Re: [PLUG] Java errors

[John Jason Jordan]

Didn't make any difference. The net came up with other possibilities
(swing?), but so far I haven't hit on one that makes any difference. The
window size is also a bit too big, but my attempts there have also
failed.

Michael Ewan  dijo:
>Try removing the -Dsun.java2d.uiScale=3 option and let the JVM choose.

>On Fri, Oct 6, 2023 at 1:57 AM John Jason Jordan 
>wrote:
>
>> SUCCESS!!
>>
>> This command did the job:
>>
>> PATH=/usr/java/jre1.8.0_381/bin/
>>
>> followed by the launch script.
>>
>> The above path is for Oracle Java 8 which I downloaded and installed
>> manually myself. I could probably just as easily have used openjdk17
>> or Oracle 17, both of which are also installed, but with Oracle 8
>> JRE I knew where it was and with the other two I'd have to figure it
>> out.
>>
>> Now my only question is, does that PATH command create a permanent
>> path, or will it go away, like after a reboot. Never mind. I just
>> added it to the launch script, which now says:
>>
>> #!/bin/bash
>> cd /home/jjj/Software/TreeForm103/TreeForm/
>> PATH=/usr/java/jre1.8.0_381/bin/
>> java -Dsun.java2d.uiScale=3 -Xmx256m -Xms64m -jar
>> TreeForm.jar
>>
>> The program runs perfectly, except that the fonts are about 1mm. I
>> can still use the program because I know what the text says, even if
>> I can no longer really read it. I added the
>> '-Dsun.java2d.uiScale=3,' which was suggested to me as a way to
>> increase the size of fonts, but in this case it had no effect. The
>> two -Xm... comments were in the original launch scrip written by the
>> developer. I don't know what they do, but leaving them out makes no
>> difference.
>>
>> If anyone can suggest alternatives for the -Dsun... option to
>> increase font sizes in Java applications, I'd be happy to do some
>> experimenting.
>>
>>
>> On Mon, 2 Oct 2023 13:35:05 -0700
>> John Jason Jordan  dijo:
>>  
>> >   Exception in thread "main" java.lang.NoClassDefFoundError:
>> >   sun.misc.Launcher at
>> >   userInterface.Start.(Start.java:54)
>> >
>> >I get the same error message whether I run from the launch script
>> >with its two options or straight from the command line without them.
>> >
>> >The only way I know how to tell what versions of Java are installed
>> >is to launch LibreOffice Writer and go to Tools > Options >
>> >Advanced, where it displays the versions of Java that LO found. It
>> >now lists Oracle 8, Oracle 17, and Debian 17, where it used to
>> >include Oracle 21 as well. There's probably a faster, simpler way
>> >to get a list of installed Javas from the command line, but I don't
>> >know it. For testing it might also be useful if I could specify
>> >which of the installed Javas the app is supposed to use. As it is,
>> >it just picks one based on who knows what criteria.
>> >
>> >In Xubuntu it appears that the installed Javas include
>> >JB-Java-jdk8.d. On that the net leads me to openjdk8, and more
>> >links. The TreeForm app there just runs, and I can't tell what Java
>> >it's actually using.
>> >
>> >
>> >On Sun, 1 Oct 2023 10:32:59 -0700
>> >Michael Ewan  dijo:
>> >  
>> >>Unfortunately it sounds like there is a static string in the Java
>> >>source or possibly in the jar file.
>> >>You can disassemble the jar file with the "jar" command from the
>> >>Java runtime.
>> >>
>> >>
>> >>On Sat, Sep 30, 2023 at 6:31 PM John Jason Jordan 
>> >>wrote:
>> >>  
>> >>> I opened the launch script in a text editor and all it says is
>> >>>
>> >>> cd /home/jjj/Software/TreeForm103/TreeForm/
>> >>> java -Xmx256m -Xms64m -jar TreeForm.jar
>> >>>
>> >>> With the command line in ~/Software/TreeForm103/TreeForm I just
>> >>> copied the above line and ran it, and I got the same error
>> >>> message. Then I deleted the -Xm... options and ran it bare, and
>> >>> again I got the same error message. It might be useful to know
>> >>> what those options do. Window size, maybe? In any event, they
>> >>> don't seem related to the error message.
>> >>>
>> >>> Checking in Synaptic I have had OpenJDK installed the whole
>> >>> time. Is there a way to tell the java -jar command which java to
&g

Re: [PLUG] Java errors

[John Jason Jordan]

SUCCESS!!

This command did the job:

PATH=/usr/java/jre1.8.0_381/bin/

followed by the launch script.

The above path is for Oracle Java 8 which I downloaded and installed
manually myself. I could probably just as easily have used openjdk17 or
Oracle 17, both of which are also installed, but with Oracle 8 JRE I
knew where it was and with the other two I'd have to figure it out.

Now my only question is, does that PATH command create a permanent
path, or will it go away, like after a reboot. Never mind. I just added
it to the launch script, which now says:

#!/bin/bash
cd /home/jjj/Software/TreeForm103/TreeForm/
PATH=/usr/java/jre1.8.0_381/bin/
java -Dsun.java2d.uiScale=3 -Xmx256m -Xms64m -jar TreeForm.jar

The program runs perfectly, except that the fonts are about 1mm. I can
still use the program because I know what the text says, even if I can
no longer really read it. I added the '-Dsun.java2d.uiScale=3,' which
was suggested to me as a way to increase the size of fonts, but in this
case it had no effect. The two -Xm... comments were in the original
launch scrip written by the developer. I don't know what they do, but
leaving them out makes no difference.

If anyone can suggest alternatives for the -Dsun... option to increase
font sizes in Java applications, I'd be happy to do some experimenting.


On Mon, 2 Oct 2023 13:35:05 -0700
John Jason Jordan  dijo:

>   Exception in thread "main" java.lang.NoClassDefFoundError:
>   sun.misc.Launcher at
>   userInterface.Start.(Start.java:54)
>
>I get the same error message whether I run from the launch script with
>its two options or straight from the command line without them. 
>
>The only way I know how to tell what versions of Java are installed is
>to launch LibreOffice Writer and go to Tools > Options > Advanced,
>where it displays the versions of Java that LO found. It now lists
>Oracle 8, Oracle 17, and Debian 17, where it used to include Oracle 21
>as well. There's probably a faster, simpler way to get a list of
>installed Javas from the command line, but I don't know it. For testing
>it might also be useful if I could specify which of the installed Javas
>the app is supposed to use. As it is, it just picks one based on who
>knows what criteria.
>
>In Xubuntu it appears that the installed Javas include JB-Java-jdk8.d.
>On that the net leads me to openjdk8, and more links. The TreeForm app
>there just runs, and I can't tell what Java it's actually using.
>
>
>On Sun, 1 Oct 2023 10:32:59 -0700
>Michael Ewan  dijo:
>
>>Unfortunately it sounds like there is a static string in the Java
>>source or possibly in the jar file.
>>You can disassemble the jar file with the "jar" command from the Java
>>runtime.
>>
>>
>>On Sat, Sep 30, 2023 at 6:31 PM John Jason Jordan 
>>wrote:
>>  
>>> I opened the launch script in a text editor and all it says is
>>>
>>> cd /home/jjj/Software/TreeForm103/TreeForm/
>>> java -Xmx256m -Xms64m -jar TreeForm.jar
>>>
>>> With the command line in ~/Software/TreeForm103/TreeForm I just
>>> copied the above line and ran it, and I got the same error message.
>>> Then I deleted the -Xm... options and ran it bare, and again I got
>>> the same error message. It might be useful to know what those
>>> options do. Window size, maybe? In any event, they don't seem
>>> related to the error message.
>>>
>>> Checking in Synaptic I have had OpenJDK installed the whole time. Is
>>> there a way to tell the java -jar command which java to use? As for
>>> paths, do you mean the path to the java? I know where the Oracle
>>> Java 8 is located, because there were 'install' instructions telling
>>> me to put it in /usr/java/jre1.8.0_381, so that's where I put it. It
>>> must be correct because LibreOffice found it and listed it as
>>> available. I don't know where the others are.
>>>
>>> There is a Help file with TreeForm, which says at the top:
>>>
>>> To install this software in your computer, unzip the folder
>>> in any folder of your choosing.
>>> This software requires Java 1.4.2 or higher to run. The Java
>>> runtime can be downloaded from the Java
>>> Website.
>>>
>>> Michael Ewan  dijo:
>>> >Your start script may be calling for a specific path rather than a
>>> >relative path in your JRE.
>>> >Also try using OpenJDK instead of Oracle Java. Do a text search in
>>> >your source code for that path.
>>>
>>> >>$ ./TreeForm_launch_script
>>> >>Error: could not open
>>> >> `/usr/lib/jvm/jdk-21-oracle-x64/lib/jvm.cfg'
>>>
>>>

Re: [PLUG] Move cron jobs to new computer

[John Jason Jordan]

On Fri, 29 Sep 2023 16:22:47 -0700
John Jason Jordan  dijo:

>I'm just going to have to muddle through and rewrite them.

When I sat down at the computer this morning the gxmessage window
was in the middle of the screen announcing that rsync ran at 3am and at
4am to make mirrors of / and ~/ with exit code 0. I had put a dummy
file on both partitions before going to bed, so I checked and it was in
both mirrors. Problem solved. Except that I still haven't figured out
where cron saves the files. At least now I have the sense to copy them
to a text file in ~/.

Re: [PLUG] Java errors

[John Jason Jordan]

This little app is currently running fine on Xubuntu 22.04.3, as it
always has, but the Java versions installed there are somewhat
different. I decided to experiment, and for my fist venture I
uninstalled Oracle JDK 21. Afterwards the error message changed:

Exception in thread "main" java.lang.NoClassDefFoundError:
sun.misc.Launcher at userInterface.Start.(Start.java:54)

I get the same error message whether I run from the launch script with
its two options or straight from the command line without them. 

The only way I know how to tell what versions of Java are installed is
to launch LibreOffice Writer and go to Tools > Options > Advanced,
where it displays the versions of Java that LO found. It now lists
Oracle 8, Oracle 17, and Debian 17, where it used to include Oracle 21
as well. There's probably a faster, simpler way to get a list of
installed Javas from the command line, but I don't know it. For testing
it might also be useful if I could specify which of the installed Javas
the app is supposed to use. As it is, it just picks one based on who
knows what criteria.

In Xubuntu it appears that the installed Javas include JB-Java-jdk8.d.
On that the net leads me to openjdk8, and more links. The TreeForm app
there just runs, and I can't tell what Java it's actually using.


On Sun, 1 Oct 2023 10:32:59 -0700
Michael Ewan  dijo:

>Unfortunately it sounds like there is a static string in the Java
>source or possibly in the jar file.
>You can disassemble the jar file with the "jar" command from the Java
>runtime.
>
>
>On Sat, Sep 30, 2023 at 6:31 PM John Jason Jordan 
>wrote:
>
>> I opened the launch script in a text editor and all it says is
>>
>> cd /home/jjj/Software/TreeForm103/TreeForm/
>> java -Xmx256m -Xms64m -jar TreeForm.jar
>>
>> With the command line in ~/Software/TreeForm103/TreeForm I just
>> copied the above line and ran it, and I got the same error message.
>> Then I deleted the -Xm... options and ran it bare, and again I got
>> the same error message. It might be useful to know what those
>> options do. Window size, maybe? In any event, they don't seem
>> related to the error message.
>>
>> Checking in Synaptic I have had OpenJDK installed the whole time. Is
>> there a way to tell the java -jar command which java to use? As for
>> paths, do you mean the path to the java? I know where the Oracle
>> Java 8 is located, because there were 'install' instructions telling
>> me to put it in /usr/java/jre1.8.0_381, so that's where I put it. It
>> must be correct because LibreOffice found it and listed it as
>> available. I don't know where the others are.
>>
>> There is a Help file with TreeForm, which says at the top:
>>
>> To install this software in your computer, unzip the folder
>> in any folder of your choosing.
>> This software requires Java 1.4.2 or higher to run. The Java
>> runtime can be downloaded from the Java Website.
>>
>> Michael Ewan  dijo:  
>> >Your start script may be calling for a specific path rather than a
>> >relative path in your JRE.
>> >Also try using OpenJDK instead of Oracle Java. Do a text search in
>> >your source code for that path.  
>>  
>> >>$ ./TreeForm_launch_script
>> >>Error: could not open
>> >> `/usr/lib/jvm/jdk-21-oracle-x64/lib/jvm.cfg'  
>>
>>  

Re: [PLUG] Java errors

[John Jason Jordan]

I opened the launch script in a text editor and all it says is

cd /home/jjj/Software/TreeForm103/TreeForm/
java -Xmx256m -Xms64m -jar TreeForm.jar

With the command line in ~/Software/TreeForm103/TreeForm I just copied
the above line and ran it, and I got the same error message. Then I
deleted the -Xm... options and ran it bare, and again I got the same
error message. It might be useful to know what those options do. Window
size, maybe? In any event, they don't seem related to the error message.

Checking in Synaptic I have had OpenJDK installed the whole time. Is
there a way to tell the java -jar command which java to use? As for
paths, do you mean the path to the java? I know where the Oracle Java 8
is located, because there were 'install' instructions telling me to put
it in /usr/java/jre1.8.0_381, so that's where I put it. It must be
correct because LibreOffice found it and listed it as available. I don't
know where the others are.

There is a Help file with TreeForm, which says at the top:

To install this software in your computer, unzip the folder in
any folder of your choosing.
This software requires Java 1.4.2 or higher to run. The Java
runtime can be downloaded from the Java Website.

Michael Ewan  dijo:
>Your start script may be calling for a specific path rather than a
>relative path in your JRE.
>Also try using OpenJDK instead of Oracle Java. Do a text search in your
>source code for that path.

>>$ ./TreeForm_launch_script
>>Error: could not open `/usr/lib/jvm/jdk-21-oracle-x64/lib/jvm.cfg'

[PLUG] Java errors

[John Jason Jordan]

I have a little application for drawing phrase structure trees that was
written in Java, in roughly 2013. The author included a launch script,
and it runs fine on Xubuntu 22.04.3, but on Sparky Linux I get:

   $ ./TreeForm_launch_script
   Error: could not open `/usr/lib/jvm/jdk-21-oracle-x64/lib/jvm.cfg'

There is a document, also written in 2013, stating that it needed JRE
1.42 or later. Such is not available for SparkyLinux, but it wasn't
available in Xubuntu either. The default in SparkyLinux is Debian 17,
and I installed jre-8u381-linux-x64 and both jdk17 and jdk21, all
downloaded from Oracle. I assume they are 'installed' correctly,
because LibreOffice needs a JRE for some functions, and it has a
utility to choose which java it is supposed to use; and all four appear
in the list of available javas.

I have no idea what the above error message means, and less than that
about how to fix it. I know zero about such matters. Can someone hit me
over the head with a clue?

Re: [PLUG] Move cron jobs to new computer

[John Jason Jordan]

On Fri, 29 Sep 2023 14:57:35 -0700 (PDT)
Rich Shepard  dijo:

>On Fri, 29 Sep 2023, John Jason Jordan wrote:
>
>> Well, that didn't come out well. In Xubuntu I did 'sudo crontab -e,'

>When you want to only look at the crontab contents use `crontab -l' to
>list them.

I did that as well, and as jjj and as root, and all I found was the T-
Mobile reminder.

>If you know the file names, or part of them you can do either `locate
>' or `sudo find / -name '.

Names of what? I still haven't found where cron is storing the jobs.
The cron man page is silent on the subject, and online I haven't found
it mentioned yet. It's not /var/spool/cron; empty on both systems. And
it's not in /etc/cron.d or any of the /etc/cron. folders, because
all they have is the standard default folders and files, all with old
timestamps.

Several GUI cron alternatives are mentioned on the net, but none seem
to be in the Debian or Xubuntu repositories. I also haven't found a cron
tutorial that is useful. I don't need instructions, I just need several
pages of examples that I can crib from. Web 'tutorials' mostly just copy
and paste man cron, which has zero examples.

I'm just going to have to muddle through and rewrite them.

Re: [PLUG] Move cron jobs to new computer

[John Jason Jordan]

>On Thu, 28 Sep 2023 20:58:41 -0700
>wes  dijo:
>
>>On Thu, Sep 28, 2023 at 6:21 PM John Jason Jordan 
>>wrote:
>>> >I don't know where those distros keep cron files.

>>they are kept in /var/spool/cron.
>>But, look first at  
>>these are different from crontab. each has its own strengths and
>>weaknesses.  

>There is a /var/spool/crontabs folder in Xubuntu's /, but it is empty.
>And I read about anacron and a couple more, but what I did before
>worked well. I have it on tomorrow's to-do list to reboot to Xubuntu,
>go to crontab -e, copy my jobs to a text file, and get back to Sparky.
>In theory that's all I need to do.

Well, that didn't come out well. In Xubuntu I did 'sudo crontab -e,'
and it had one small job: a monthly reminder to pay my T-Mobile bill,
The daily jobs to back up /home and / were nowhere to be seen. There was
also nothing in anacron. Yet, in Xubuntu /home and / were getting
backed up every night, and there was another daily cron job to pop up a
message with the date and rsync exit code. In Xubuntu every morning that
message was on the middle of my screen. But right now I have no idea
what was making the rsync scripts run in Xubuntu.

Sparky's crontab is empty, so I can paste in the T-Mobile job from
Xubuntu, but it looks like I'll have to rewrite the two backup jobs and
the job to pop up the rsync exit codes. The scripts are here and they
run fine when I run them manually. I remember writing these jobs in
Xubuntu's crontab a couple years ago. What happened to them? And if the
jobs are no longer in Xubuntru's crontab, why are they still running? 

Re: [PLUG] Move cron jobs to new computer

[John Jason Jordan]

On Thu, 28 Sep 2023 20:58:41 -0700
wes  dijo:

>On Thu, Sep 28, 2023 at 6:21 PM John Jason Jordan 
>wrote:
>
>>  
>> >I don't know where those distros keep cron files.  
>
>
>they are kept in /var/spool/cron.
>
>But, look first at
>> >/etc/; there should be a set of cron directories there; e.g.,
>> >cron.d/, cron.daily/, cron.hourly/, cron.monthly/, cron.weekly/.  
>>
>>  
>these are different from crontab. each has its own strengths and
>weaknesses.

There is a /var/spool/crontabs folder in Xubuntu's /, but it is empty.
And I read about anacron and a couple more, but what I did before
worked well. I have it on tomorrow's to-do list to reboot to Xubuntu,
go to crontab -e, copy my jobs to a text file, and get back to Sparky.
In theory that's all I need to do.

It's taken about a week, but I'm almost done. And I'm liking Sparky
more than Xubuntu. And there's nothing snapping at me or flat packages
to deal with.

Re: [PLUG] Move cron jobs to new computer

[John Jason Jordan]

On Thu, 28 Sep 2023 16:13:28 -0700 (PDT)
Rich Shepard  dijo:

>I don't know where those distros keep cron files. But, look first at
>/etc/; there should be a set of cron directories there; e.g., cron.d/,
>cron.daily/, cron.hourly/, cron.monthly/, cron.weekly/.
>
>Second, use the `slocate' command, a.k.a. `whois'. (`locate' should
>also work); e.g., `locate cron' will show all files with cron in their
>names.

Yeah, both distros have /etc/cron/daily folders, which is where
both cron jobs should be; they ran daily only. Most of the files are
identical between the two distros. None have my name on them or
anything that looks like it might be something that I produced. On
Xubuntu I originally had 20.04, then 22.04, so if the timestamp on a
file is before 2020 it couldn't be mine. But here's something that
makes no sense: In the cron/daily folder for Xubuntu there is a file
'brave-browser ... 9.27.2023.' WTH? The last time I booted to Xubuntu
was a week ago, for only a couple of hours, and the OS hasn't been run
since. My gast is flabbered.

