ETPS/2 Elantech Touchpad as /devices/platform/i8042/serio1/input/input2
8 --- snip ---

Run sudo dmesg | grep -i synaptics, sudo dmesg | grep -i elantech and see what you get in the printout.

You may also want to consider posting (some?) information about the hardware and Devuan release you are using so that you can get more help.

Best,

A.

#518 Re: Hardware & System Configuration » Which work environment to choose for a processor two cores and four th » 2024-05-31 11:06:50

Hello:

Welcome to Dev1. 8^)

John22233 wrote:

... a computer with a processor  2015 year. It has 16 GB of RAM and an AMD processor with two cores and four threads.

Your 2015 hardware is less than 10 years old. 

I use a ca. 2007 Sun Ultra24 workstation with an Intel Core 2 Quad Q9550 processor and 8.0Gb RAM.

The system boots from a 120Gb/SATA3 SSD drive, has four different sized SAS HDDs SAS drives running from a ca. 2012 LSI SAS1068E SAS controller for storage and such and two ca. 2009 Nvidia Quadro FX580/512Mb cards driving a pair of ca. 2012 Samsung SyncMaster 940n and a ca. 2015 Dell P1914s displays.

There's also a ca. 2000 Adaptec 2940U/UW Pro PCI controller to run a ca. 1995 Umax S6-E SCSI-2 scanner and an external DDS SCSI tape drive should ever need to do such a thing.

As you can see, all of the hardware I am listing is quite dated/has been discontinued, some of it for many years now.
Yet it all works wonderfully well with Devuan Linux Daedalus.

As for the desktop, Mate or XFCE will do fine.
I think you should have no issues with your box.

Best,

A.

#519 Re: Installation » [SOLVED] Chimaera to Daedalus upgrade broke WINE » 2024-05-28 21:11:34

Hello:

aluma wrote:

... probably be right to remove my simple comments from this thread ...

I beg to differ, it would not be right.

There is absolutely no need for that, please don't remove/delete anything.

We were simply exchanging opinions in a civilised manner, albeit drifing off topic.
Nothing nothing more / nothing less.

At worse, agreeing to disagree.  8^)

I do apologise if my tone/wording made you think otherwise.

Best,

A.

#520 Re: Installation » [SOLVED] Chimaera to Daedalus upgrade broke WINE » 2024-05-28 12:44:37

Hello:

This thread is/was about the Daedalus upgrade breaking my Wine installation but is drifting in another direction.
Taking that into account and having been labelled as [solved], I'll say no more than this:

I wrote:

... Richard Harris' Pegasus Mail.

Like I wrote: not any Win based email client.

wikipedia wrote:

... described as "one of the web's oldest and most respected email clients".

https://en.wikipedia.org/wiki/Pegasus_Mail#cite_note-4

Martin von Wittich wrote:

Do not rely on Wine to magically protect you ...

Who could think this was even a remote possibility?

I have would not ever rely on Wine to protect me from malware or anything else.
I protect my system from malware by, first and foremost, not running Wine as root.
ie: release notes and FAQs are an important part of the documentation.

wiki.winehq wrote:

... NEVER run Wine as root! Doing so gives Windows programs (and viruses) full access to your computer and every piece of media attached to it.

https://wiki.winehq.org/FAQ#Should_I_run_Wine_as_root?

In all the years (~30) I have used Pegasus Mail, I have never (ever) had a problem with malware coming in via email.
This in both MS and Linux based systems.

Of course, and it goes without saying, I have never opened suspicious email attachments.
Nor did Pegasus Mail automatically/suicidally open them for me.  8^)

aluma wrote:

WINE itself takes up about the same amount of memory...

Not something I took or ever needed to take into consideration.

As a Linux user, my need was to find a Linux based email client that could match what Pegasus Mail offered.
I was never able to find such a thing, much less in Thunderbird which I tried many years ago and found to be lacking.

For the reasons given in my previous post and in spite of it being MS based and not open source, I have continued to use it in every Linux distribution I have gone through, without any issues.

Like I said, YMMV.

Thank you for your input.

Best,

A.