The cron jobs were just one-liners in crontab, so an easier way to deal
with this would be to boot to Xubuntu, run crontab -e, copy the jobs to
a small text file somewhere accessible, then reboot to Sparky and paste
them into Sparky's crontab-e.

[PLUG] Move cron jobs to new computer

[John Jason Jordan]

Several years ago I created a couple shell scripts to make rsync create
mirrors of / and /home on another drive installed in the computer.
Although it took awhile I created cron jobs to run the scripts every
day at 3am and 4am.

My efforts worked perfectly for years, but now I'm migrating from
Xubuntu 22.04.3 to SparkyLinux 7.0, and I'd like to migrate cron jobs
in the process. SparkyLinux automatically mounts / and /home from
Xubuntu, so I can just copy config files from the old OS to the new OS.
However, I can't figure out where cron on Xubuntu kept the jobs.
They're in some config file somewhere on those partitions, but where
they are is a mystery. Can someone give me a clue?

Re: [PLUG] Mint, Ubuntu, Debian (and Centos and Redhat/IBM)

[John Jason Jordan]

On Mon, 11 Sep 2023 02:13:25 +
MC_Sequoia  dijo:

>If you want a default lightweight Debian distro, I'd recommend either
>BunsenLabs or Crunchbang++.
>
>Crunchbang is Debian 12, Bullseye w. Linux Kernal 6.1 Long Term
>Support.
>
>Both offer an installation script so the user can decide which
>packages to install based on the feature support wanted and/or use
>case for the computer.

Very interesting. I've never tried either one. I'm currently in the
middle of getting SparkyLinux set up on my main computer. No snaps or
flatpaks, although both are in the repository if you want to install
them.

Re: [PLUG] Trying to install Debian 12 as dual boot with Ubuntu

[John Jason Jordan]

On Wed, 20 Sep 2023 13:48:39 -0700
John Jason Jordan  dijo:

>I'm ready for install number 11, but at the end where it sets up grub I
>need someone sitting behind me to tell me how, because it is clear that
>I do not know how.

My 11th effort also failed, but I finally got #12 to boot. It installed
Grub to the / partition of the new drive so it would boot Debian, left
the Xubuntu Grub alone on the old drive, and installed no Grub dual
boot menus. I tried using the utility in the BIOS (hit F12 after the
BIOS screen) that lists the various drives, and lets me choose which
one I want to boot to. I had tried this with previous installs and all
I got was a black screen with a cursor-like underscore flashing in the
upper left corner. There was no mouse and the keyboard was dead, so
reboot and start over.

But after install #12, with no Grub dual boot menu, my F12 maneuver
worked; it booted straight to Debian. Before continuing I shut down
Debian and rebooted, this time choosing the Xubuntu drive, and Xubuntu
came up. Yay! Who needs a Grub dual boot menu?

Then I spent about eight hours configuring and installing in Debian,
when I suddenly realized I needed my account number for Mullvad, my
VPN, and the only way to get it was from the utility as it was running
under Xubuntu, so I shut down Debian and booted to Xubuntu. When I had
the account number I shut down Xubuntu and rebooted Debian. It came up,
but most of my hours of work were gone. Most of the applications were
still there, including Brave browser, the last thing I installed before
shutting down. But the screen was plastered with about 40 icons, each
for one of the tabs in Brave browser, and all my work at fixing font
sizes, getting rid of default artwork, configuring apps, and more, was
lost. I rebooted to Recovery mode, but there was nothing to be done. If
I'm to get it working I'm going to have to make reinstallation #13 and
start over from scratch. Eight hours down the drain.

I wonder if it's possible to install Ubuntu and leave out the snapping
turtles and flat things.

[PLUG] Trying to install Debian 12 as dual boot with Ubuntu

[John Jason Jordan]

So far I've done this ten times. Only once was I actually able to boot
to Debian. And sitting in front of my new Debian I spent several hours
installing and configuring, which included deleting all the hundreds of
Thai fonts (Why?), which ended with Synaptic deleting half the OS. It
wouldn't run, nor would it reboot, so back to the drawing board.

This morning I followed some suggestions from a user on the Debian
forums which ended with sudo update-grub. That command edited
grub.cfg in the new Debian / folder as well as grub.cfg in the Xubuntu
/ folder, and now I no longer even get the (useless) menu; it just boots
straight to Xubuntu.

I'm ready for install number 11, but at the end where it sets up grub I
need someone sitting behind me to tell me how, because it is clear that
I do not know how.

Re: [PLUG] Change hard drive FS

[John Jason Jordan]

I'm back in Ubuntu, following my ninth failed attempt to install Debian
12. But I should say at the beginning that, like Rich, I've used ext4
for many years and have never had a problem. My issues with installing
Debian 12 is that it won't boot because of a failure in setting up Grub.
And the root cause of that is because I am too dumb to properly follow
the instructions in the Debian installer. Xubuntu is on a Samsung
M.2 drive of 1TB, and I'm trying to install Debian on a new
Samsung M.2 drive of 2TB. I can easily tell which is which when they
give me the Samsung product name, but while setting up Grub it asks
'Install grub to your primary drive?' Well, which drive is that? Both
drives have a primary partition for / and a second logical partition
for /home. 

It probably doesn't matter. I've answered that question 'yes,' and
'no,' and either way the installed Debian appears in the Grub list, but
it still won't boot. I'd rather have it on the new 2TB drive rather
than the three-year-old 1TB drive, but that probably also doesn't
matter. 

And speaking of primary partitions, the Debian installer set the
boot flag for / on the new drive, but the / partition on the old drive
(which unfailingly boots Ubuntu) has no boot flag, at least according to
GParted. More stuff I don't understand.


Michael Ewan  dijo:

>I am glad you have not had any problems.  I have had the opposite
>experience with ext4 but never a problem with xfs, hence my suggestion.
>
>On Tue, Sep 19, 2023 at 1:25 PM Rich Shepard 
>wrote:
>
>> On Tue, 19 Sep 2023, Michael Ewan wrote:
>>  
>> > You will ultimately have problems with a corrupted file system
>> > with ext4, almost guaranteed.  Xfs is a much more robust file
>> > system but if you do  
>> not  
>> > trust it, then try zfs or btrfs.  
>>
>> Michael,
>>
>> I've used ext2, ext3, and ext4 with no issues on any of them. I'll
>> stay with
>> what's worked flawlessly with me since 1997.
>>
>> Thanks,
>>
>> Rich
>>
>>  

Re: [PLUG] Can't boot to new dual boot

[John Jason Jordan]

On Mon, 18 Sep 2023 12:26:22 -0500
Bill Barry  dijo:

>> This would probably be easy to fix, if I only knew how. I'm anxiously
>> looking for suggestions so I can look at Debian 12 on the new drive.

>I had a similar problem after I did an update yesterday. Only one of
>my systems was bootable. It turned out there was a line in  the file
>/etc/default/grub
>like this
>GRUB_DISABLE_OS_PROBER=false
>which was commented out with a # in front of it.
>Uncommenting it then running update-grub fixed the problem.

At this point I'm pretty disgusted with the whole install process. As I
write this I have started over and reinstalled Debian 12 eight times.
Once I actually got the dual boot working so I could easily boot into
Debian 12 or Xubuntu 22.04.3. I spent several hours in Debian fixing
everything, except HDHomeRun wouldn't even launch, so I left that
problem for later. I was in the middle of uninstalling the 172 Thai
fonts (Why?), when Synaptic screwed me up. I had a bunch of fonts
selected to remove, and when I hit the Apply button I watched in horror
as Synaptic spent 15 minutes removing half the OS. Synaptic needs an
'Undo' button. The result was so bad that it wouldn't even boot. I had
to do a fresh install and start over from scratch. After several more
installations that ended in non-functional boot loaders, I decided to
take MX Linux for a spin.

I downloaded the MX Linux ISO and burned it to a flash drive, then
booted to it. I spent an hour trying to get it installed, but I have a
4K screen and the default fonts are too tiny to read without a
magnifying glass. In the desktop GUI I changed things so it looked
reasonable, but when I clicked on the Install button it popped up a
tiny window, complete with unreadable fonts. Evidently the Install
utility bypasses the desktop and uses base settings that can't be
changed. I might have been able to install it, but MX Linux wasn't
worth the effort of spending hours trying to read unreadable text;
instructions that are hard to follow even if you can actually read
them. I wasn't very excited about MX Linux anyway, because of their
small user base, so I'm giving it a pass.

The most recent installation of Debian 12 is still there, and the Grub
menu comes up and offers it, but it won't boot. I can select Ubuntu and
that works, but I'm evidently not smart enough to understand the Debian
install utility boot options.

Re: [PLUG] Can't boot to new dual boot

[John Jason Jordan]

On Mon, 18 Sep 2023 12:26:22 -0500
Bill Barry  dijo:

>> I've been reading up on how to get this to work and I haven't found
>> the answer yet. Both drives have a separate partitions for / and
>> /home, and each of them has a /boot/grub/grub.cfg file in the /
>> partition. At the top of the menu entries, the one in the Debian
>> drive has Debian and Debian-Alternative followed by 80 (believe it
>> or not) menu entries for Xubuntu. On the Xubuntu drive the file has
>> menu entries only for Xubuntu, although only about 20 of them.
>> Methinks some serious tidying up is overdue, but that can wait.
>> Maybe a command to update grub is the right way to do it.

>I had a similar problem after I did an update yesterday. Only one of
>my systems was bootable. It turned out there was a line in  the file
>/etc/default/grub like this
>GRUB_DISABLE_OS_PROBER=false
>which was commented out with a # in front of it.
>Uncommenting it then running update-grub fixed the problem.

I had already done that in both the grub.cfg files, but no luck.
However, I finally kludged together something that got me booted into
the Debian OS, and as you can see, I can reboot into Xubuntu.

I had noticed that the grub.cfg file in the Debian installation had two
menu entries at the top, which were missing in the Xubuntu file. I had
been spending all my efforts trying to get the BIOS to find the Debian
grub.cfg file, but I finally decided 'fine, if all it can find is the
Xubuntu file, the all I need to do is add the Debian menu entries to
the top of the Xubuntu file.' It took some finagling because of dealing
with two files owned by root, but eventually I got the whole two menu
entries from the Debian file pasted into the top of the Xubuntu menu
entries. As I rebooted I was telling myself 'there is no way this is
going to work, surely the computer won't boot to anything.' I had made
a copy of the Xubuntu file, and I had visions of having to find a
Knoppix disk or something to use so I could put the copy back, but
guess what! It booted straight into Debian 12!

Just before it booted I saw what looked like a Grub menu flash by in
the upper left corner of my screen. It was gone way too fast to read,
but it looked like there were eight lines in tiny text, text that
happens on a 4K screen before it gets to a GUI. From past experience
each distro in the menu probably had a main line, then a recovery line,
followed by a couple lines for memtest. After I shut down Debian to
come back here I remembered that hitting Esc after the BIOS gets you
the Grub menu, so that's what I did to get back here.

I never did get the BIOS to boot to the Grub folder in the new Debian
drive. But at least it's working, and all I need to do is figure out
how to get the Grub menu into a readable font, and make it come up
always, without having to remember to hit Esc.

Being a Linux user for years has accustomed me to using fudges and
pokes to get things to work. Today was proof of that. :)

[PLUG] Can't boot to new dual boot

[John Jason Jordan]

Xubuntu is on a three-year-old 1TB M.2 card and it has worked well for
years. I think I want to change to Debian, so I downloaded the Xfce
flavor of the Debian 12 ISO and burned it to a USB drive. Then I went
out and bought a new 2TB M.2 drive and installed it in the computer in
a spare, unused slot. I booted to the flash drive and went through the
installation, specifying the new 2TB drive, creating a 200GB partition
for / and 1800GB for /home. At the end the installer stopped and said
it had found the Xubuntu installation, and did I want to create a dual
boot? I answered 'yes,' and it said that upon booting I would have the
choice of which OS to boot to. After the installation completed and I
rebooted it went straight into Xubuntu; no option to boot to Debian,
like the new drive wasn't even installed. While in Xubuntu I noted that
the / and /home partitions on Debian had been mounted, so I looked at
them and all the files appeared to be in the partitions.

The BIOS has a feature where you can choose which disk drive you want
to boot to; all you have to do is hold down F12 and you will get a
menu. I did so, and there was the new M.2 drive in the list, so I
selected it. Unfortunately it did not boot. Instead, a few minutes later
I was staring at a black screen with a flashing underscore in the upper
left corner. The keyboard was inactive, and there was no mouse.

I had one more trick up my sleeve - go to the the BIOS directly. In the
BIOS I swapped the boot order of the drives so the new M.2 drive would
be first to boot. But when I booted it still wouldn't boot Debian; all I
got was the same black screen and flashing underscore.

I've been reading up on how to get this to work and I haven't found the
answer yet. Both drives have a separate partitions for / and /home, and
each of them has a /boot/grub/grub.cfg file in the / partition. At the
top of the menu entries, the one in the Debian drive has Debian and
Debian-Alternative followed by 80 (believe it or not) menu entries for
Xubuntu. On the Xubuntu drive the file has menu entries only for
Xubuntu, although only about 20 of them. Methinks some serious tidying
up is overdue, but that can wait. Maybe a command to update grub is the
right way to do it.

This would probably be easy to fix, if I only knew how. I'm anxiously
looking for suggestions so I can look at Debian 12 on the new drive. :)

Re: [PLUG] Mint, Ubuntu, Debian (and Centos and Redhat/IBM)

[John Jason Jordan]

On Sun, 17 Sep 2023 16:19:41 -0700
"Kevin Williams"  dijo:

>John, what was your specific error that you got when using SSH to your
>computer? What is the local SSH server not already running?

Way too long go to remember.

>Do I understand correctly that you got the silicondust App to see the
>TV tuner on Mint and Ubuntu, but not on Debian?

Yes. On Debian the utility wouldn't even launch, with the error message
that it couldn't find the tuner. Then I discovered that I could ping the
tuner from Debian (192.168.1.112), and from Debian I was even able to
download new firmware and install it, with a command furnished by
Silicondust. But the command had a place to put the tuner's ID number,
and the command failed. When I replaced the ID number with the IP
address the command worked and the new firmware is now installed. I
worked on getting the utility to search for the IP address instead of
the ID number, but couldn't find a way to do it. I even played with the
utility's .desktop file, but without success.

>You said you wanted to move away from snap and a bunch too, and you
>didn’t like the artwork on Mint. Are you still looking to try more
>distro’s?

Maybe. Right now I want to get Debian 12 installed on the new drive.
It's installed, but I can't get it to boot. Once I solve that problem
I'll immediately see if I can get it to see the TV tuner. Everyone
blames VirtualBox for the tuner not being seen, in spite of the fact
that I found a thread on the Debian forums where over a dozen people
had the same problem, and none of them succeeded. If Debian can see the
tuner, then Debian will be my new OS. If not, then I am still distro
shopping.

>If you prefer to stick with apt-based distributions, what about MX
>Linux? It’s based on Debian, but has many papercuts worked out,
>possibly, including the one for the TV tuner.

I have heard of MX Linux, and it's on my list of possibilities. 

>If you’re willing to try an Linux distribution not based on apt, I can
>say that Fedora and OpenSUSE Tumbleweed have improved tremendously in
>usability and reliability in the last five years.
>If you’re not happy with Mint, and if MX Linux does not work with the
>TV tuner either, would you consider trying those rpm-based distros?

I ran Fedora for about two years, until I finally got tired of their
forced six month upgrades. I understand why Red Hat has that
requirement, and I respect their decision. But it's not for me. As for
others, I once briefly tried OpenSUSE, but not a lot of others. I want
a combination of cutting edge, good multimedia support, yet high
stability, a tall order for any distro. I don't find those features as
readily in RPM distros as in Debian. Plus, it takes quite a lot of
different commands and ways of doing things.

Re: [PLUG] Make secure shell work

[John Jason Jordan]

On Sun, 17 Sep 2023 15:08:24 -0700
Russell Senior  dijo:

>On Sun, Sep 17, 2023 at 11:05 AM John Jason Jordan 
>wrote:
>
>> Several times over the years I gave tried to get ssh working, and I
>> always failed.  
>
>Wait, what?
>
>Ssh shouldn't be hard. Like, at all. Let's fix that.  What doesn't
>work?

Good idea, but I'm currently in the middle of installing Debian 12
Bookworm as a dual boot, using the brand new 2TB M.2 drive that I
bought this morning from Best Buy at Jantzen Beach. I recently bought
one for $120, but the price today was only $100. Ssh will have to wait,
but I'm eager to learn how to do it.

Re: [PLUG] Mint, Ubuntu, Debian (and Centos and Redhat/IBM)

[John Jason Jordan]

On Sat, 16 Sep 2023 22:22:00 -0700
wes  dijo:

>On Sat, Sep 16, 2023 at 11:02 AM John Jason Jordan 
>wrote:
>
>> >>> Normally release is done with the Right-Ctrl key, but if you're
>> >>> in the host and the guest has exclusive control of mouse and
>> >>> keyboard, you have no Right-Ctrl key  
>> >  
>> >>something is not right here.  
>>
>> I'm still stuck using the power button.

>one possible thing to try would be to connect to this system remotely.
>if you have set up ssh access, you can connect in and terminate the
>virtualbox process from the command line. it would be interesting to
>find out whether this returns everything to normal.

Several times over the years I gave tried to get ssh working, and I
always failed.

>I sort of doubt that the window manager really died - the existing
>windows should no longer display if that were the case. or at least
>you should lose some of the surrounding detail (title bars, clock,
>notification tray, etc).

Good points. Let me revise my theory: Window manager did not die
completely, but some piece of it died. 

But that has become more or less moot.The Debian installation was
always temporary to let me experiment to be sure I could get all my
applications installed. That worked out reasonably well; Debian has
snap in its repositories, so it can be added, but it's not installed by
default, and I got almost everything running. 

The hangup was the GUI config program for my Silicondust television
tuner on my home network. I got the program installed in Debian, but it
couldn't see the tuner. Yet I could ping the tuner from Debian. I spent
a lot of time trying to solve this, and in the process I finally
discovered a long thread in the Debian forums, where dozens of people
had the same problem, and all of them failed, and none of them were
running in any virtual machine. Reluctantly I had to bid Debian
goodbye. Yes, I could buy a separate television set and find a place
for it on my desk, but why? 

On my Xubuntu 22.04.3 I have managed to replace the few snap packages
that I had with apt packages, and my next effort will be to see if I
can completely uninstall snap. 

I also installed Mint 21.2 Xfce in a VirtualBox machine and got it
running. Like Debian, Mint makes snap available, but discourages
installing it. But Mint has the ugliest artwork of any distro, and I
can't bear the thought of having to stare at it constantly. However, I
can minimize the Mint window and restore it without problems.

I started this whole voyage because I have come to have a strong
dislike of snap and was looking for an alternative to Ubuntu. Assuming
that I found a good solution I was going to buy a 2TB M.2 card to
install in my Thinkpad in its second, currently unused, M.2 slot, then
install the replacement OS on it, leaving the existing 1TB M.2 card with
Xubuntu on it installed as a dual boot. I recently bought a Samsung 2TB
M.2 card from Best Buy at Jantzen Beach for $120 for my desktop
computer, and it is damn fast. 

Re: [PLUG] Mint, Ubuntu, Debian (and Centos and Redhat/IBM)

[John Jason Jordan]

On Mon, 11 Sep 2023 14:10:31 -0700
John Jason Jordan  dijo:

>On Mon, 11 Sep 2023 13:37:15 -0700
>wes  dijo:
>
>>On Mon, Sep 11, 2023 at 1:32 PM John Jason Jordan 
>>wrote:  
>>> Normally release is done with the Right-Ctrl key, but if you're in
>>> the host and the guest has exclusive control of mouse and keyboard,
>>> you have no Right-Ctrl key  
>
>>something is not right here. there should never be a case where the
>>virtualbox guest can prevent virtualbox from acting on the "host key"
>>(right-ctrl by default, as you noted). I don't have any specific
>>suggestions at this point, but I would encourage you to peruse through
>>the settings related to this to see if anything sticks out.  
>
>I currently have three other virtual machines (all Windows), and in the
>past I have had numerous others. I have never had this happen before,
>so my guess is that you are right, 'something is not right here.' I
>will poke at this more later today. 

OK, I have more information, although I don't have a fix. I thought the
problem had just disappeared, but today it happened again. And while
things were locked up I spent an hour trying different things to see if
I could repair whatever was wrong. One of the things I discovered was
that all running programs were still working, e.g., I was downloading a
distro by qBittorrent, whose window was still on the screen, and I
suddenly noticed numbers shifting as the download continued. In
qBittorrent I tried moving around inside the window with the tab key,
moving to different buttons to click on and it worked, except clicking
failed. Eventually I discovered that the keyboard was fully functional
unless I tried to minimize or maximize a window, or start or stop a
program, or get to a terminal. The mouse cursor was working fine, but
no clicks were possible with any of the keys or scroll wheel. At the
end I came to the conclusion that what had gone wrong was that the
window manager had been killed. On the net via my phone I read about
key commands to restart the window manager, but none of them worked.

I also discovered that it only happens when I minimize the window that
is holding the Debian guest OS. I hadn't done this for a couple days,
which is why I thought that the problem was resolved, but today I did it
again, and sure enough, instantly I had no window manager. And since I
never found a solution I'm still stuck using the power button. 

Well, at least I've made some progress. And now I can easily replicate
the problem for testing purposes. 

Re: [PLUG] Debian 12 on VirtualBox

[John Jason Jordan]

On Thu, 14 Sep 2023 14:01:09 -0400
Tomas Kuchta  dijo:

>On Thu, Sep 14, 2023, 13:13 John Jason Jordan  wrote:
>
>> ... snip ...
>Before someone takes the trouble to point out to me that to make shared
>> folders work you need to have VirtualBox, the Guest Additions, and
>> the Extensions all installed, and each of the same version. Yes, I
>> do have all of them installed, and they're all the same version.
>> That is, they're installed in the 22.04 host, but not Debian,
>> because they're not in the existing default Debian repositories.

>You may not want to try yet another vbox fix, me neither, so to that
>end I install both vbox + additions from virtualbox.org, not from the
>distro repositories. Virtualbox.org also host update repositories, so
>nothing breaks over and over after updates.
>
>This (above) has saved me a lot of anxiety, wasted time, . anyway.

I had already thought of that, but hadn't yet tried it. So just now I
did so, installing 7.0, plus 7.0 guest additions on the host and 7.0
guest additions on the guest, and still nada.

But if you notice my reply to wes, I have now successfully logged into
one of my two Synology NAS machines, so shared folders are no longer
worth wasting more time on. Nevertheless, from now on I'm going to
follow your advice and install from Oracle. :)

Re: [PLUG] Debian 12 on VirtualBox

[John Jason Jordan]

On Thu, 14 Sep 2023 11:05:11 -0700
wes  dijo:

>On Thu, Sep 14, 2023 at 10:13 AM John Jason Jordan 
>wrote:
>> I'm having a terrible time getting the guest OS to see the shared
>> folders that I specified in the setup for the Debian 12 machine.

>I have often had problems getting shared folders to work in vbox too.
>I use network based file transfers instead. which type depends on the
>situation. "sshfs" would be the closest to a shared folder, but might
>be complex to set up for someone who is not super familiar with how to
>use ssh. I seem to recall that you have used nfs in the past, which
>would also be a good solution to this particular need.

Yes, I did once spend a lot of time getting nfs working, but I could
never successfully get it working, and eventually I gave up. 

Debian 12 works great at the internet. And your comment made me think
of my two Synology diskstations. Just now, in Firefox on Debian I typed
in the address, which popped up a login window, whereupon I logged in,
et voilà!

To hell with shared folders. Oracle can go suck eggs. :)

[PLUG] Debian 12 on VirtualBox

[John Jason Jordan]

I now have Debian 12 Bookworm installed and running sort of OK in
Virtualbox. Thinking about switching t it from Xubuntu 22.04 I've been
testing it to see if I can install all the applications that I need.
Toward that end I made a list of 70+ programs to install, and started
adding them. From the list it looks like about half of them are not in
the default repositories. For one of my next adventures I'll see about
adding additional repositories.

I note that snap and flatpak are not installed in the default, but both
can be added. I also discovered that ntp and sntp were not installed,
so after 24 hours the clock displayed in the panel was way off. I've
never installed a distro that didn't install them by default.

I'm having a terrible time getting the guest OS to see the shared
folders that I specified in the setup for the Debian 12 machine. The
only thing it can see is the optical drive, but that's not an easy way
to test newly installed applications with my existing files. Virtualbox
is a grand PITA with shared folders, and always has been. I searched up
and down the web and found lots of pages with instructions for getting
shared folders working, but still no success. You open one web page of
guaranteed how-tos, and you start following the list of things to do,
and then you come to a line that doesn't work. When you finally decide
to move on to the next page of instructions you get a completely
different list of things to do, so you start out, only to eventually
get to a line that won't work. And on to the next set of stuff that
also eventually fails. And every time you think you've finally got it
you have to shut down the guest OS and reboot it, always discovering
that it still can't see the shared folders. Yesterday I rebooted Debian
12 at least 50 times, always without success. And yet, at the very
beginning I swear that the shared folders appeared.

I've got some blank 25GB Blu-ray disks. Maybe I should just burn a
bunch of files to a disc so I'll finally be able to test newly
installed applications on Debian 12.

Before someone takes the trouble to point out to me that to make shared
folders work you need to have VirtualBox, the Guest Additions, and the
Extensions all installed, and each of the same version. Yes, I do have
all of them installed, and they're all the same version. That is,
they're installed in the 22.04 host, but not Debian, because they're
not in the existing default Debian repositories.

Wish me luck. :)

Re: [PLUG] Mint, Ubuntu, Debian (and Centos and Redhat/IBM)

[John Jason Jordan]

On Mon, 11 Sep 2023 13:37:15 -0700
wes  dijo:

>On Mon, Sep 11, 2023 at 1:32 PM John Jason Jordan 
>wrote:
>> Normally release is done with the Right-Ctrl key, but if you're in
>> the host and the guest has exclusive control of mouse and keyboard,
>> you have no Right-Ctrl key

>something is not right here. there should never be a case where the
>virtualbox guest can prevent virtualbox from acting on the "host key"
>(right-ctrl by default, as you noted). I don't have any specific
>suggestions at this point, but I would encourage you to peruse through
>the settings related to this to see if anything sticks out.

I currently have three other virtual machines (all Windows), and in the
past I have had numerous others. I have never had this happen before,
so my guess is that you are right, 'something is not right here.' I
will poke at this more later today. 

Re: [PLUG] Mint, Ubuntu, Debian (and Centos and Redhat/IBM)

[John Jason Jordan]

On Mon, 11 Sep 2023 12:50:09 -0700
John Jason Jordan  dijo:

>Nevertheless I did get it installed, and now it's on to figuring out if
>Debian 12 will really work for me. :)

And once more I had to power down he computer because the virtual
machine had captured the mouse and keyboard, and *after* I had
minimized the machine, so I had no way to make it release the mouse and
keyboard.

Normally release is done with the Right-Ctrl key, but if you're in the
host and the guest has exclusive control of mouse and keyboard, you have
no Right-Ctrl key; you have no keyboard or mouse at all. If the window
for the guest is minimized in the host, you are dead in the water, and
sunk to the bottom.

I can't remember what they're called, but they're little windows that
an application pops up with some attribute that completely locks the OS
until the little window gets what it wants. I suspect that one of these
little windows popped up in the guest right after I minimized its
window in the host.

This is a VirtualBox issue, not a Debian problem. I need a solution. I
can't keep having my computer locked up every half hour or so.

Re: [PLUG] Mint, Ubuntu, Debian (and Centos and Redhat/IBM)

[John Jason Jordan]

On Mon, 11 Sep 2023 12:12:18 -0500
Bill Barry  dijo:

>> C'mon Debian people, don't you test this stuff?
>
>I don't see that installing from the Live OS is the recommended path
>for installing Debian. The netinst method is one of the recommended
>installation methods.

I know that net install is more common. But if the Live ISO has a
button on its main page that says 'Install Debian,' then it ought to
work. I was not impressed in my installation experience with Debian 12
Bookworm. More follows.

The continuing saga of a Live ISO install:

I did finally get it installed. The Live ISO initially boots to a Grub
menu, and one of the options is to boot to an 'Install' version. This
version immediately went into installation, bypassing the need to deal
with the worthless 'Install Debian' button. This presented more or
less the same questions (time zone? language? etc.), but the screens
asking these questions were different.

Things still did not go well. When I got to how to partition the
20GB VirtualBox .vdi image I told it to create / of 6GB and the
remainder for /home. It errored on formatting these partitions,
claiming that the media could not be formatted. I spent an hour trying
to get past partitioning, and finally succeeded when I went back to the
beginning of the partitioning section and chose the option to let it
just create one partition whatever way it wanted. Too many bugs in the
'manual' partition process.

Finally it started copying files from the DVD, but after fifteen
minutes or so it stopped and popped up an error message that the source
refused to deliver data. I tried again, and after half an hour the
mouse became locked to the guest so I couldn't do anything in the host,
not even powering down the guest was possible. Eventually I decided the
only way out was to use the computer's power button, killing
everything. After rebooting the host I opened VirtualBox and proceeded
with attempts to install Debian 12. I opened the Debian machine and
tried again, and this time at about halfway through the Live ISO
apparently had a power manager timeout and locked the screen, demanding
my password to continue. At the very beginning I had told it to boot
automatically without a password, so I had never created one. I
couldn't get the screen unlocked, but the optical drive was still
grinding and clucking along, so I let it sit. When the drive finally
went silent I killed the VirtualBox window, and then rebooted. It came
up complete without error messages, so evidently I had finally gotten
it installed.

The first thing I did was to turn off screen locking, and then I went on
with configurations. I quickly needed to install some things, so I
launched Synaptic, which popped up a little authentication window
asking for my non-existent password. Eventually I discovered that if I
left the password blank and just proceeded the Synaptic window came up
anyway. Not the smoothest way to handle a user without a root password,
but eventually it worked.

The whole installation took about six hours, although maybe a third of
that was because VirtualBox refused to boot from the USB drive, so I
had to install from the much slower DVD. Another third or so was
clunking around fixing failures to install.

My main purpose in this exercise was to see if Debian 12 Bookworm would
provide an alternative to my current Xubuntu 22.04.3, and a big part
of the answer is determining if all the applications I need are
available in the repos or otherwise; and if I can stay completely away
from snaps. Xubuntu 22.04.3 is the host, so I can just pop back and
forth to make sure I don't leave out any critical applications.

It has been about three months since the Xfce Live ISO went public, and
I'm not terribly impressed. Bugs happen, bad design decisions happen,
and lots of other problems crop up, but three months after the public
started downloading the ISO I would not have expected all the issues
that I experienced. Admittedly, VirtualBox was responsible for some of
them, like not being able to boot from the USB drive (which it does
see). If Debian labels the Bookworm installation as 'not supported as a
virtual machine,' that's a cop out and they can forget about me as a
possible user.

Nevertheless I did get it installed, and now it's on to figuring out if
Debian 12 will really work for me. :)

Re: [PLUG] Mint, Ubuntu, Debian (and Centos and Redhat/IBM)

[John Jason Jordan]

On Sun, 10 Sep 2023 22:26:42 -0700
John Jason Jordan  dijo:

>On Sun, 10 Sep 2023 16:55:57 -0700
>Keith Lofstrom  dijo:
>
>Re Debian 12 Hookworm
>
>Just for kicks I downloaded the Live Xfce ISO and burned it to a small
>USB 3.0 drive. I may see what happens if I try to install it in
>VirtualBox. Meantime I'm trying to find out details about it,
>specifically its potential snapitude, and a few other things.

I spent two hours trying to get Virtualbox to recognize either the
optical drive or a USB drive where I had installed the Live Debian 12
ISO that I downloaded. I finally succeeded at getting the virtual
machine to use the optical drive, although I failed with the USB drive.

Once I finally got the Live OS to boot there was an 'Install Debian'
icon, so I clicked on it. That popped up an error message that
'Calamares' (the install program) was not executable. There was a
button to make it executable, so I clicked on it and things progressed.
It finally got to the end of the questions and answers and presented a
summary of what it was going to do in the install. I looked all over
for a button that said 'OK' or 'Continue,' or anything to move on and
let it install the OS, but there was no way to get past that point.

C'mon Debian people, don't you test this stuff?

Re: [PLUG] Mint, Ubuntu, Debian (and Centos and Redhat/IBM)

[John Jason Jordan]

On Sun, 10 Sep 2023 16:55:57 -0700
Keith Lofstrom  dijo:

Re Debian 12 Hookworm

Just for kicks I downloaded the Live Xfce ISO and burned it to a small
USB 3.0 drive. I may see what happens if I try to install it in
VirtualBox. Meantime I'm trying to find out details about it,
specifically its potential snapitude, and a few other things.

[PLUG] Ubuntu messed up LO!

[John Jason Jordan]

I had LO downloaded from LO (not the Ubuntu repositories) and installed
on my Xubuntu 22.04 laptop. Today I decided to do a system update, using
the Update Manager utility. When the Update Manager finished sniffing
the internet and my computer it came up with over 250 packages that
needed to be updated. I scanned through the list and saw a lot of LO
packages, so I painstakingly went through the entire list and
deselected all the LO packages. If I had not done so, I reasoned, the
updates would overwrite my installed non-Ubuntu version. When I
finished I clicked the button to start the update.

As I watched the progress, imagine my horror as LO packages displayed as
being installed. WTH? I don't know how this happened, but I figured the
thing to do was let the update finish, reboot, and then open Synaptic
package manager, select all the LO packages and mark them to be
deleted, then click on Synaptic's Apply button to finish the task. My
plan was to get rid of god only knows what mess was installed after the
update, then download the current latest from LO and install it.

Sadly, this failed. Synaptic said it couldn't delete the packages
because I had 'broken' packages that needed to be fixed. Guess which
packages were broken? Yup, you guessed it - every one of the LO
packages that I was trying to uninstall.

Just for kicks and giggles, before further attempts to resolve the
problems, I decided to see what would happen if I tried to launch
Writer, the only package I really use. Amazingly a Writer window
opened, ready to start writing. The only problem was that my floating
toolbar for formatting was missing, but that is easily fixed.

Questions:

1) I had 7.3.7.2 (that's what it says in Writer's Help - About). If I
download and install 7.4.7 or 7.5.4 (the latest from LO), will it
overwrite all the messed up packages? Or will it fail to complete the
installation and leave me with an even bigger mess?

2) What would be a command line tool to fix the broken pages? I'd
settle for just uninstalling them. After all, they're going to be
replaced anyway.

Observations and suggestions welcome!

Re: [PLUG] "sold as 16GB" PNY flash drives store >30GB

[John Jason Jordan]

On Sun, 28 May 2023 00:37:28 -0700
Keith Lofstrom  dijo:

>I bought a "PNY 16GB Attaché 3 USB 2.0 Flash Drive 10-Pack"
>from Amazon for $27.98, to transfer files from some ancient
>CentOS machines to replacement Debian machines.
>
>The four I've tested so far format to more than 30GB (!!!)
>of ext3 file systems.  Hard to believe, so I read/write-
>tested two of them with random patterns for a few days. 
>No bits lost.

In the past I've bought USB drives advertised as 1TB that appeared in
GUI file browsers as 1TB, but which could not take more than 32GB of
data. The end of the story is that they were really 32GB drives that
evildoers in China had hacked to make them appear as 1TB drives. If you
bring up a selection of 1TB drives on eBay you will find many at the
beginning of the list priced at around $16, and I can guarantee you that
every one has been similarly hacked. Then there will be a big empty
space in the listings, followed by more 1TB drives at the $100+ price.
And it's not just eBay; I find the same thing at Amazon and Newegg,
although responsible vendors at least try to hold the bad guys at bay.
And if you want to buy a 1TB SD card you'll find that the same thing is
happening. 

And one more point - the hacked ones were originally 32GB (usually),
and many were name brands like PNY, Kingston, etc; in other words, the
seller bought a big batch of 32GB name brand drives, then hacked them
to sell them as 1TB drives, so you can't assume that if it has a name
brand on the case it is really as big as the seller claims. And one
final point, in US law the principle of 'let the buyer beware' (caveat
emptor) hasn't been true since I was a child; nowadays it has become
discloseat vendor (apologies to ancient Romans). Unfortunately, Chinese
law imposes little or no penalty for fraud, and ditto for many other
countries, and even if what the seller does is really illegal, good luck
finding the seller and serving them with legal process.

I do not understand how the hackers accomplish their machinations, nor
do I know how to undo the hack. But I do have abundant knowledge of
what a dollar should buy, and I am certain as to how many of them are
in my bank account. This gives me one sure way to tell if a drive is
real; just go to (e.g.) pny.com and see what they charge for a drive of
the size you want to buy, and if your vendor is selling a drive of that
capacity for a fraction of the price, give it a pass.

The above is not directly on point to Keith's plaint, but it is related
and may be useful.

Re: [PLUG] Zoneminder server build instructions

[John Jason Jordan]

On Mon, 8 May 2023 21:54:08 -0500
Chuck Hast  dijo:

>Yep and for that a few cops with laser/Radar guns with cameras, that
>will slow them down somewhat.

Do Portland police still do that? I thought they no longer do speed
monitoring (formerly revenue patrol), in addition to no longer
arresting people for drugs, shoplifting and assaults. The purpose, as I
understand it, is to make Portland crime statistics plummet, while
simultaneously reducing the cost of police.

Re: [PLUG] Brave browser behavior changed

[John Jason Jordan]

On Mon, 24 Apr 2023 07:56:24 -0700 (PDT)
Rich Shepard  dijo:

>I'm building and installing many packages on the ThinkPad T430.
>Yesterday, after several hours the browser would not download a file
>until I reloaded the web page. I've not seen this happen on any
>browser or host.
>
>It's a bit annoying, not a critical issue, and I'm curious why it may
>have started yesterday. The same issue is occurring today.

I can't answer your question directly, but my recent experience may
give you some ideas.

I first started using the Brave browser a week or two ago, and
yesterday I finally got it configured the way I want it. It took so
long because:

1) Documentation is thin and out of date. My version (installed on
Xubuntu 22.04 via snap) is Version 1.50.125 Chromium: 112.0.5615.165
(Official Build) (64-bit). But when I search for help on a particular
problem almost every hit gives me solutions that are impossible because
the folders/files/buttons don't exist in my Brave. Apparently every
single upgrade, even tiny ones, requires completely rewriting the user
interface, yet all the web sites with help instructions refer to
something several versions back.

2) There are bugs galore, frequently things that change how things
work. For example, I installed three extensions and they worked well. I
added a four extension and it wouldn't finish downloading. I tried
several others and they wouldn't finish downloading either. It took me
six hours of searching everywhere until I finally, accidentally,
discovered that the problem was that I wasn't logged in to my Google
account. That solved the problem, but then several days later
extensions wouldn't complete the download again. In the settings I
discovered that I was no longer logged into Google. What a PITA. I
never had to log in to Google to install things in Chromium, Firefox,
or any other browser. I don't know whose idea that was, but it wasn't a
very good one. And if you do have to be logged in to Google in order to
install an extension, an error message stating that would have been
nice.