#521 Re: Hardware & System Configuration » CUPS 2.4.2 - Daedalus - "Filter failed" error » 2024-05-27 11:55:56

Hello:

I wrote:

Installing the libcupsimage2 instantly solved the problem.

I see that the CUPS package is maintained by the 'Debian Printing Team'.

This issue with (what seems to be) a dependency problem with libcupsimage2 ...
Who would have to take care of it? ie: CUPS, Debian, Devuan?

My uneducated guess would be that package dependencies are something that involves the developers.
ie: the list of dependencies is generated somewhere along the line and added to the source code so that when the package is compiled and 'made' by maintainers, the binary will install them all with the proper versions.
ie: maintainers would not add dependencies on their own.

Is this correct?
If not, who would have to take care of the problem with libcupsimage2?

~$ aptitude why libcupsimage2
i   cups                 Suggests   foomatic-db-compressed-ppds | foomatic-db
p   foomatic-db          Recommends printer-driver-all                       
p   printer-driver-all   Recommends printer-driver-splix                     
p   printer-driver-splix Depends    libcupsimage2 (>= 1.4.0)                 
~$ 

From what I can make of the above aptitude printout, cups suggests foomatic-db / foomatic-db-compressed-ppds which in turn recommends printer-driver-all.

In turn, printer-driver-all recommends printer-driver-splix but this last package depends on libcupsimage2.

Now, my system does not have the suggested packages foomatic-db-compressed-ppds / foomatic-db installed because I do not install suggested packages because of their suggested nature.
ie: they are not dependencies.

But this being so ended screwing up my printing because the cups-filters package will fail if the libcupsimage2 is not installed.

It seems that cups-filters depends on libcupsimage2 being present to work properly.
But there is no mention of that here.

I'd appreciate some input on this.

Thanks in advance.

Best,

A.

#522 Re: Installation » [SOLVED] Chimaera to Daedalus upgrade broke WINE » 2024-05-26 16:55:24

Hello:

chris2be8 wrote:

... ls -lR /home/groucho/.wine where the R is to list subdirectories recursively.

The dosdevices directory contains a list of all the drives Wine would access.
And the drive_c directory is not really related to it.

It has all the Windows directories it needs plus the directories for the MS based applications installed, in this case, Pegasus Mail.

~/.wine$ cat wineserver
wine-1X5UO
~/.wine$
~/.wine$ cat wrapper.cfg
ww_name="All fonts"
~/.wine$ 

Best,

A.

#523 Hardware & System Configuration » CUPS 2.4.2 - Daedalus - "Filter failed" error » 2024-05-26 15:20:05

Altoid
Replies: 1

Hello:

After upgrading to Daedalus, I came across a severe problem with CUPS 2.4.2.
When I tried to print something, it failed with this output in dmesg:

--- snip ---
~$ sudo dmesg
usb 5-1: new high-speed USB device number 10 using ehci-pci
usb 5-1: New USB device found, idVendor=04e8, idProduct=3321, 
bcdDevice= 1.00
[usb 5-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
usb 5-1: Product: M2020 Series
usb 5-1: Manufacturer: Samsung Electronics Co., Ltd.
usb 5-1: SerialNumber: 0731B8GJBB029TD
usblp 5-1:1.0: usblp1: USB Bidirectional printer dev 10 if 0 alt 0 
proto 2 vid 0x04E8 pid 0x3321
usblp1: removed       # last line
~$

The last line appears after sending the file to print.
The printer (Samsung M2020W USB) was correctly detected at boot time:

~$ sudo dmesg
--- snip ---
[    6.482475] usb 7-1: New USB device found, idVendor=04e8, 
idProduct=3321, bcdDevice= 1.00
[    6.491763] uhci_hcd 0000:00:1d.2: detected 2 ports
[    6.500910] usb 7-1: New USB device strings: Mfr=1, Product=2, 
SerialNumber=3
[    6.500913] usb 7-1: Product: M2020 Series
[    6.500914] usb 7-1: Manufacturer: Samsung Electronics Co., Ltd.
[    6.500916] usb 7-1: SerialNumber: 0731B8GJBB029TD
--- snip ---
~$

The CUPS UI informed "Filter failed".