It took many hours to get all the tabs that I had in Chromium finally
working in Brave, mostly because they required usernames and passwords
which, of course, I had forgotten. And in the process I often needed to
move a tab, but that was a nightmare. Dragging a tab almost always
results in it falling out of the channel, and when that happens the tab
becomes a new window and the tab cannot be dragged back to the window
it was in. And even worse, as soon as the tab becomes a new window
Brave opens a new empty window every second until you release the mouse
button. Now you've got to close the many extra empty windows, and then
you need to open a new empty tab in the original window, and copy and
paste the URL from the one you were trying to move. Dragging tabs needs
serious work, like scrap what's there and start over from scratch.

In spite of the problems, I sort of like the Brave browser. I no longer
need DuckDuckGo, and it seems a tiny bit faster than other browsers.
Most importantly I feel more private, although whether that's really
true, and if it is, how much more, are difficult for me to figure out.
I've uninstalled Chromium, and I'm not planning on going back.

Re: [PLUG] 3rd party vpn Defense evasion

[John Jason Jordan]

On Tue, 18 Apr 2023 17:38:23 +
Ted Mittelstaedt  dijo:

>It's not going to be possible to block all VPNs.

I've been using a VPN for several years now, currently Mullvad
(Stockholm based). I'm curious about the efficacy of various VPN
services. I selected Mullvad (and PIA previously) based on how badly
they cut my gigabit access speed. In that respect most give you about
10% of what our ISP provides, if you're lucky. But that was the only
way I had to shop for a VPN. I have no idea how good they are at all
the many other issues you brought up. 

I do know that web sites think I am in the city where I am connected to
one of the servers, Houston at the moment. I know that because, e.g., I
tried to access safeway.com while connected to Mullvad in Stockholm,
which rudely told me that I was not in the US, so I was denied
access because I was clearly a terrorist or other evildoer.

It would be nice if VPN services advertised how effectively they stop
others from finding out who and where you really are.

Re: [PLUG] Google Bard - entry level sys-admin, learning fast?

[John Jason Jordan]

On Thu, 30 Mar 2023 11:01:55 -0700
Galen Seitz  dijo:

>I was going to put this on the plug-talk list for John to chew on, but 
>since the subject came up here...
>
>You Are Not a Parrot
>

Too much to chew on simply and easily. And it's anglo-centric, meaning
that it ignores features from non-Indo-European languages, e.g.,
polysynthetic languages, and even agglutinative languages.

To my mind there are two points of focus in linguistics:

1) Traditional linguistics: Syntax, phonetics, phonology, morphology

2) Discourse analysis: How we react to conversation.

(2) is way harder to deal with. An example might be a classroom at PSU
with multiple cameras to record instruction while teaching English to
speakers of other language, in order to analyze the teacher-student
interactions to improve their communication. Analyzing conversation is
not a simple process; far more complex than other aspects of
linguistics.

Linguists involved in (2) are also well versed in (1), because it is
considered the basis of linguistics, plus they had to study (1) on the
way to their doctorates. Yet, incorrect discourse expressions can
render misunderstanding, even while syntactically accurate:

a) She discovered the answer.

Who discovered the answer? A syntactician would simply render this
utterance as a legal expression under the rules of English. A discourse
analyst would want to know who 'she' is; was the person referred to
earlier, was she the subject of a previous conversation, etc.

I belong with the (1) linguists, but as a writer I find the work of (2)
linguists very useful.

Re: [PLUG] Thumbnails Missing

[John Jason Jordan]

On Mon, 27 Mar 2023 09:00:01 -0700
Dick Steffens  dijo:

>Normally, image files on my desktop, or some Thunar windows, display a 
>thumbnail of that image. Now, I get an icon with the letters GIF or
>JPG instead. I can fix the problem by restarting the machine, but
>shouldn't there be a way to fix it without restarting? It's not true
>for all directories, but other than the fact that the ones where a
>thumbnail is displayed are very old directories, I can't explain why.

Occasionally I use Thunar, but I prefer PCManFM. There are others as
well. Try a different one and see what happens; maybe it'll give you a
clue.

Re: [PLUG] Remind or Tkremind

[John Jason Jordan]

On Wed, 21 Dec 2022 16:08:21 -0800
Steve Dum  dijo:

>    A field may be an asterisk (*), which always stands for 
>``first-last''.

Ah! That's what I missed!

The new line (for tomorrow) is:

0 9 22 * * MAILTO="" DISPLAY=:0.0 gxmessage -font "Junicode 24"
Pay T-Mobile

That should pop it up at 9am tomorrow, and just once, and leave it until
I click on its 'OK' button.

Re: [PLUG] Remind or Tkremind

[John Jason Jordan]

Oops. Something is wrong with this line:

* * 21 * * MAILTO="" DISPLAY=:0.0 gxmessage -font "Junicode 24" Pay
  T-Mobile

I expected it to pop up a message today (the 21st), and it did so, but
60 second later it opened up another one, and again every sixty seconds.
I couldn't figure out how to shut down gxmessage. It's not in Top or
Task Manager. Eventually I opened crontab again and put a # in front of
the line, then saved it and exited from the editor. That did the job,
but now I have to figure out why '* * 21 * *' repeats the message every
60 seconds. 

Steve,

Thanks for the suggestion. I spent some time trying to figure out how
to do it with cron and gxmessage. I added this line to crontab:

* * 21 * * MAILTO="" DISPLAY=:0.0 gxmessage -font "Junicode 24" Pay
  T-Mobile

That is, I set it at the 21st, because that's tomorrow, to see if cron
pops up the message. If I do just the gxmessage part at the command
line it pops up the message, with an 'OK' button to close the message,
so it ought to work if I got the crontab line correct. 

Crossing fingers for tomorrow. :)


On Mon, 19 Dec 2022 23:42:07 -0800
Steve Dum  dijo:

>my ubuntu 22.04 release comes with a program called notify_send that 
>does what you request
>it takes a -t n option to set the time it stays on screen and totally 
>ignores the parameter.
>However it takes a urgency argument, -u critical seems to leave the
>note on screen until you delete it.
>notify-send -u critical "time to write that check"
>seems to do what you want. But if you do more than 1 notify-send they 
>appear serially so you don't see the next one
>until you close the current one.
>looking at the online man page, it has tabs for every supported ubuntu 
>version so it's been around a while.
>steve
>
>John Jason Jordan wrote:
>> I need an little application that will pop up a reminder on the same
>> day of every month, and crucially, *leave it on the screen until I
>> manually close the window*. There are lots of notify tools, but they
>> just pop up a little notice for 10-20 seconds and then the notice
>> disappears.
>>
>> So far the only thing I have found is Remind, and its GUI tool,
>> Tkremind. I've installed both, and they run, but there is no
>> documentation other than a man page, which I can't figure out. That
>> is, I created a reminder, but it never appears, so clearly I haven't
>> done it right.
>>
>> The purpose is to pay a bill that comes due every month on the 20th,
>> and each month with a different amount. I could set up an auto-pay
>> with my bank, but not if the amount varies. The vendor also has an
>> auto-pay option, but to use it I have to give them my credit card
>> number and authorization to charge it for whatever they want. No
>> thanks, I'm not that stupid.
>>
>> I could probably figure out how to use cron, but cron needs to call a
>> program to do the announcement. Or at least I think it does. Cron is
>> a mystery to me.
>>
>> Surely there exists a simple, understandable reminder app. Any
>> suggestions?
>>  
>

[PLUG] Remind or Tkremind

[John Jason Jordan]

I need an little application that will pop up a reminder on the same
day of every month, and crucially, *leave it on the screen until I
manually close the window*. There are lots of notify tools, but they
just pop up a little notice for 10-20 seconds and then the notice
disappears.

So far the only thing I have found is Remind, and its GUI tool,
Tkremind. I've installed both, and they run, but there is no
documentation other than a man page, which I can't figure out. That is,
I created a reminder, but it never appears, so clearly I haven't done
it right.

The purpose is to pay a bill that comes due every month on the 20th,
and each month with a different amount. I could set up an auto-pay with
my bank, but not if the amount varies. The vendor also has an auto-pay
option, but to use it I have to give them my credit card number and
authorization to charge it for whatever they want. No thanks, I'm not
that stupid. 

I could probably figure out how to use cron, but cron needs to call a
program to do the announcement. Or at least I think it does. Cron is a
mystery to me.

Surely there exists a simple, understandable reminder app. Any
suggestions?

Re: [PLUG] Restart mdadm? SOLVED

[John Jason Jordan]

On Mon, 28 Nov 2022 11:59:14 -0800
Russell Senior  dijo:

>Probably a good idea to record what /proc/mdstat looks like when
>things are working, ahead of time, too.

Done, and added to mdadm.txt file. Thanks!

Re: [PLUG] Restart mdadm? SOLVED

[John Jason Jordan]

On Mon, 28 Nov 2022 11:26:01 -0500
Tomas Kuchta  dijo:

>TB is hot plug interface. As such, I'd sort of expect that HW and SW
>assumes that it can safely traverse power states and that connection
>resets may be OK.
>
>If that is what is happening, it is of course pretty nasty thing to do
>to md arrays. Native higher level file systems btrfs/zfs may fare
>better, but I would be skeptical and expect similar problems. JBOD
>should be more resilient.
>
>In any case system logs, particularly power state changes should
>confirm what is actually happening.

Russell Senior  dijo:

>Another good starting point is: cat /proc/mdstat

Both comments above added to my mdadm.txt file for next time this
happens. Thanks!

Re: [PLUG] Restart mdadm? SOLVED

[John Jason Jordan]

On Sun, 27 Nov 2022 22:39:48 -0800
Paul Goins  dijo:

>When this occurs, have you tried "dmesg -T" or checked
>/var/log/kern.log or /var/log/syslog for suspicious log entries around
>the time of the error? Might be worth taking a peek as well.

Your suspicion was correct - I had not tried either of those options,
and because I was unaware of them. I have added your suggestions to my
file for mdadm solutions. Next time this happens I will see what they
produce.

Re: [PLUG] Restart mdadm? SOLVED

[John Jason Jordan]

On Sun, 27 Nov 2022 19:06:35 -0600
Bill Barry  dijo:

>On Sun, Nov 27, 2022 at 7:02 PM Bill Barry  wrote:
>>
>> On Sun, Nov 27, 2022 at 6:55 PM John Jason Jordan 
>> wrote:  
>> >
>> > On Sun, 27 Nov 2022 16:07:09 -0600
>> > Bill Barry  dijo:
>> >  
>> > >So reassembling is what was needed, but you don't need to force
>> > >it. Now if you just had some way to automate that.  Have you
>> > >connected your external enclosure to the UPS so that it is
>> > >protected from temporary power disruptions?  
>> >
>> > Yes, everything is on my three massive APCs (connected together),
>> > even the stereo.
>> >
>> > I keep thinking that the problem lurks somewhere in the crazy
>> > setup:
>> >
>> > 4 Intel 8TB NVMe drives, mounted two each on two long PCIe cards,
>> > inside a 4-bay PCI enclosure, in turn connected to a Thunderbolt 3
>> > port on a Lenovo dock, in turn connected to my Lenovo P73 laptop
>> > via one of its TB3 ports. The whole setup is close to two years
>> > old, during which time this is the sixth case where the RAID has
>> > failed.
>> >
>> > When I got up this morning everything was running, so it couldn't
>> > have been caused by a power glitch. There are lots of other
>> > electrical devices around the house that tell me when PGE hasn't
>> > been playing nice
>> > - e.g., the clocks on the range and microwave will be flashing, the
>> >   bedroom television that runs all night will be off, among
>> > others.  
>>
>> If it's not a power problem then maybe cabling. How long are your
>> Thunderbolt cables?
>>
>> Bill  
>
>I see lots of posts out there claiming that their Thunderbolt
>connection  randomly disconnects. That could certainly cause your
>problem.
>
>Bill

Ah! Now that is news I was not aware of! I will do some searching and
see what additional information I can come up with. And (crossing
fingers) a solution.

Regarding the cables, I have several TB3 cables, and after the first
couple of failures I swapped them, with no change in the problem. Of
course, there are lots more TB3 connection suspects - the three ports,
for example. But I should have mentioned before that the two PCIe cards
and their connections in the enclosure, and the connections of the four
drives to the cards are the least likely suspects, because after each
of the failures all four drives appear correctly and 'mdadm --query' and
'mdadm --detail' both list them as available without a hint of a
problem.

Re: [PLUG] Restart mdadm? SOLVED

[John Jason Jordan]

On Sun, 27 Nov 2022 16:07:09 -0600
Bill Barry  dijo:

>So reassembling is what was needed, but you don't need to force it.
>Now if you just had some way to automate that.  Have you connected
>your external enclosure to the UPS so that it is protected from
>temporary power disruptions?

Yes, everything is on my three massive APCs (connected together), even
the stereo. 

I keep thinking that the problem lurks somewhere in the crazy setup:

4 Intel 8TB NVMe drives, mounted two each on two long PCIe cards,
inside a 4-bay PCI enclosure, in turn connected to a Thunderbolt 3 port
on a Lenovo dock, in turn connected to my Lenovo P73 laptop via one of
its TB3 ports. The whole setup is close to two years old, during which
time this is the sixth case where the RAID has failed.

When I got up this morning everything was running, so it couldn't have
been caused by a power glitch. There are lots of other electrical
devices around the house that tell me when PGE hasn't been playing nice
- e.g., the clocks on the range and microwave will be flashing, the
  bedroom television that runs all night will be off, among others.

Re: [PLUG] Restart mdadm? SOLVED

[John Jason Jordan]

On Sun, 27 Nov 2022 12:49:54 -0800 (PST)
Rich Shepard  dijo:

>On Sun, 27 Nov 2022, John Jason Jordan wrote:
>
>> I've searched documentation on mdadm and I can't figure out if there
>> is a way to just restart the array. Suggestions?  

>How do I reactivate my MDADM RAID5 array? - Super User You should try
>stopping and re-starting the array: mdadm --stop /dev/md0 mdadm
>--assemble --scan to re-assemble the array and if that doesn't work,
>you may need to update your mdadm.conf, see for example this question
>for details on how to do that. Share Improve this answer Follow edited
>Mar 20, 2017 at 10:17 Community Bot 1
>
>mdadm RAID array stop and restart - Ask Ubuntu https://askubuntu.com ›
>questions › 1398907 › mdadm-raid-array-stop-and-restart Mar 23,
>2022Now to stop the array and restart it , I'm following below steps:
>mdadm --stop device mdadm --create . (the same command which I
>used to create the array at first) --- > Is it correct process to stop
>and restart RAID arrays? --- > Should I run "mdadm --zero ..." to
>clear superblock after stopping array ? Is it mandatory step to stop
>array?

Rich, I found that AskUbuntu web page, but I note that the question
received no answers. In any event, stopping the array was a problem:

sudo mdadm --stop /dev/md0
mdadm: Cannot get exclusive access to /dev/md0:Perhaps a running
process, mounted filesystem or active volume group?

The error message is hardly a surprise, since /dev/md0 was mounted -
sudo mount gave me dozens of lines, including: 
/dev/md0p1 on /media/jjj/Movies type ext4

And then sudo umount /dev/md0p1 worked. But when I tried to remount it
I got:
sudo mount -a (it's in fstab)
mount: /media/jjj/Movies: can't read superblock on /dev/md0p1

So then I did 'sudo mdadm --stop /dev/md0,' which returned:
mdadm: stopped /dev/md0

And then I found a page that gave me the final solution:

sudo mdadm -A -s
mdadm: /dev/md/0 has been started with 4 drives.

Where -A = --assemble and -s = --scan, which scans for all arrays
located in /etc/mdadm/mdadm.config - checked earlier so I knew /dev/md0
was in there with its UUID. I note that the documentation for --assemble
is confusing. Nowhere did I find anything that said its purpose is to
*restart* an already created array. I was afraid it would create a new
array.

I could have continued with mounting it on the command line, but my GUI
file manager auto-mounts everything in 'Places' just by clicking on it,
so I clicked on it, et voilà!! I'm back in business!

This array has been a pain for the past year and a half, so I'm going
to copy this whole discussion into my 'Commands.txt' file for future
reference. I still need to figure out why the array keeps failing, but
at least now I can recover somewhat gracefully.

Thanks for poking me in the right direction!

[PLUG] Restart mdadm?

[John Jason Jordan]

It's happened again. One of my RAID0 arrays just stopped working. The
mount command shows it as mounted, but neither the command line nor any
GUI file managers can see any files, and if I try to open a file that I
know is there I get "Error reading file: Input / output error."
Attempts to umount it fail because mount thinks it is busy, and "mount
-a" executes without error, but does nothing.

From previous experience the solution is to reboot, which restarts the
array and all is well again. I've never lost any data, and if I did I
have a full mirror on a Synology, but it's a major PITA to have to
reboot. 

I've searched documentation on mdadm and I can't figure out if there is
a way to just restart the array. Suggestions?

Re: [PLUG] I hate the mount command

[John Jason Jordan]

On Tue, 15 Nov 2022 14:58:56 -0800
John Jason Jordan  dijo:

>OK, I may be making some headway here.
>
>sudo mdadm -D /dev/md0
>/dev/md0:
>   Version : 1.2
> Creation Time : Mon Jan 25 13:53:32 2021
>Raid Level : raid0
>Array Size : 30005334016 (27.94 TiB 30.73 TB)
>  Raid Devices : 4
> Total Devices : 4
>   Persistence : Superblock is persistent
>
>   Update Time : Mon Jan 25 13:53:32 2021
> State : broken, FAILED 
>Active Devices : 4
>   Working Devices : 4
>Failed Devices : 0
> Spare Devices : 0
>
>Layout : -unknown-
>Chunk Size : 512K
>
>Consistency Policy : none
>
>Number   Major   Minor   RaidDevice State
>   0 25920  active sync   missing
>   1 25911  active sync   missing
>   2 25902  active sync   missing
>   3 25933  active sync   missing
>   
>Also tried sudo mdadm --run /dev/md0, but there was no output. 
>
>Is this repairable without wiping out the data and recreating the
>array?

I did some searching and found several suggestions involving mdadm
commands, but the suggestions just offered the commands with no
explanation of what the command does, nor could I find clear and
understandable explanations on the man page. For example, one
suggestion used the mdadm command 'assemble,' and when I looked in the
man page it said 'use assemble to assemble the raid.' Using the term to
define the term is a useless pleonasm.

In the end I closed down all running programs and then shut down the
computer. When I rebooted everything was running as it should,
including the RAID. In the year and a half since I bought the drives
and set up the RAID this is at least the fifth time that the RAID has
failed while in use. In each case a mere reboot cured the problem. But I
have to say that the next time I need to set up a RAID mdadm will be my
last choice.

Re: [PLUG] I hate the mount command

[John Jason Jordan]

OK, I may be making some headway here.

sudo mdadm -D /dev/md0
/dev/md0:
   Version : 1.2
 Creation Time : Mon Jan 25 13:53:32 2021
Raid Level : raid0
Array Size : 30005334016 (27.94 TiB 30.73 TB)
  Raid Devices : 4
 Total Devices : 4
   Persistence : Superblock is persistent

   Update Time : Mon Jan 25 13:53:32 2021
 State : broken, FAILED 
Active Devices : 4
   Working Devices : 4
Failed Devices : 0
 Spare Devices : 0

Layout : -unknown-
Chunk Size : 512K

Consistency Policy : none

Number   Major   Minor   RaidDevice State
   0 25920  active sync   missing
   1 25911  active sync   missing
   2 25902  active sync   missing
   3 25933  active sync   missing
   
Also tried sudo mdadm --run /dev/md0, but there was no output. 

Is this repairable without wiping out the data and recreating the array?

Re: [PLUG] I hate the mount command

[John Jason Jordan]

On Tue, 15 Nov 2022 01:46:13 -0800
Russell Senior  dijo:

>It says it's already mounted. Can you post the results of a plain mount
>command? That is, show what the system thinks is mounted.

Here's the plain mount command:

$ mount
sysfs on /sys type sysfs (rw,nosuid,nodev,noexec,relatime)
proc on /proc type proc (rw,nosuid,nodev,noexec,relatime)
udev on /dev type devtmpfs
(rw,nosuid,relatime,size=16099964k,nr_inodes=4024991,mode=755,inode64)
devpts on /dev/pts type devpts
(rw,nosuid,noexec,relatime,gid=5,mode=620,ptmxmode=000) tmpfs on /run
type tmpfs
(rw,nosuid,nodev,noexec,relatime,size=3227136k,mode=755,inode64)
/dev/nvme4n1p1 on / type ext4 (rw,relatime) securityfs on
/sys/kernel/security type securityfs (rw,nosuid,nodev,noexec,relatime)
tmpfs on /dev/shm type tmpfs (rw,nosuid,nodev,inode64) tmpfs on
/run/lock type tmpfs
(rw,nosuid,nodev,noexec,relatime,size=5120k,inode64) cgroup2 on
/sys/fs/cgroup type cgroup2
(rw,nosuid,nodev,noexec,relatime,nsdelegate,memory_recursiveprot)
pstore on /sys/fs/pstore type pstore (rw,nosuid,nodev,noexec,relatime)
bpf on /sys/fs/bpf type bpf (rw,nosuid,nodev,noexec,relatime,mode=700)
systemd-1 on /proc/sys/fs/binfmt_misc type autofs
(rw,relatime,fd=29,pgrp=1,timeout=0,minproto=5,maxproto=5,direct,pipe_ino=27752)
mqueue on /dev/mqueue type mqueue (rw,nosuid,nodev,noexec,relatime)
hugetlbfs on /dev/hugepages type hugetlbfs (rw,relatime,pagesize=2M)
debugfs on /sys/kernel/debug type debugfs
(rw,nosuid,nodev,noexec,relatime) tracefs on /sys/kernel/tracing type
tracefs (rw,nosuid,nodev,noexec,relatime) tracefs on
/sys/kernel/debug/tracing type tracefs
(rw,nosuid,nodev,noexec,relatime) fusectl on /sys/fs/fuse/connections
type fusectl (rw,nosuid,nodev,noexec,relatime) configfs on
/sys/kernel/config type configfs (rw,nosuid,nodev,noexec,relatime) none
on /run/credentials/systemd-sysusers.service type ramfs
(ro,nosuid,nodev,noexec,relatime,mode=700) nfsd on /proc/fs/nfsd type
nfsd (rw,relatime) /var/lib/snapd/snaps/bare_5.snap on /snap/bare/5
type squashfs (ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/core_13741.snap on /snap/core/13741 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/core_13886.snap on /snap/core/13886 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/core18_2566.snap on /snap/core18/2566 type
squashfs (ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/gnome-3-28-1804_145.snap on
/snap/gnome-3-28-1804/145 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/gnome-3-28-1804_161.snap on
/snap/gnome-3-28-1804/161 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/gnome-3-38-2004_115.snap on
/snap/gnome-3-38-2004/115 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/gnome-3-38-2004_119.snap on
/snap/gnome-3-38-2004/119 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/gtk-common-themes_1534.snap on
/snap/gtk-common-themes/1534 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/gtk-common-themes_1535.snap on
/snap/gtk-common-themes/1535 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/hunspell-dictionaries-1-7-2004_2.snap on
/snap/hunspell-dictionaries-1-7-2004/2 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/meteo_108.snap on /snap/meteo/108 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/meteo_111.snap on /snap/meteo/111 type squashfs
(ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/tesseract_2504.snap on /snap/tesseract/2504 type
squashfs (ro,nodev,relatime,errors=continue,x-gdu.hide)
/var/lib/snapd/snaps/vobsub2srt_34.snap on /snap/vobsub2srt/34 type
squashfs (ro,nodev,relatime,errors=continue,x-gdu.hide) /dev/nvme4n1p1
on /var/snap/firefox/common/host-hunspell type ext4 (ro,noexec,noatime)
/dev/nvme4n1p2 on /home type ext4 (rw,relatime) /dev/sda1 on
/media/jjj/Data type ext4
(rw,nosuid,nodev,noexec,relatime,errors=remount-ro,user) binfmt_misc on
/proc/sys/fs/binfmt_misc type binfmt_misc
(rw,nosuid,nodev,noexec,relatime) sunrpc on /run/rpc_pipefs type
rpc_pipefs (rw,relatime) net_cls on /sys/fs/cgroup/net_cls type cgroup
(rw,relatime,net_cls) 192.168.1.164:/volume1/Synology on
/media/jjj/Synology2 type nfs
(rw,nosuid,nodev,noexec,relatime,vers=3,rsize=131072,wsize=131072,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=192.168.1.164,mountvers=3,mountport=892,mountproto=tcp,local_lock=none,addr=192.168.1.164,user)
tmpfs on /run/user/1000 type tmpfs
(rw,nosuid,nodev,relatime,size=3227132k,nr_inodes=806783,mode=700,uid=1000,gid=1000,inode64)
gvfsd-fuse on /run/user/1000/gvfs type fuse.gvfsd-fuse
(rw,nosuid,nodev,relatime,user_id=1000,group_id=1000) portal on
/run/user/1000/doc type fuse.portal
(rw,nosuid,nodev,relatime,user_id=1000,group_id=1000) tmpfs on
/run/snapd/ns type 

[PLUG] I hate the mount command

[John Jason Jordan]

A couple hours ago PGE did it to me again. This happens every couple of
months. The lights flickered off, and then immediately came back on.
Now, I have three massive APC backup units that run everything, even
the stereo. I mean, all the network components, both computers, the
Lenovo dock and everything connected to it. And when the power failed
they all remained running, except the enclosure that holds four 8TB
NVMe drives in RAID0 to give me a 32TB mdadm device, known as md0p1. It
auto-mounts via a line in fstab.

Unfortunately, after the power glitch I cannot access anything on the
device. The GUI file managers (PCManFM, Thunar, etc.) displayed the
folders, but the contents were blank. I could open a text file in
mousepad, but I could not save it because it was read only. How this
happened is a mystery; before the power glitch it was mounted r/w, as
it always is.

After a couple hours of trying to fix the problem I have managed to
umount the device. However, attempting to remount it has failed:

sudo mount -a
mount: /media/jjj/Movies: /dev/md0p1 already mounted or mount point busy



I also tried lsof, hoping that it might tell me what was hanging things
up, but I couldn't understand the results:

sudo lsof /dev/md0p1
lsof: WARNING: can't stat() fuse.gvfsd-fuse file system
/run/user/1000/gvfs Output information may be incomplete.
lsof: WARNING: can't stat() fuse.portal file system /run/user/1000/doc
  Output information may be incomplete

Final note: This is not the first time that this device has given me
grief. From past experience, if I reboot the computer it comes back
fine. At the moment I don't want to reboot. But if rebooting fixes the
problem, it makes me suspect mdadm. 

Suggestions welcome!

Re: [PLUG] PLUG Venue Options

[John Jason Jordan]

On Thu, 3 Nov 2022 15:18:08 -0700
Russell Senior  dijo:

>https://docs.google.com/forms/d/e/1FAIpQLSc58ccHKEQrlDoJDud-iRTw9I8jsHIAovb8-MAnGzFYahrD_A/viewform

Requires a Google account, which I don't have or want. But I would
answer Yes to all, with the comment that I need Max train access; buses
don't work well when you have to stand and wait in the rain at 10pm,
and then have to wait a couple times more for transfers. 

Re: [PLUG] Nordic, Russian keyboards

[John Jason Jordan]

On Tue, 25 Oct 2022 23:29:00 -0400
Tomas Kuchta  dijo:

>May I ask - what the . is Compose key? Is that Apple computer
>keyboard thing?
>
>I don't think I have ever seen/used keyboard with it. Perhaps, I am
>lucky??

It's a setting that you can add to most desktop environments. In Xfce
it's under Settings > Keyboard > Layout. This is also where you can set
up keyboards for other languages / countries. I don't know where other
desktop environments might had hidden the setting, but it's probably
somewhere similar. It's probably also accessible from the command line,
but I have no ken of that/.

Re: [PLUG] Nordic, Russian keyboards

[John Jason Jordan]

On Tue, 25 Oct 2022 08:29:05 -0700
Galen Seitz  dijo:

>On 10/24/22 22:05, John Jason Jordan wrote:
>...
>> I could do the same thing for any other language that requires
>> completely non-Latin characters. But I also sometimes write in
>> Spanish, French and German, and for those I just use the Compose key
>> rather than switch to a completely different keyboard. When I
>> enabled the Compose key I assigned it to the Windows key, otherwise
>> useless on Ubuntu. If I want (e.g.) an ä I hold down the Windows
>> key, type a double quote, then the a, and the process gives me an ä.
>> All I have to remember is that a double quote gives me an umlaut, a
>> single quote gives me an acute accent, a back quote produces an
>> accent grave, and an > gives an accent circumflex. There are lots
>> more possibilities, and I have memorized the sequences for a few
>> more, like ß, !, ¿, etc.  
>
>I just tried it here and it works as you described.  I wonder if this 
>works due to key rollover support. 

I wasn't sure sure what 'key rollover' meant, but the Wikipedia
article explained it. Thanks for the link.

While I have long used the Compose key, I should have mentioned that
when you enable it (turned off in Xubuntu by default) you discover that
it has disabled continuous entry of characters. That is, normally, if
you hold a key down, it enters the character repeatedly until you
release the key. That does not work for me, and it stopped working
right after I enabled the Compose key. I poked around trying to find a
way to get it working again, but failed. I just live without continuous
character insertion.

>I should have read the wikipedia article first.  Entering the 
>diacritical first allows for the use of more than one diacritical.  It 
>sounds like the Compose key may or may not be a modifier key.  It is 
>system or configuration dependent.

I knew that you can enter multiple diacritics, and occasionally I have
done so, because we need that now and then with IPA. The problem is
that in IPA the rules say that you can enter the diacritic above or
below the character, and the usual practice that I see when reading
academic journal articles is to put the 'ordinary' diacritic (as used in
many languages) above, and the special diacritic used only by linguists
below the character. Although rare, I have even seen three diacritics
on a character. However, I have never figured out how the Compose key
decides where to put the diacritic. 

I also always assumed that the Compose key works by selecting a
character that has the diacritics that the user has entered, but if the
font being used in the text does not contain that character, then it
enters the character from another installed font that does have the
required character, if available and, if not available it just enters
an empty space. This is very rare for me, because my workhorse font for
academic papers is Junicode, which has not only complete IPA
characters, but also characters for most major languages, plus
full combining diacritics. 

And if I need some diacritic(s) that are unavailable with the Compose
key, combining diacritics rescue me. To use them I just enter them with
Ctrl-Shift-u + hex code number. The only problem with combining
diacritics is that they have a uniform side bearing, which creates a
diacritic which is slightly off center when applied to m, w, or i, l.
In a full-fledged word processor or Scribus I can manually adjust the
position of the diacritic, but not here: e.g., a syllabic m comes out
m̩ because Claws Mail lacks entry for sophisticated typography.

Thanks again for the Wikipedia link. The article cleared up a few
questions for me.

Re: [PLUG] Nordic, Russian keyboards

[John Jason Jordan]

On Mon, 24 Oct 2022 20:57:13 -0700
Ali Corbin  dijo:

>On Mon, Oct 24, 2022 at 8:24 PM Keith Lofstrom 
>wrote: .
>
>> Does anyone on the plug list have experience using multiple
>> keyboards and alternate character sets?  Suggested vendors
>> for those keyboards?  Helpful Linux tools for linguistic
>> cripples?

>I regularly switch back and forth between Latin, Cyrillic, and
>(ancient)  Greek.  But I simply switch the layouts, using a single physical
>keyboard for each.  Which takes some memorization.  Since I never
>learned to touch-type in Russian, I can use a phonetic Cyrillic
>keyboard and mostly press the Latin key that sounds like the Cyrillic
>one.  I do have to memorize where the extra letters are, or bring up
>an image of the keyboard layout, or even bring up the character map
>and click the letters into the paste buffer.

That's pretty much how I do it for Ancient Greek (polytonic), although I
also often use Onboard, which came with Xubuntu. Onboard gives you an
on-screen keyboard so you can click on characters. I assigned
Shift-Ctrl-Alt to switch between normal US English keyboard and Ancient
Greek. When I'm in Ancient Greek the Onboard keys change to Greek, or I
can just use the keyboard on my Thinkpad, in which case I have to
remember which keys give me which Ancient Greek characters.

I could do the same thing for any other language that requires
completely non-Latin characters. But I also sometimes write in Spanish,
French and German, and for those I just use the Compose key rather than
switch to a completely different keyboard. When I enabled the Compose
key I assigned it to the Windows key, otherwise useless on Ubuntu. If I
want (e.g.) an ä I hold down the Windows key, type a double quote, then
the a, and the process gives me an ä. All I have to remember is that a
double quote gives me an umlaut, a single quote gives me an acute
accent, a back quote produces an accent grave, and an > gives an accent
circumflex. There are lots more possibilities, and I have memorized the
sequences for a few more, like ß, !, ¿, etc.

I also frequently need IPA characters. I could install a special
keyboard for them, but over years of using them I find it's faster just
to use Ctrl-Shift-u + Unicode hex number. The thing is that IPA
characters are needed just as a single character inserted into a paper
that is otherwise completely in English. Over the years I've memorized
the most frequently used ones, and I made a little popup message tool
that displays them when I need one that I haven't memorized. That is,
the popup is created by gxmessage, (in the repos) which I call with a
little bash script that also loads the text file that contains the data
to be displayed.

The nice thing about my setup is that I need no extra physical
keyboards. I can do everything quickly and easily with the US English
keyboard that Lenovo built into my current Thinkpad. And if I travel I
need bring no extra baggage in order to have everything available.

Re: [PLUG] Text color issue

[John Jason Jordan]

On Sat, 24 Sep 2022 14:46:54 -0700 (PDT)
Rich Shepard  dijo:

>I didn't think of looking at the man page for a solution. I was
>surprised to not find any text color 34 in /etc/DIR_COLORS and didn't
>think of changing the color in the v,t.

This is probably of no help to the problem, but I just wanted to
mention a trick that I discovered  long time ago for basic colors. One
day I was trying to tell a color wheel in some program I was using that
I wanted the color white, not some number that meant nothing to me; I
wanted plain 'white.' In frustration I just typed 'white' in the box
that was expecting a number, et voilà! I got pure white!

In later work I discovered that it only works for basic colors. I have
a house that is painted 'pale flowers,' the name of a paint color sold
by Lowes. If I typed that in the box I'm sure it might cause my
computer screen to melt. 

And I don't know where the translation to a number happens; in the
program I'm running? In Xfce4? I don't know. I also don't know the
extent of the colors that are available. Maybe there's a list somewhere
with the actual number that a given text entry gets translated to.

Re: [PLUG] Typing accented characters in text files

[John Jason Jordan]

On Thu, 15 Sep 2022 15:01:45 -0700 (PDT)
Rich Shepard  dijo:

>My locale is set to LANG=en_US.UTF-8. I write primarily in text-based
>applications such as emacs and alpine, and I want to learn how to enter
>accented characters so I can type, for example, Résumé rather than
>Resume.

Others have mentioned Ctrl-Shift-u, which gives you the opportunity to
enter the Unicode value for the character you want. I need a lot of
these for linguistics work, so I made up a little table with two
columns, one for the character as it appears and its name, and another
for the Unicode hex number. I found a little message utility that pops
up my list on the screen whenever I need to see it. 

But I also added the Compose key to my keyboard which is faster when
all I need are normal diacritics for things like ¡résumé! I assigned
the Windows key to be the Compose key, so to type an é I hold down the
Windows key, type an apostrophe, then type an e, and it comes out é.
The only problem with the Compose key is that setting it up disables
repeat typing. 

I also added a keyboard for Polytonic Greek, which I can activate by
clicking on a little flag in my panel - the American flag changes to a
Greek flag or back again, or I can switch it with Ctrl-Shift-Alt. 

[PLUG] Can't print from Chromium

[John Jason Jordan]

The printer is a Brother HL-2280-DW connected as 192.168.1.25. I can
ping it, so the connection is good. I can print to it from any computer
and from any program except Chromium. When I print from Chromium I look
at the GUI print queue for the printer, and no print jobs appear. In
other words, the problem is that Chromium is not even sending the print
job. A print job appears in the queue when I print from anything else.

I've poked all over Chromium's settings looking for some permissions or
restrictions issue, but I can't find anything. And Chromium obviously
has no problem accessing the network. It's just printing that doesn't
work. And I mean that it won't print from any web page. If I copy the
URL of a Chromium web page and paste it into Firefox, then the page
prints fine. Why would Chromium refuse to print?

Re: [PLUG] Vobsub2srt

[John Jason Jordan]

On Mon, 12 Sep 2022 14:54:24 -0700
John Jason Jordan  dijo:

>I remember when I first installed Ubuntu in 2005. At that time the
>repos had packages for just about everything, but in case you needed
>something that was available only in RPM, Ubuntu came with 'alien,'
>which could convert an RPM to a DEB. I used alien half a dozen times,
>and it always worked perfectly.

I did it! I found an RPM for vobsub2srt, downloaded it, alien-ed it to
a DEB, and used Gdebi GUI to install it. It now appears in Synaptic as
installed!

The bad news is that it still can't find all the pieces of tesseract
that it wants, and I don't know what happened to the snap version - did
it get overwritten by the Debian package, or are both installed, or
what?

Re alien, I should have mentioned this before. Once at the Clinic there
was a fellow with a Fedora computer who needed to install a program
that was available only as a DEB. Fedora didn't have anything like
alien, so I did an experiment. I downloaded the DEB file and did 'sudo
alien pkg_name.deb pkg_name.rpm. We both sat and watched my screen as
alien whirred, and then there appeared the RPM file. The fellow
installed it on his Fedora machine and it ran perfectly. We both sat
there giggling at our success. I never would have guessed that alien
would work in either direction.

Re: [PLUG] Vobsub2srt

[John Jason Jordan]

On Mon, 12 Sep 2022 15:01:14 -0500
Bill Barry  dijo:

>The instructions for installing in in Ubuntu 12.1 are given in this
>github repository
> https://github.com/ruediger/VobSub2SRT

Doesn't work on 22.04.

I had a PPA for vobsub2srt, but it was removed by the upgrade from
20.04 to 22.04. I re-enabled it, and now when I run sudo apt update I
get an Atlanta warning for 185.125.190.52 443, because that's what's
encoded in the PPA, and now Ubuntu needs . . . 53.

I decided to continue trying to build vobsub2srt from source, starting
by removing the snap package. 'Sudo snap remove vobsub2srt' exited
without error, but also without removing vobsub2srt. That is to say,
'vobsub2srt filename' still tries to execute and the man page is still
installed. Apparently installing a snap package is forever.

I remember when I first installed Ubuntu in 2005. At that time the
repos had packages for just about everything, but in case you needed
something that was available only in RPM, Ubuntu came with 'alien,'
which could convert an RPM to a DEB. I used alien half a dozen times,
and it always worked perfectly.

Note to Ubuntu: If you don't like Debian and apt, that's fine with me,
but give me an 'alien' so I can convert DEBs to snaps, or flatpacks, or
appimages, or whatever you think is God's gift to mankind. Don't
convert half the packages and leave the rest flopping around and
incompatible with everything else. Do it all the way or not at all.