The problem (solution found here) is caused by a missing package: libcupsimage2, which for some reason is not listed as a dependency for CUPS (!)

~$ aptitude why libcupsimage2
i   cups                 Suggests   foomatic-db-compressed-ppds | foomatic-db
p   foomatic-db          Recommends printer-driver-all                       
p   printer-driver-all   Recommends printer-driver-splix                     
p   printer-driver-splix Depends    libcupsimage2 (>= 1.4.0)                 
~$ 

Installing the libcupsimage2 instantly solved the problem.

Best.

A.

#524 Re: Installation » [SOLVED] Chimaera to Daedalus upgrade broke WINE » 2024-05-26 11:48:52

Hello:

aluma wrote:

... wouldn’t use the Win mail client under Linux ...

I get the idea, of course, but we are not talking about any Win based email client.

This is Richard Harris' Pegasus Mail.

I have used Pegasus Mail since my early W3.10 days, cannot recall the exact year but I am quite sure it was pre W95 by at least a year which makes for ~30 years' continued use without ever giving me any grief or problems that I could not fix myself.

Meaning PMail under W3.XX straight through W95/W98/W2000/XP -> a number of Linux distributions and now Devuan, from Jessie to Daedalus.

Never lost any emails and when I did, it was because of the well known 'between the chair and the keyboard' problem.

I started using PMail after trying out all the email clients available (at the time) for the MS environment but not one came close to PMail.
ie: Eudora, OE, Netscape (?) and a couple of others I cannot recall.

It has a very useful 'selective mail download' feature, something I have not seen in any other email client.
That plus an easy to use UI that has not changed significantly in all these years has made it my email client of choice, even under Linux.

More than enough for me to use and continue using it, without any reservations.
Even if it is closed source.

Did I mention that it is free?
Of course, YMMV.

Thanks a lot for your input.

Best,

A.

#525 Re: Installation » [SOLVED] Chimaera to Daedalus upgrade broke WINE » 2024-05-25 17:22:39

Hello:

chris2be8 wrote:

... a bit late now ...

Never too late ... 
Or so one hopes.

chris2be8 wrote:

... but did you ...

I did, just in case.

Non working .wine saved elsewhere.

~$ ls -lr /media/stuff/.wine
total 2956
-r-------- 1 groucho groucho      11 Nov  3  2020 wineserver              ### <-
-rw-r--r-- 1 groucho groucho    3314 May 25 07:16 userdef.reg
-rw-r--r-- 1 groucho groucho   60783 May 25 07:16 user.reg
-rw-r--r-- 1 groucho groucho 2946462 May 25 07:16 system.reg
drwxr-xr-x 8 groucho groucho    4096 Feb 18  2022 drive_c
drwxr-xr-x 2 groucho groucho    4096 May 25 07:16 dosdevices
~$ 

Previously/now working .wine from BackInTime snapshot.

~$ ls -lr /home/groucho/.wine #(*.reg backup files removed for clarity)
total 6020
-rw-r--r-- 1 groucho groucho      20 May 25 10:13 wrapper.cfg
-rw------- 1 groucho groucho      11 Nov  3  2020 wineserver             ### <-
-rw-r--r-- 1 groucho groucho    3314 May 25 13:45 userdef.reg
-rw-r--r-- 1 groucho groucho   61038 May 25 13:45 user.reg
-rw-r--r-- 1 groucho groucho 3065505 May 25 13:45 system.reg
drwxr-xr-x 8 groucho groucho    4096 Feb 18  2022 drive_c
drwxr-xr-x 2 groucho groucho    4096 May 25 13:44 dosdevices
~$ 

As you can see the non-working directory is missing the wrapper.cfg file and the wineserver files do not have matching attributes.
ie: -r-------- in the non-working version but -rw------- in the working version.   
Not too clear to me how that got to be so.

That said, check the timestamp of the non-working version of the wineserver file: 20201103 -> four years ago.
Same as the working version. Attribute changes do not set a new timestamp?

No matter, problem solved.

The moral of the tale?
If you have a decent back-up at hand, you also have a chance to fix whatever went awry. 

Thank you very much for your input.

Best,

A.

Board footer

Forum Software