Re: [PLUG] Vobsub2srt

[John Jason Jordan]

On Mon, 12 Sep 2022 14:57:57 -0400
Tomas Kuchta  dijo:

>The issue is your tesseract is snap and vobsub2srt is not. Snap
>packages cannot share libraries, that is the point of snaps.
>
>You need to:
>a) find snap version of vobsub2srt which includes tesseract libs.
>b) find not snap tesseract package.
>
>If your vobsub2srt is snap, do a) else b)

Actually, my tesseract is apt and my vobsub2srt is snap. And there is
no snap for tesseract and there is no apt for vobsub2srt. There are
also a couple of PPAs to install later versions of tesseract, including
5.0, but that doesn't sound like the right move.

Maybe I can find an old version of vobsub2srt from before snap existed.

Re: [PLUG] Vobsub2srt

[John Jason Jordan]

On Mon, 12 Sep 2022 11:54:06 -0400
Tomas Kuchta  dijo:

>Try creating link with the name to your current tesseract lib file in
>the lib directory.

Thanks for the suggestion. However, the error message doesn't say where
it is looking for the lib file:

vobsub2srt: error while loading shared libraries:
libtesseract.so.3: cannot open shared object file: No such file
or directory

A seach on libtesseract* in Catfish gave me:

/snap/vobsub2srt/34/usr/lib/libtesseract.so.3
/snap/vobsub2srt/34/usr/lib/libtesseract.so.3.0.4
/snap/vobsub2srt/34/usr/share/doc/libtesseract3
/usr/lib/x86_64-linux-gnu/libtesseract.a
/usr/lib/x86_64-linux-gnu/libtesseract.so
/usr/lib/x86_64-linux-gnu/libtesseract.so.4
/usr/lib/x86_64-linux-gnu/libtesseract.so.4.0.1
/usr/share/doc/libtesseract-dev
/usr/share/doc/libtesseract4
/var/lib/dpkg/info/libtesseract-dev:amd64.list
/var/lib/dpkg/info/libtesseract-dev:amd64.md5sums
/var/lib/dpkg/info/libtesseract4:amd64.list
/var/lib/dpkg/info/libtesseract4:amd64.md5sums
/var/lib/dpkg/info/libtesseract4:amd64.shlibs
/var/lib/dpkg/info/libtesseract4:amd64.triggers

I begin to see what happened - vobsub2srt was installed with snap,
which apparently put libtesseract.so.3 in a /snap folder (why??), and
tesseract doesn't look in there. I have little knowledge of how to
create links, but I do know how to copy and paste so, guessing that the
file should be in /usr/lib/x86_64-linux-gnu, I copied the file into
there, and that solved the problem. Except that now I have more things
that tesseract can't find. Stay tuned. :)

[PLUG] Vobsub2srt

[John Jason Jordan]

Vobsub2srt has been around for a long time to convert bitmap subtitles
to srt (text) subtitles. It uses tesseract for the OCR work. I have
tesseract installed  According to the man page

vobsub2srt --lang en filename

Is supposed to do the job, where 'filename' is the name given to each of
the .sub and .idx files, without appending the extension. This is what I
get:

vobsub2srt: error while loading shared libraries:
libtesseract.so.3: cannot open shared object file: No such file
or directory

I can't find libtesseract.so.3. The problem may be that vobsub2srt
hasn't been updated since the turn of the millennium, where tesseract
keeps coming out with updates.

Any suggestions?

Re: [PLUG] Fonts to me crazy drive

[John Jason Jordan]

On Mon, 29 Aug 2022 17:49:42 -0700
John Jason Jordan  dijo:

>The font (four .ttf files) resides in ~/.fonts. But it turns out that
>none of the fonts in that folder appear in applications, e.g., LO
>Writer. Looking at the results of 'fc-cache -fv' I find the line:
>
>   /home/jjj/.local/share/fonts/Unknown Vendor: skipping, looped
>   directory detected

I have a small clue. At the top of the ~/.fonts folder there is a
small file .uuid, and it opens in a text editor, displaying what looks
like a uuid code. More importantly, I cannot delete it; that is, if I
delete the file it instantly reappears. And when it reappears the uuid
code is different.

Furthermore, there is also a .uuid file in /usr/share/fonts, and
another one in each of its subfolders. And deleting any of them
produces a new one in a few seconds, always with a new uuid code.

WTH?

[PLUG] Fonts to me crazy drive

[John Jason Jordan]

Having completed the dist-upgrade from Xubuntu 20.04.x to 22.04.1 I am
in the process of tidying up. And over the years I have collected way
too many fonts, installed by programs whose developers think it's OK to
pollute their users' computers. For the past couple of hours I have
nuked about a third of the fonts, but in the process I have done
something evil to my main workhorse font, which I can no longer access.

The font (four .ttf files) resides in ~/.fonts. But it turns out that
none of the fonts in that folder appear in applications, e.g., LO
Writer. Looking at the results of 'fc-cache -fv' I find the line:

/home/jjj/.local/share/fonts/Unknown Vendor: skipping, looped
directory detected

OK, the vendor is Mark Shuttleworth. Isn't that good enough?

Suggestions?

Re: [PLUG] Still waiting on a venue... informal meetings for now?

[John Jason Jordan]

On Thu, 02 Jun 2022 18:31:12 -0700
Michael Rasmussen  dijo:

>On 2022-05-31 21:43, Michael Dexter wrote:
>> Hello all,
>>
>> I am told that it will be several months until PSU will welcome PLUG
>> back and in the interim, I welcome everyone to either join the
>> previous Jit.si URL at the regular meeting time:
>> https://meet.jit.si/PDXLinux
>>
>> ... or that someone is welcome to propose something new. I will need
>> to take a break from finding speakers until we have an in-person
>> venue.
>
>What about a library meeting room?
>There are five open until 8:00pm
>Midland
>Hollywood
>Hillsdale
>Gresham
>Belmont

Intriguing possibilities, but they close at 8, while our meetings always
go longer. On the other hand, what if we started at 6:30 instead of 7,
and speakers and public are told they must finish by 8, lest we all get
booted out on our tushies? I like that idea, because I always thought
it is difficult to get home from afters while TriMet is still operating.

And TriMet brings up another another issue - which library locations are
on a convenient TriMet stop? And convenient to a Lucky Lab?

Re: [PLUG] Wifi not working

[John Jason Jordan]

On Mon, 2 May 2022 14:23:21 -0600
Tomas Kuchta  dijo:

>On Mon, May 2, 2022, 13:23 John Jason Jordan  wrote:
>> I also question why I must be root to fly in an airplane.

>I do not need to be root to do this - because I am using network
>manager to manage networks on laptops. Maybe, you also use network
>manage - in which case you could install + start network manager
>applet to do all your network related settings without root.

Network manager was another thing I tried, but if it was installed the
installer did not create a menu item to launch it. 'No problem,' I
thought; I'll just launch it from the command line. But that failed
also, because every command I tried failed with 'no such file.' I found
several command suggestions on the net, but they also all failed.

I might have been able to find it in Synaptic, but by then I was too
frustrated to bother. A program is pretty useless if the user can't
launch it.

Re: [PLUG] Wifi not working

[John Jason Jordan]

On Mon, 2 May 2022 05:52:01 -0700 (PDT)
Rich Shepard  dijo:

>On Mon, 2 May 2022, Russell Senior wrote:
>
>> There is a switch or button on your computer that turns off wifi and
>> bluetooth, sort of an "airplane mode". Find it, and toggle it to the
>> other position, see if that fixes it.
>
>All my laptops had/have a tiny slide switch on the front edge. After
>accidently turning it off once I learned to ensure that it was on.

That happened to me many years ago with my second ever Linux computer.
I fumbled around for an hour trying to figure out what was wrong, until
I finally found the switch. So this time that was the first thing I
tried. It turns out that Lenovo decided to save a few pennies and left
out the switch. My P73 has no switch. And Lenovo says the way to turn
it on or off is in the Windows 10/11 taskbar. Sadly, the network icon
in the Xfce4 panel says nothing about wifi; when it's turned off it's
just not there at all.

Fortunately Ubuntu told me the solution:

sudo rfkill unblock wifi; sudo rfkill unblock all

It worked like a charm. I probably needed only the first command,
because I note that Bluetooth is now also turned on. The post also said
that airplane mode can get turned on automatically if there is a delay
in the connection so, bass ackward as that sounds to me - a delay should
restart wifi, not turn it off - it must be why Xubuntu turned it off.

I also question why I must be root to fly in an airplane. Maybe the TSA
is behind this. It sounds more logical that airplanes should emit a
signal that will turn off wireless connections in all computers within
range. Oh wait ... that would cost airplane manufacturers a few
pennies. Never mind.

[PLUG] Wifi not working

[John Jason Jordan]

It used to be that if wifi was not working you just did 'sudo wlan0
up.' Well, that doesn't work, but 'ifconfig wlan0' got me:

wlp82s0   IEEE 802.11  ESSID:off/any
  Mode:Managed  Access Point: Not-Associated   Tx-Power=off
  Retry short limit:7   RTS thr:off   Fragment thr:off
  Power Management:on

So apparently my wifi is now wlp82s0. What an improvement (not!).

So then I did:

$ ifconfig wlp82s0 up
SIOCSIFFLAGS: Operation not permitted
jjj@Devil-Thinkpad:~$ sudo ifconfig wlp82s0 up
SIOCSIFFLAGS: Operation not possible due to RF-kill

Now I'm stuck. What is RF-kill? Where did it come from and why? How do
I make it go away?

Re: [PLUG] useful/cheap DVI switch

[John Jason Jordan]

On Sat, 30 Apr 2022 19:18:18 -0700
Keith Lofstrom  dijo:

>Hardware tamed!  Back to Ubuntu distro software, misleading
>700 page Linux books, just plain evil advice from random
>idiots on the web, and the superb and generous guidance
>from the wizards of PLUG.

¡De acuerdo!

My big problem with advice from the web is that more often than not it
is so hopelessly out of date that it is useless. I can ask DuckDuckGo
to limit the search to the last month (or other time periods), but it
turns out that people somehow refresh their five-year old web page so
DDG thinks it was created only a few days ago. Result - I have to waste
hours wading through worthless junk.

PLUG saves my sanity!

Re: [PLUG] Ubuntu LTS upgrade timing

[John Jason Jordan]

On Wed, 13 Apr 2022 20:22:16 -0700
Keith Lofstrom  dijo:

>Questions for you folks with years of Ubuntu experience:
>
>1) How easy/fraught is a dist-upgrade, say 16.04 to 20.04?
>
>2) Should I wait for 22.04.1 (estimated 3+ months from
>now) before playing with 22.04 Jammy?

I left this thread alone for a while because I didn't want to get into
the dist-upgrade vs. fresh install argument that it invites. I'll just
say that the last time I did a fresh install it took eight days to get
the system to about 80% usability for all my needs.

At the same time I keep my systems up to date with the dist-update
feature. (Dist-update /= dist-upgrade.) Once in a while there is a
problem, but it's usually trivial. The dist-update feature is kind of a
compromise between a rolling release distro and no changes until there
is a new full release. Ubuntu does it by coming out with a stack of new
stuff every month or so, including sometimes even new kernels.

When I need to do a dist-upgrade (e.g., the impending need to upgrade
my Xubuntu 20.04.something computers to 22.04), things get a bit more
hairy. Before continuing I should state that Ubuntu will release 22.04
imminently as I write this (mid April, 2022), but only for fresh
installs. Ubuntu's policy is not to make the dist-upgrade available
until the first dot release. Usually that is a month or two later, but
for 18.04 to 20.04 it was not available until nearly October.

I should add that Ubuntu allows you to tell their Software Update app
not to bother you with dist-upgrades until the next LTS version is
available, or you can set it to announce it for every six-month upgrade.
My preference is for LTS only, but some people prefer to do the
dist-upgrade for each version, every six months for Ubuntu and Fedora.

I should also mention that everyone's needs are different. Many still
run releases that were released five and even ten years ago. I hate
having to upgrade the system, but I find myself having to decide
whether it is better to be gored from the front or from the back. If I
don't upgrade, apps slowly lose features or fail to run at all, and
worse, I can't install the latest version of an app. If I do upgrade
then other apps suddenly won't work, but I gain things that I couldn't
do before. Life is always six of one and half a dozen of another.

My experience over more than a decade of Ubuntu dist-upgrades is that
after a dist-upgrade there is always a day or two of patching stuff
that breaks. Usually this actually results in a better outcome anyway; I
just bitch at the time I have to spend putting my computer back
together. I should add that, although occasionally I have had serious
issues, I have never had a computer fail to boot after a dist-upgrade.

I guess, to summarize, the closer your computer is to what you get with
a fresh install, the fewer problems you will have with a dist-upgrade.
The more you are like me - hundreds of strange apps and configurations
- the more likely it is that you will feel pain. But then, you still
  save yourself the hassle of spending a week getting everything
  working like you need it to on a fresh install.

Re: [PLUG] Mouse glide repair

[John Jason Jordan]

On Wed, 30 Mar 2022 15:19:41 -0700
Keith Lofstrom  dijo:

>I still use ancient Logitech ball mice, three button plus
>scroll wheel, tethered to a USB-A connector.  They drive
>half a dozen linux computers around my home and office.

For many years I have used smallish Logitech wireless mousies that I
buy for ~$10 at Wally World. I've never had the feet fail, but they
develop other problems, mostly with the sensor and the scroll wheel.
I've never figured a way to repair these issues.

However, while shopping for my recently acquired Dell 2-in-1 confuser I
stumbled upon a similar sized mouse for $7.95 with free shipping on the
Lenovo eBay store. I bought one, and I kind of like it, more than the
Logitechs. Lenovo's regular price on eBay is $12.95, but periodically
they lower the price to $7.95. I waited until I could get the lower
price and then ordered a couple more.

I realize that I'm not very earth friendly by not spending a lot of
effort to repair these things, and my hat is off to those who do, but I
can do more for the earth in other ways that coast me less effort.

Re: [PLUG] Login and keyring ... why?

[John Jason Jordan]

On Sat, 26 Mar 2022 19:14:46 -0700
Keith Lofstrom  dijo:

I have been sorely vexed by the login keyring requirements, and created
the thread 'Login Keyring' on Saturday 22/2/12 which ultimately resolved
most of my issues. It included ~20 responses covering a lot of matters
that I previously had no ken of.

My problem was two of my computers which auto-boot without logins. I
did that because neither ever leaves the house, nor do they contain
sensitive information. It turns out the the keyring is set by the
login, so without a login I was constantly being harassed for the login
keyring password.

If you can find the thread it might be a useful read.

Re: [PLUG] Rotating greeter screen in Mate Ubuntu 20.04

[John Jason Jordan]

On Wed, 23 Mar 2022 19:44:07 -0700
Keith Lofstrom  dijo:

>How do I configure lightdm for a screen rotated right?
>
>However, the login screen for mate-ubuntu uses the lightdm
>display manager, and I haven't yet figured out how to
>rotate the screen for that login page.
>
>Frankly, I would prefer to log in from the same text
>screen as boot, THEN have a LOGIN SCRIPT fire up Mate
>instead of a GUI greeter, but I don't know how to make
>that happen either.

I have a similar problem with one of my computers, which has a 4K
screen. I use Xfce, but lightdm still produces the login screen, and it
is tiny, unreadable. Luckily, all I need do is type in my password and
hit enter, so it's not a huge deal, but it would be nice if I could
double the size of the minute login window.

I haven't tried doing this, but could you boot to a root prompt, log
in, then do 'start x' or whatever the current command to engage the GUI
is? Oh, and I tried this as a solution to my problem, but as soon as
the boot process starts the dmesg-type text scrolling past jumps down to
about 4-5 points, again, unreadable.

I think the problem is that the initial boot process assumes a VGA
display, and if you're using something higher, too bad for you.

Re: [PLUG] USB-A port colors and speeds

[John Jason Jordan]

On Tue, 15 Mar 2022 15:06:31 -0700 (PDT)
Rich Shepard  dijo:

>A new Probox support USB-3.0 or 3.1 speeds. I'll order a new one if
>the old one has failed.

I have a Mediasonic box that I formerly connected with its USB 3.0 port
with a 3.0 cable (everything blue). One day it started disconnecting on
me randomly, and eventually wouldn't connect at all. Next to its USB
3.0 port it had an eSATA port, so I bought an eSATA cable and used it
to connect the box to an eSATA port on my laptop. But then I got a new
laptop that lacked ] an eSATA port, so I contacted Mediasonic and for
about $20 they sold me a new card with both connectors on it, which I
used to replace the original card in the box. The card was just about
6cm square and replacing it was trivial.

I mention this experience just so you know that I had a good experience
with Mediasonic replacement parts.

I still have the original card, now with a dead USB port and a working
eSATA port. I saved it just 'cause I'm a clutter-butt.

Re: [PLUG] Linux and Windows VPN?

[John Jason Jordan]

On Mon, 14 Mar 2022 17:28:58 -0700
Dick Steffens  dijo:

>My wife is interested in VPNs, but she runs Win7. Is there a VPN that
>runs on both Win7 and Linux (Ubuntu based)?

There are a ton of VPN services available. Most will run on any OS, but
few have an application to configure their service, and of those that
do, even fewer have a Linux app. Private Internet Access (PIA) and
Mullvad are two that do. I started with PIA and later switched to
Mullvad. Mullvad provides a .deb package and PIA has a script, and both
were just a couple of clicks to install.

Another thing to consider is that running a VPN will kick your
bandwidth in the tush. My gigabit fiber connection gives me about
800Mb/s normally, and with PIA it dropped to about 250, and with
Mullvad to about 300-350.

I should add that browsers report your location to web sites. Once I
tried to open safeway.com and was greeted with 'access denied" because
I was connecting via a Mullvad server in Stockholm. Of course, a big
outfit like Safeway has to be extra careful, because we all know that
everyone in Sweden is a terrorist. When I'm sitting at my desk my main
laptop computer has the VPN, but I have a desktop computer right next
to it, and I deliberately did not install the VPN on it. If I have an
urgent need to view safeway.com I can swivel my chair to the desktop.

You should also be aware that there are many VPN services just for web
browsers, and many of them are free. They are available as extensions.
Of course, they give you privacy and protection only for browser
activity. Plus, if you use more than one browser you'd need to add the
extension to all of them.

>Is it worth the effort?

Depends on how much you value you privacy.

Re: [PLUG] Replacement laptop

[John Jason Jordan]

On Fri, 25 Feb 2022 20:30:13 -0500
TomasK  dijo:

>I cannot recommend Dells shipping with Linux right now - they ended up
>being devoid of even basic number of USB ports recently. If you do not
>need more than 1 or 2 USB-C ports (one taken by the charger) - and can
>receive Fed-Ex delivery (I cannot - no office front) - Dell's XPS13 are
>pretty nice computers too.

Yeah, current Dells are short on USB ports - my new Latitude 7275 has
only two, and one is for power. Not only that, they are right next to
each other. I had to invest in a dock; inconvenient for a laptop. Not
only that, no ethernet port.

Re: [PLUG] Login keyring

[John Jason Jordan]

On Mon, 14 Feb 2022 23:20:59 -0800
John Jason Jordan  dijo:

>On Mon, 14 Feb 2022 17:23:00 -0800
>wes  dijo:
>
>>if you right-click on the keyring, do you get any options around
>>setting a password?
>
>It just pops up the box to enter the password, and if I leave it blank
>and then click on OK it gives an error message that I didn't enter the
>right password. It won't take a blank password.
>
>I can't find an option to launch Seahorse from the menus, but
>'seahorse' from the command line did the trick. It provides a window
>titled 'Passwords and Keys,' which gives me a panel on the left:
>
>Passwords
>   Default keyring
>   Login
>Keys
>   GnuPG keys
>   OpenSSH keys
>Certificates
>   System Trust
>
>Clicking on 'Default keyring' gives me a panel on the right with an
>image of some keys, underneath 'Keyring is locked,' and below that a
>button 'Unlock.' Clicking on Unlock prompts me to enter my password,
>but if I leave it blank it gives an error message that the password was
>not correct.
>
>Clicking on any of the other items gives me a window on the right that
>says 'This collection seems to be empty.'
>
>The Seahorse GUI seems unable to accept an empty password.

I should have added that I can right-click on Login and there is an
option to change password that pops up a box for the new password. I
have to enter the old password first, then Continue, and then the new
password. I can leave the new password blank, but then it says if I
leave it blank I have no security and anyone can change my password. I
left it at that, so maybe I solved this problem.

Re: [PLUG] Login keyring

[John Jason Jordan]

On Mon, 14 Feb 2022 17:23:00 -0800
wes  dijo:

>if you right-click on the keyring, do you get any options around
>setting a password?

It just pops up the box to enter the password, and if I leave it blank
and then click on OK it gives an error message that I didn't enter the
right password. It won't take a blank password.

I can't find an option to launch Seahorse from the menus, but
'seahorse' from the command line did the trick. It provides a window
titled 'Passwords and Keys,' which gives me a panel on the left:

Passwords
Default keyring
Login
Keys
GnuPG keys
OpenSSH keys
Certificates
System Trust

Clicking on 'Default keyring' gives me a panel on the right with an
image of some keys, underneath 'Keyring is locked,' and below that a
button 'Unlock.' Clicking on Unlock prompts me to enter my password,
but if I leave it blank it gives an error message that the password was
not correct.

Clicking on any of the other items gives me a window on the right that
says 'This collection seems to be empty.'

The Seahorse GUI seems unable to accept an empty password.

Re: [PLUG] Login keyring

[John Jason Jordan]

On Mon, 14 Feb 2022 12:24:09 -0800
wes  dijo:

>On Mon, Feb 14, 2022 at 11:04 AM John Jason Jordan 
>wrote:
>
>> On Sun, 13 Feb 2022 17:14:19 -0800
>> wes  dijo:
>>
>> >On Sun, Feb 13, 2022 at 12:49 PM John Jason Jordan 
>> >wrote:
>> When I ran ifconfig it gave me enx+long-number which was the
>> ethernet, and no IP address for it, and it gave me virbr0 with an
>> address of 192.168.122.1. This has to be the wifi, but where did the
>> 122 come from? My network is192.168.1.x.
>
>
>virbr0 is for virtualization software. virtualbox, vmware, etc. that's
>where the 122 comes from.

That's interesting. Of course, I spent hours yesterday fiddling with
VirtualBox, and finally got it working. But since last night it
has been shut down. And just to be sure there are no virtual machines
still running, just now I rebooted, but ifconfig still shows virbr0,
with the same funny IP address. Is it possible that VirtualBox running
Windows 10 (the only machine that I ran) created this 'device' for the
machine? And if it did, when I shut down Windows 10 and VirtualBox, why
didn't it delete it? I might add that it doesn't appear on my other two
computers.

I thought it might be a neighbor's wifi, so I tried iwlist. It showed
half a dozen wifi networks (including my own), but iwlist doesn't
list the IP address of the network, and ifconfig doesn't display the
name for 192.168.122.1.

Before running iwlist I tried the network icon in the panel, because if
you click on it you can see all the networks available. But it's not in
the panel on the Latitude, nor is it available to be added to the
panel. I can't find it in Synaptic. It's in the panel on my other
Xubuntu computers, where it appears in the Notification Area panel
plugin. Apparently on Xubuntu 21.10 the Notification Area has been
removed. Xubuntu 21.10 does have a penal plugin for the network, called
'Net,' but if you click on it all you get is the name of the connection
device and the last four messages - no options to connect to other
networks, list of available networks - basically useless.

I also tried to run network-manager, which Synaptic says is installed,
but evidently it never occurred to the writers of the documentation to
tell users HOW TO LAUNCH IT! I tried nm, network-manager,
network-manager-applet (with and without dashes), and numerous others,
all of which return 'no such file.'

Meantime, I can ping 192.168.122.1, although the ping command just says
that it connected and the time it took for the response. And I tried to
connect to it by putting it in the URL bar in Chromium, but got
'refused to connect.'

>> At this point I was finally able to launch seahorse. I explored it a
>> bit, but couldn't figure out how to use it. One of the buttons that I
>> clicked on displayed 'Keyring Locked.' I could probably have unlocked
>> it, but I left it for now because I want it to be unlocked at login,
>> but that option must be buried someplace that I didn't look at.

>there isn't a specific option to have the keyring unlocked by default.
>you have to implicitly do this by unlocking the keyring, and then
>setting the password to empty.

I click on Unlock Login Keyring and it just asks for my password. If I
enter it all it does is unlock the keyring; there is no option to set
the password to empty, although there is an option 'Unlock this keyring
when I'm logged in.' But the checkbox has always been selected, yet it
still asks me repeatedly to unlock the keyring.

I mostly agree with points about security from Tomas. The Latitude will
be used just for reading electronic books and articles; it has no
sensitive information on it, not even my bookmarks for Firefox and
Chromium. Security needs are negligible. Yet Ubuntu forces me to give
the secret handshake for permission to go to the bathroom. Bah.

Re: [PLUG] Login keyring

[John Jason Jordan]

On Sun, 13 Feb 2022 17:14:19 -0800
wes  dijo:

>On Sun, Feb 13, 2022 at 12:49 PM John Jason Jordan 
>wrote:
>
>> Apparently Xubuntu is supposed to be using seahorse, but it's not
>> installed on my Lenovo laptop with 20.04.2 nor on my new Latitude
>> with 21.10. I tried the link, but again it says to use seahorse. I
>> can install seahorse, but must be something else that is causing the
>> problems.

>now that you mention it, this triggers my memory. I have always had to
>install seahorse to solve this problem.
>
>I don't think changing your user account's password will help. you
>need to change the keyring password (which is frequently the same as
>the login password for convenience, but is not necessarily). this is
>what seahorse allows you to do

Over the weekend I poked at a lot of things, without luck So this
morning I installed seahorse, and seahorse-daemon. Not sure id the
daemon was required, but I did it anyway. I did this from the command
line, and I got a lot of errors, mostly missing packages. I launched
Synaptic and did a Reload, which also produced a lot of errors. Back to
the command line I entered ifconfig, which produced more strange
results.

Note that this new Latitude has wifi, but no built-in ethernet port,
although I have a USB 3.0 'dock' plugged in that provides an ethernet
port. To use it I have to run a cable from another room, and to avoid
having a wire to trip over I have always used just the wifi with the
Latitude. When I ran ifconfig it gave me enx+long-number which was the
ethernet, and no IP address for it, and it gave me virbr0 with an
address of 192.168.122.1. This has to be the wifi, but where did the
122 come from? My network is192.168.1.x. Well, at least it explains why
apt couldn't download anything. While waiting for my brain to come up
with an explanation for the 122 I plugged in an ethernet cable, and
then I was able to connect, and apt properly installed seahorse and
seahorse-daemon.

At this point I was finally able to launch seahorse. I explored it a
bit, but couldn't figure out how to use it. One of the buttons that I
clicked on displayed 'Keyring Locked.' I could probably have unlocked
it, but I left it for now because I want it to be unlocked at login,
but that option must be buried someplace that I didn't look at.

Using the Latitude I logged into the D-Link DIR860L B1 and noted that
there is a connection at 192.168.1.171 marked 'Latitude,' that must be
the ethernet. Of course, 192.168.122.1 did not appear.

So now I have two questions:

1) How to use Seahorse to get rid of the keyring hassle.
2) What the heck happened to make the wifi on the Latitude think that
its network was 192.168.122.x?

I could use some suggestions.

Re: [PLUG] VirtualBox Problems

[John Jason Jordan]

I have finally succeeded. It took seven hours to get it working. During
that time I also installed KVM, which I had never heard of before, and
tried to install Windows 10 Pro 64-bit, using the same install media
(ISO file) that I had used to install it on VirtualBox several years
ago. I had also tried to install it under VirtualBox on the new
Latitude, and it appeared to install, but when I tried to start it I
got a black screen with the message that it couldn't find anything to
boot from. I got this same error message with the KVM install.

Note that in the past whenever I had a new install of whatever the
current version of Xubuntu was I just installed VirtualBox from the
repository, which created ~/.VirtualBox/VDI (and many more folders),
and then copied the .vdi files from the old computer to the /VDI folder
on the new installation. When I launched the new VirtualBox there would
be all my machines, working perfectly. But this time that didn't work.

I also noted that the new version defaults to importing an OVA or OVF
file which was exported from the old VirtualBox/VDI folder. I
successfully exported my old .vdi file, both as OVA and OVF, but the
import feature on the new install always ended in an error message that
the import had failed.

The problem was exacerbated by the fact that the old installation was
VirtualBox 1.6.16 on Xubuntu 20.04.2, and the new one was version
1.6.26 on Xubuntu 21.10. Between 1.6.16 and 1.6.26 Oracle completely
changed the configurations, including putting the .vdi files into their
own folder ~/VirtualBox VMs/. I also tried putting my .vdi files into
the new folder, but they still didn't appear.

Eventually I found a place in Preferences for the new 1.6.26 where I
could specify the location for .vd files. The default was ~/VirtualBox
VMs, so I changed that to ~/.VirtualBox/VDI, and used that as the
folder to copy my .vdi files to, but they still didn't appear.

Finally I wiped out all the configurations for the new VirtualBox, then
copied the entire ~/.VirtualBox folder (recursively) from the old
computer to a USB stick, and then pasted the whole thing (105GB) to ~/
on the new computer. When I launched VirtualBox finally all my old
machines appeared. I tried Windows 10 and it ran perfectly as it always
did on the old computer.

Note that each of these attempts took from 15-20 minute to over an
hour, partly because of time spent browsing the net for howtos, and
partly for copying the files. It takes a long time to copy 105GB from
the old computer to the new one. It was a PITA, but it's finally done.

Thanks for all the suggestions!


On Sat, 12 Feb 2022 14:47:00 -0500
Tomas Kuchta  dijo:

>I have no intention to judge other people choices - just my opinion +
>this is pLUG.
>
>No usb issues with Virtual Box here for about decade now - using Oracle
>repo + virtualBox Additions.
>
>I have no issues using usb with KVM either. I feel KVM is better and
>more performant than VirtualBox these days.
>
>Given my bad and costly experience with various commercial SW quality
>and/or long term support, both personal and at work ---> I am of the
>opinion that commercial software does not have positive ROI, fiscally
>and intellectually - if there is a choice to be made. Most commercial
>solutions look great on paper, if you can find what they are suppose
>to do at all, but . Clearly commercial SW companies know it -
>hence the shift to subscriptions, lobbying, competition removal, and
>various lock-in schemes.
>
>My 2c - I would not touch VMware again, and vBox either, if not for
>legacy reasons. KVM works and will keep working better.
>
>-T
>
>On Sat, Feb 12, 2022, 10:44 Chuck Hast  wrote:
>
>> VirtualBox USB issues drove me to VMWare, and
>> since for personal use it is free, I started using it.
>>
>> Never had any issues moving a VM from one
>> VMWare machine to another.  I never tried VB
>> again due to the USB issues I encountered.
>>
>> On Fri, Feb 11, 2022 at 11:21 PM John Jason Jordan 
>> wrote:
>> > I discovered that I can export from the old VirtualBox to OVF
>> > format just by manually changing the name of the file it will
>> > create from .OVA to .OVF. And the new VirtualBox on the Latitude
>> > will import it, except that at the end the import fails:
>> >
>> > VirtualBox Error
>> > Failed to import appliance ... Windows 10.ovf
>> > Detail
>> > Result code NS_ERROR_INVAL ID_ARG (0x80070057)
>> >
>> > I tried it both ways - exporting as OVA and exporting as OVF, and
>> > the new VirtualBox will try to import either one, but at the end
>> > always gives the above error message.
>> >
>> > I also tried copying the Windows 10.vdi file to ~/VirtualBox/VDI on
>> > the new computer (where I had to cr

Re: [PLUG] Login keyring

[John Jason Jordan]

Apparently Xubuntu is supposed to be using seahorse, but it's not
installed on my Lenovo laptop with 20.04.2 nor on my new Latitude with
21.10. I tried the link, but again it says to use seahorse. I can
install seahorse, but must be something else that is causing the
problems. And seahorse is apparently a front end for gnome-keyring or
gnome-keyring-daemon, but those aren't installed either. I went into
Users and Groups where I am listed, but I couldn't find any settings
for password, so maybe that's not where I should be looking.


On Sun, 13 Feb 2022 11:57:23 -0500
Brian Stanaland  dijo:

>Ubuntu uses the keyring to store passwords for all kinds of things.
>It looks like Xubuntu uses seahorse or gnome-keyring as the keyring
>manager. I'm running Kubuntu so it uses KDE Wallet.
>Anyway, if you open whichever one Xubuntu uses, there should be either
>a "login" or "wallet" password stored. By default it uses the user
>account password and just opens when you log in. Since you don't enter
>the password to log in, the wallet doesn't get opened. Setting the
>login password there as blank should fix the problem.
>This post is 7 years old but it should still be close to what you'll
>see.
>https://askubuntu.com/questions/533324/change-keyring-password-on-xubuntu#533326
>
>Brian
>
>On Sun, Feb 13, 2022 at 10:59 AM Tomas Kuchta
> wrote:
>
>> Such is life with desktop/server these days. It is pretty annoying
>> that the security zealots who implemented authentication for just
>> about anything (filesystem, video, sound, usb, applications, etc.)
>> on your system did not think/care of this.
>>
>> Anyway, the desktop login dialog unlocks keyring for you. If you
>> disable it, this is the outcome. I am not familiar with xubuntu -
>> there is probably a way to unlock the ring by authenticating with
>> some special app...
>>
>> This of course defeats the purpose of no-login dialog. The easiest is
>> probably to enable login and set empty password - that is if the
>> security zealots did not enforce password complexity.
>>
>> Hope that helps,
>> -T
>>
>> PS: Another annoying example: On most systems today, even if you
>> arhenticate, you cannot play sounds or display anything remotely.
>> Meaning that you cannot practically turn modern linux to a media
>> player. One has to turn to not secure, special flake, totally
>> unsecured distros
>>
>> On Sun, Feb 13, 2022, 02:02 John Jason Jordan  wrote:
>>
>> > I have my desktop computer and my new laptop set up to boot Xubuntu
>> > without requiring a login. That part works perfectly, but the
>> > instant I try to do anything I get a popup:
>> >
>> > Authentications required
>> > The login keyring did not get unlocked when you logged into
>> > your computer
>> > Password: [ ]
>> >
>> > What's the point of booting without logging in if I have to log in
>> > to do anything? Not only that, if I enter my password to satisfy
>> > the login keyring (whatever that is), the next time I try to
>> > launch an application I have to login again ... and again ... and
>> > again.
>> >
>> > On my main computer I set it up to require a login, but the
>> > desktop and the new laptop don't have anything sensitive on them
>> > and they never leave the house.
>> >
>> > Is there any way to get rid of the stupid keyring requirements?

[PLUG] Login keyring

[John Jason Jordan]

I have my desktop computer and my new laptop set up to boot Xubuntu
without requiring a login. That part works perfectly, but the instant I
try to do anything I get a popup:

Authentications required
The login keyring did not get unlocked when you logged into
your computer
Password: [ ]

What's the point of booting without logging in if I have to log in to
do anything? Not only that, if I enter my password to satisfy the login
keyring (whatever that is), the next time I try to launch an
application I have to login again ... and again ... and again.

On my main computer I set it up to require a login, but the desktop and
the new laptop don't have anything sensitive on them and they never
leave the house.

Is there any way to get rid of the stupid keyring requirements?

Re: [PLUG] VirtualBox Problems

[John Jason Jordan]

I discovered that I can export from the old VirtualBox to OVF format
just by manually changing the name of the file it will create from .OVA
to .OVF. And the new VirtualBox on the Latitude will import it, except
that at the end the import fails:

VirtualBox Error
Failed to import appliance ... Windows 10.ovf
Detail
Result code NS_ERROR_INVAL ID_ARG (0x80070057)

I tried it both ways - exporting as OVA and exporting as OVF, and the
new VirtualBox will try to import either one, but at the end always
gives the above error message.

I also tried copying the Windows 10.vdi file to ~/VirtualBox/VDI on
the new computer (where I had to create the folder first), but the new
VirtualBox does not see the machine. I also tried putting it in
~/VirtualBox VMs, which is the default folder that the installer
created on the new computer, but it still is not visible.

At the end I decided to just reinstall Windows 10 in VirtualBox on the
new computer from the .ISO that I still have. It took a couple of hours
of swearing, but I finally got it installed, except when I launch it I
get a screen with a lot of little squares instead of a Windows 10
screen. In VirtualBox the icon for Windows 10 appears and the Settings
to the right also appear, and appear correct, but I can't figure out
how to fix the video problem.

TomasK  dijo:
>So, you need OVF to import it?
>Try export Virtual Appliance from menu then select one of the OVF
>formats.

The format is not selectable, but (see above) the file that it will
create appears in a box below and you can just edit the name to .ova or
.ovf.

>I do not bother - just:
>  rsync -a "VirtualBox VMs" .../home/$USER/
>  rsync -a .config/VirtualBox .../home/$USER/.config/
>
>Then start your new virtualbox and you should see everything as it used
>to be.

I don't understand the rsync commands above. The .vdi files are on one
computer and they need to be copied to a new computer, and the
computers are not directly connected. It is faster and easier just to
copy files from the old computer to a USB drive and plug the drive into
the new computer. And I already did that on the new computer, both to
~/.VirtualBox/VDI (which I had to create) and to ~/VirtualBox VMs
(created by the installer of 6.1.26).

Tomorrow I'm going to scour the VitualBox forums hoping to find step by
step 'for dummies' instructions for how to install Windows 10 from an
ISO.

I should add that this is not a crucial matter. My old computer has
Windows 2000, Windows XP and Windows 10, and none of them have been
run for at least a couple of years; Windows 2000 for at least five
years. If I had any brains I'd just forget about it.

I should add that the old computer has Xubuntu 20.04.2 with VirtualBox
6.1.16 and the new computer has Xubuntu 21.10 with VirtualBox 6.1.26.

[PLUG] VirtualBox Problems

[John Jason Jordan]

I have had VirtualBox on my Lenovo laptop and its predecessors for many
years, and it always just works. Now that I have a new computer I
decided to install it there also, and copy the VDI file for Windows 10
over there also (I have multiple licenses). In the past this always
worked when I needed to make a fresh install.

Installing VirtualBox on the new computer was no problem, but when I
tried to copy the VDI file the new installation it will only let me
import OVF files, which are apparently an exported version of your
existing VDI files. After much muttering I swiveled around to my Lenovo
laptop to export the VDI file, only to discover that VirtualBox was not
installed. All the VDI files are still there in ~/.Virtualbox, but the
program just ain't there.

I did not uninstall it. Something else uninstalled it, but what? Why?
How?

I opened Synaptic and installed it on the Lenovo, and when I opened
VirtualBox there were all my virtual machines, including Windows 10.
The reinstalled program found the machines in ~/.VirtualBox. But there
is yet another problem. The version that I just installed on the Lenovo
laptop will only export in OVA format, and the version on the new
computer will import only OVF files. And both versions of VirtualBox
control the file manager so if it wants an OVF file that's all that it
allows to appear in the folder.

More putzing around. :(

Re: [PLUG] usb-storage process continues to run

[John Jason Jordan]

On Fri, 4 Feb 2022 13:19:23 -0800 (PST)
Rich Shepard  dijo:

>On Fri, 4 Feb 2022, Rich Shepard wrote:
>My limited understanding of RAIDs is that striping would increase
>writing speed but lose the redundance of the hdds themselves. Am I
>mistaken?

That is sort of true. With two physical disks RAID0 turns them into one
device. In theory your OS reads/writes to both at the same time,
speeding things up by a factor of two. In reality the speedup is far
less than two. And since both disks have been turned into one device,
you have no place to put a mirror. And another bad point about RAID0 is
that if one disk fails, you lose both disks, because half of file was
written to one disk and the other half to the other. Trying to recover
means you have to match all the parts; not generally possible.

All my RAID arrays are RAID0 because each one is mirrored nightly to
another RAID0 of the same size.

[PLUG] Can't shut down

[John Jason Jordan]

I my travails with the new Dell laptop, while trying Mint I clicked on
something and the panel disappeared. It was Mint Xfce so I tried
xfce4-panel -r, but that did nothing. Then I decided to try the Install
Mint button, and suddenly the screen blacked out with text rolling down
saying

1529.99 SQASHFS error squashfs_read_data failed to read
block0x40f81eeb

The block number on the right remains constant, but the left number
increments with each new line, with a new line every millisecond or so.

I can't get a terminal open, and even the power button won't function.
It looks like I'm just going to have to sit here until the battery runs
down.

Maybe I should just throw it off a bridge.

Re: [PLUG] Can't get into BIOS, Dell Latitude 7275

[John Jason Jordan]

On Thu, 3 Feb 2022 16:11:11 -0800
John Jason Jordan  dijo:

>>Does anyone know the secret way to boot to the BIOS?
>
>Never mind, I finally succeeded. Xubuntu Live is running. The
>touchscreen and the wifi are both working, so I'm going to hit the
>Install icon.

A slight problem before I install Xubuntu: With Xubuntu Live the
touchscreen is working, but when I rotate the screen it stays in
landscape mode. Not good for reading books and journal articles.

Before I bought this computer I asked the eBay seller if the screen
automatically rotated when you hold it portrait, and he said that it
did. Of course, all he had was Windows 10. And it does work with
Windows 10, which I verified the few times it started to boot Windows
before I finally got the BIOS/UEFI straightened out.

Does anyone know of a simple utility to make Xubuntu automatically
rotate the 1920x1080 screen? The machine must have a sensor that
Windows gets its orientation information from, but I don't have any
idea how to access it. All I can think of is xrandr, but that's kind of
clumsy if you have to switch back and forth.

Re: [PLUG] Can't get into BIOS, Dell Latitude 7275

[John Jason Jordan]

On Thu, 3 Feb 2022 16:41:32 -0800
VY  dijo:

>Can you share steps how you get into the BIOS to boot from a linux
>live USB?
>
>I have one coming soon.
>I am a bit concern that these newer laptops may have something
>incompatible so I also want to try with a live USB first.
>( I am planning to use Linux Mint)

It's kinda tricky because it doesn't always work. My technique is to
shut the computer completely down and then, with the Xubuntu Live USB
plugged in, hold down F12 and keep holding it down while pressing the
power on button. If it works the screen shows the big Dell logo while
in the upper right corner it says 'preparing boot options.' A few
moments later you get a GUI showing the computer's hard drive and
'Windows,' and under it the USB stick with the identification that the
burn program put on it, 'Xubuntu 21.10' in my case. The touchscreen was
working, so I just selected Xubuntu, et voilà, It booted the Live
distro. 

Re: [PLUG] Can't get into BIOS, Dell Latitude 7275

[John Jason Jordan]

>Does anyone know the secret way to boot to the BIOS?

Never mind, I finally succeeded. Xubuntu Live is running. The
touchscreen and the wifi are both working, so I'm going to hit the
Install icon.

[PLUG] Can't get into BIOS, Dell Latitude 7275

[John Jason Jordan]

My Dell Latitude 7275 2-in-1 was delivered an hour ago. My plan was to
boot to an Xubuntu 21.10 live USB to figure out if it would even work
and, if it does work, then install it, wiping out the Windows 10
already installed.

I have the Xubuntu USB plugged in, but the laptop just boots directly
to Windows. According to Dell Support for the Latitude 7275:

https://www.dell.com/support/kbdoc/en-us/000131221/how-to-boot-to-an-external-device-using-usb-type-c-connection

Please enable below two options to support boot from external
device.

At boot, Press the F2 key (or alternately press the F12 key
then select the option to enter the BIOS setup). In POST
Behavior, Select - Fastboot the select the Thorough option
(Figure 1):

I tried tapping F2 repeatedly after pressing the power button, but it
has no effect - it still goes straight into Windows. I also tried
holding F2 down continuously, but it still boots directly to Windows.
And I did the same using F12, with no success.

Does anyone know the secret way to boot to the BIOS?

Re: [PLUG] Left over "artifact?" when doing screenshots, sometimes

[John Jason Jordan]

On Sun, 23 Jan 2022 19:56:47 -0800
Dick Steffens  dijo:

>Caja is the GUI file browser used in Linux Mint Mate.
>> because the file existed for a split
>> second. Caja was notified of the file's creation, but not of the
>> file's subsequent removal. this is why it doesn't show up in
>> terminal: it's actually not there. Caja may maintain some sort of
>> cache or running tally of which files exist, which it updates
>> periodically. I would guess that if you just wait a while, the .part
>> file will probably go away. it may also be possible to instruct Caja
>> to refresh its index without rebooting.

>It doesn't go away until after a reboot. I haven't tried anything
>else, though.

>Ignoring it works. I have enough screen real estate to shove it into a
>corner. I'll probably put a little time into seeing if there is a way
>to refresh something in Caja, but it's not a high priority. If
>something turns up, I'll report back.

I have Caja installed, but I seldom use it because I prefer PCManFM or
Thunar. But I get the same .part files that don't exist, usually on a
long download. The solution is to go to View > Reload Folder. It takes
only a second or two, and all the funny pieces disappear. As you've
discovered, rebooting also solves the problem, but reloading the folder
is far less work.

Re: [PLUG] Using crosh shell on a Chromebook ( was: Chromebooks and Linux )

[John Jason Jordan]

On Thu, 20 Jan 2022 23:23:06 -0700
Robert Citek  dijo:

>Here's the output from 'free' in crosh with the Debian VM running:
>
>crosh> shell
>chronos@localhost / $ sudo su -
>localhost ~ # free -tm
>  totalusedfree  shared  buff/cache
>available
>Mem:   38062732 231 287 842
>369
>Swap:  557523523222
>Total: 938150843454
>
>And here's 'free' from within the Debian VM:
>
>rwcitek@penguin:~$ free -tm
>   totalusedfree  shared  buff/cache
>available
>Mem:2777   82744   0  25
> 2769
>Swap:  0   0   0
>Total:  2777   82744
>
>This is on a Dell 3100 2-in-1.  So, this is an Intel Celeron and not a
>Snapdragon CPU.
>
>localhost ~ # grep name /proc/cpuinfo | uniq -c
>  2 model name  : Intel(R) Celeron(R) N4020 CPU @ 1.10GHz
>
>What are you wanting to do in Linux?  I'd be happy to try it out.

That looks like I might be able to run Debian in a container.

Another couple of questions have occurred to me:

* If you have the laptop folded over, in 2-in-1 mode, how can you hold
  it up in the air (like you were reading a book sitting in a
  recliner), without your fingers pressing on the keys in back?

* If you are holding it up in portrait mode, does the display
  automatically go portrait as well?

Re: [PLUG] Using crosh shell on a Chromebook ( was: Chromebooks and Linux )

[John Jason Jordan]

Robert,

The Chromebook that is at the top of my list at the moment has only 4GB
RAM. Installing and running Debian (or others) in a container worries
me. When you are running Linux in a container, how much RAM is taken in
all by Chrome OS and Debian?

I plan to use the tablet mostly for reading electronic editions of
books and academic journals. I would install an SD card for storage,
and I would never store things online or read things online. I would
boot first to Chrome OS and then to Linux, which is all I would use. The
tablet would seldom leave the house and would not be used for e-mail or
web browsing.

I realize that many consider me crazy to refuse storing things in the
cloud, but my brain is already plenty cloudy without that. When I read
something I want a copy of it on my own computer, not someplace in the
cloud that I have no control over.

On Thu, 20 Jan 2022 16:09:06 -0700
Robert Citek  dijo:

>Same scenario here.  I use the Chromebook for its standard features:
>browsing, email, chat, music. I use the Linux VM ( container? ) with
>Debian to do some light work in the shell or in a Docker container.
>For heavier loads, I use the cloud ( AWS, GCP, Azure, etc. ). I have
>Docker containers that have the corresponding CLI tools for working
>with those APIs. I just pass in my credential and run with it. For
>example, I recently had to do some work in Windows 11. I fired up an
>instance on Azure via the API and connected to it using the RDP client
>on the Chromebook.
>
>One feature I really like about the Chromebook is the Powerwash. If I
>mess things up too badly, I can restore to factory defaults really
>quickly. Chrome OS reinstalls my apps and some settings. The only
>manual steps are to enable Linux and a few other items. I then have
>scripts that configure my Linux environment and install my favorite
>packages.  For example, one script creates symlinks to my Google
>Drive. I can access all my remote files as if they were local at
>~/GoogleDrive. I’m not sure how Chrome OS mounts the remote file store
>but sshfs might be the way ( just a guess. ) All the normal tools work
>just fine on them: cat, grep, sed, awk, find, etc.
>
>Talking about things I have not figured out, Chrome OS seems to be
>running an X-server since GUI programs like xterm and ImageMagick run
>just fine.
>
>One item that is missing is a configuration manager for Chrome OS
>settings. For example, whenever I Powerwash, I have to configure
>manually my mouse and touchpad settings in addition to enabling Linux.
>It would be really nice to have Ansible, Salt, Puppet, Chef, or
>similar tool to configure those. Haven’t found anything, yet. Given
>that all that is done through the browser, I’m wondering if
>Tampermonkey might be the appropriate tool. Chrome reinstalls it after
>every Powerwash.
>
>Regards,
>- Robert
>
>On Thu, Jan 20, 2022 at 2:03 PM Timothy Scoppetta
> wrote:
>
>> My only experience with running linux (ubuntu) natively was with the
>> original CR-48 prototype devices. On those linux ran like a dream
>> but the touchpad was a nightmare with no drivers to tweak. The
>> installation process was a breeze even back when chromebooks were
>> new . My plan has always been to reinstall when I ran into something
>> that needed "real" linux. I'm still waiting for that something.
>>
>> Anything I can't do locally I do in GCP where I keep a few systems
>> powered down. I mount my homedir from a synology NAS in the closet.
>> It's a workflow that works very well for me. There's even a
>> globalprotect client for ChromeOS for those stodgy work connections.
>>
>> Tim
>>
>> On Thu, Jan 20, 2022 at 7:35 AM Robert Citek 
>> wrote:
>>  
>> > Thanks for your post, Timothy.
>> >
>> > I have almost the same environment here: read/listen to books on  
>> smartphone  
>> > or tablet or Kindle, otherwise use a Chromebook ( rarely, I'll use
>> > my MacBookPro).  In fact, if I need a "real" linux ( or Windows )  
>> environment,  
>> > I just create one in the cloud and then ssh/RDP into it from the
>> > Chromebook. I can even do that from my tablets which have Bluetooth
>> > keyboards.
>> >
>> > Have you ( or anyone else ) tried the dev_install script in the
>> > crosh  
>> shell  
>> > to then use emerge to install software in Chrome OS?  I've been
>> > able to install some things ( like tcpdump ), but not other things
>> > ( like gcc ).
>> >
>> > Have you ( or anyone else ) tried booting a different distro from
>> > either  
>> a  
>> > USB or SD card?  If so, how did that go? I have not, yet, but
>> > that's in  
>> my  
>> > queue.
>> >
>> > Regards,
>> > - Robert
>> >
>> > On Wed, Jan 19, 2022 at 12:40 PM Timothy Scoppetta
>> > > >
>> > wrote:
>> >  
>> > > I use a kindle for the original use case described but my
>> > > Pixelbook Go  
>> is  
>> > > the best linux netbook (does that term still exist?) I've ever
>> > > owned. I  
>> > run  
>> > > a full debian install in a container and have yet to find any
>> > > app or  
>> tool  

Re: [PLUG] Chromebooks and Linux

[John Jason Jordan]

On Thu, 20 Jan 2022 11:05:21 -0800
Johnathan Mantey  dijo:

>I was going to broach this subject as well Galen. I didn't because the
>original poster decided not to pursue a solution to the problem posed.
>I only recently discovered this "feature". I was already unimpressed
>with Chromebooks based on my personal experience with one. The end of
>support policy just makes them that much more unappealing.
>Johnathan
>
>
>On Thu, Jan 20, 2022 at 10:20 AM Galen Seitz 
>wrote:
>
>>
>> FYI to John and anyone else considering purchasing a Chromebook.  If
>> you are intending to use ChromeOS (i.e. as a normal Chromebook), you
>> should check the end of updates date before purchasing.  Sometimes
>> when you spot a really good deal on a Chromebook, it is because the
>> Auto Update Expiration date is approaching.  For the Lenovo IdeaPad
>> Duet 5, auto updates end in June 2029.

I had read about the end period for Chrome OS updates, although I was
unaware that 2029 was the end date for the Ideapad 5 13.3. I was
concentrating on getting Linux installed on it, or on whatever else I
ended up buying.

At this time I can't find anything that I want.

Re: [PLUG] Chromebooks and Linux

[John Jason Jordan]

On Wed, 19 Jan 2022 11:40:26 -0800
Timothy Scoppetta  dijo:

>I use a kindle for the original use case described but my Pixelbook Go
>is the best linux netbook (does that term still exist?) I've ever
>owned. I run a full debian install in a container and have yet to find
>any app or tool (GUI or command line) that I can't get working
>locally. I haven't tried heavyweights like gimp or bitwig and I
>definitely wouldn't suggest compiling on it.
>
>Happy to answer any questions about existing in the ecosystem, I use it
>daily and for me it works great.
>
>This is the one I've had for a little under a year:
>https://www.amazon.com/gp/product/B07YMGQYP6/

The Go is a good deal for someone who wants a regular type laptop. I
want something that I can use as a laptop occasionally, but 90%+ of its
use will be for reading while seated in a recliner, i.e., no table or
desk to set the screen on. I will have to hold it in the air in front
of me, as when reading a real book made of paper. For this, weight
becomes an issue, hence I want a '2-in-1' computer. Now, '2-in-1' is a
term that some vndors use to describe a laptop where the screen can be
folded completely around, but others use it to describe a computer
where the screen section and the keyboard section can be detached. It
is the latter that I want, and the Duet 5 13.3 is at the top of my list.

The great part about the Duet 5 13.3 is that it has an OLED screen,
which gets rave reviews. The bad part is that it has only 4GB RAM,
which is soldered.

Another bad point is its Snapdragon CPU, but last night I ran into a
post on Ubuntu forums where a person said that he had successfully
installed Ubuntu on a Snapdragon Chromebook and everything was working.
That may work for me, but I will likely have to run Linux in a
container instead, and then the shortage of RAM may become an issue.
Running Debian on top of Chrome OS in 4GB of RAM sounds iffy.

Another issue with the Duet 5 13.3 is that if you look for it on
lenovo.com the page says it is unavailable. Yet they are still selling
the previous version, the Duet 3 10.1, which does not offer the OLED
screen. However, the page for the Duet 5 13.3 says 'temporarily
unavailable,' not 'discontinued,' so maybe that means that they are
having supply chain problems like the rest of the world these days.

There aren't many detachable 2-in-1 computers to choose from, and OLED
displays are even scarcer.

Re: [PLUG] Chromebooks and Linux

[John Jason Jordan]

On Mon, 17 Jan 2022 22:54:15 -0800
Russell Senior  dijo:

>Not being able to print to a local USB connected printer in ChromeOS
>(you had to enroll your printer with google, and send your documents
>to google so they could send it back to your printer) used to be a big
>"NO THANK YOU" from me. I think someone told me that's no longer the
>case, but I'd already installed Gallium on my two Toshiba Chromebook
>2's. I kind of like Gallium, although new Chromebooks with higher res
>displays are hard to come by and it seems like the development pace
>has slowed recently, 3.1 is still based on Ubuntu 18.04 LTS, which is
>still supported upstream, but is coming up on 4 years old.
>
>On Mon, Jan 17, 2022 at 10:28 PM elcaseti  wrote:
>>
>> The Snapdragon 7 is an ARM CPU, so there might not be any distros
>> that would support that Duet 5 13.3.  If you do end up buying a
>> Chromebook, I agree that you really don't want to use ChromeOS.  I
>> used a Chromebook at my library recently, & it was so limiting, that
>> it was almost useless.  I was unable to print a PDF from a USB jump
>> drive, & it doesn't get more basic than that.  I asked a librarian
>> if he knew how to do this, & he said I don't think you can do that
>> on our chromebooks!

Thanks to all for the observations. I had already looked at Gallium, but
you have to download an ISO according to what CPU your device has, and
all the ISOs they offer are for Intel CPUs only, no ISOs for
Snapdragons.

My understanding is that you can run just about any distro if you
install it as a virtual OS, and current versions of Chrome OS have that
capability built in. The problem is finding a discussion somewhere
posted by someone who has actually installed a distro virtually on a
recent Chromebook, with information about what works and what doesn't.
The Lenovo that I have my eyes on has only 8GB RAM and 64GB storage,
and neither are upgradable. It's a $400 gamble, although I could
probably return it. I could live with one or two little things not
working, but what if the touchscreen or the wifi fail?

Maybe I need to put this project on the shelf and revisit it in a couple
years.

Re: [PLUG] Chromebooks and Linux

[John Jason Jordan]

On Tue, 18 Jan 2022 08:49:33 -0800
VY  dijo:

>If you have not purchased anything yet, this may be another option for
>you (though a bit experimental):
>
>https://www.indiegogo.com/projects/pocket-3-a-modular-and-full-featured-handheld-pc?fbclid=IwAR2xopyryIdeOTqD6CFLFNoTb-dH0x58-Oxx19BLdyV8YhXP56ZvJu8e5f0#/
>
>https://liliputing.com/2021/11/running-linux-on-the-gpd-pocket-3-mini-laptop.html?fbclid=IwAR1UgdQsl7Ze35pH6w5CqntOmAmCX1xHVakyXo3R8eCIPN9s9HHHvay6uXs

Too small. Also very expensive, although the RAM and storage are very
nice. I intend to use the device mostly to read books, and 11" is the
smallest screen I can consider, considering the condition of my ancient
eyes.

[PLUG] Chromebooks and Linux

[John Jason Jordan]

For some time I have been shopping for a computer to hold while
reading in a recliner, hence weight is an issue. I started looking at
tablets, thinking that I could get one with a separate keyboard for
times when the touchscreen was not adequate. But lately I've discovered
that nowadays there are lots of Chromebooks that are "2-in-1" capable,
that is, the keyboard is detachable. Of these the one currently at the
top of my list is the Lenovo IdeaPad Chromebook Duet 5 13.3, which is
available for ~$400.

The problem is that I'd much rather have a real distro instead of
Chrome OS. Yes, I know that Chrome OS is based on Linux, but it's kind
of limiting, plus I'd like to stay away from Google, if possible. The
problem is trying to find information about putting any other OS on the
Duet 5 13.3. For example, I stumbled on Gallium, an Xfce/Ubuntu 'for
Chromebooks' spin-off, but their hardware support page lists only
Chromebooks several years old. Plus, the Duet 5 13.3 has a Snapdragon 7
CPU, which they don't even mention, and I know next to nothing about
CPUs. I also checked Ubuntu forums, but didn't come up with much

Does anyone know about putting Linux on Chromebooks? Or websites
dedicated to the issue? Suggestions welcome!