Re: [PLUG] Radio silence since Apr 16

[Rich Shepard]

On Tue, 23 Apr 2024, Galen Seitz wrote:


That, or Rich is on vacation.  :-)


Ha-ha-ha.

Rich

Re: [PLUG] Want Low power long duration UPS

[Rich Shepard]

On Fri, 12 Apr 2024, Aaron Burt wrote:


We bought a shoebox-sized "portable power station" a couple years ago that
is stuffed with ~300Wh of Li-ion batteries and has a built-in inverter.

I very highly recommend having one on hand.


Thank you, Aaron. I just ordered the LiFePO4 version. It will be ideal for
business field work as well as recreational use.

Regards,

Rich

Re: [PLUG] Comcast to Ziply fiber migration, plus Ooma phone

[Rich Shepard]

On Fri, 12 Apr 2024, Russell Senior wrote:


Fwiw, every time I've called Zipy support (for my mom's residential
account), I've talked to someone with a southeastern USian accent.


Me, too. I've had better service from the Ziply business account customer
reps than I did with Frontier or Verizon. No complaints here. :-)

Rich

Re: [PLUG] Brave browser: asks for password when loading

[Rich Shepard]

On Tue, 9 Apr 2024, Ben Koenig wrote:


I think this might be the gnome keyring manager. It does this for a lot of
things, especially chrome based app. I can never remember how to make it
stop but it's one of those things that can get really annoying since it
will pop up even if you aren't in an actual gnome environment.


Ben,

How interesting.


If that's what the prompt is then there may be a way to make it stop. It's
not actually Brave doing it.


Thanks for the insight. I'll see if I can learn more about its behavior.

Much appreciated,

Rich

Re: [PLUG] Brave browser: asks for password when loading

[Rich Shepard]

On Mon, 8 Apr 2024, Michael Ewan wrote:


It is asking for the vault password, I think it uses GPG for that, it is
generally asked for when you install Brave and when you start it after a
reboot. There should be a setting for it somewhere, look for vault
settings.


Thanks, Michael. Looking at all Brave settings I found none for a vault.

Regards,

Rich

Re: [PLUG] Brave browser: asks for password when loading

[Rich Shepard]

On Mon, 8 Apr 2024, MC_Sequoia wrote:


A quick Google on "brave browser asks for password every time", returns a
few recent results regarding Brave & whatever the resident password
manager app is.


Mike,

The Brave version installed on the laptop has a setting for passwords, but
that's storage for user passwords on web sites, not a password to load
Brave. I don't mind Brave's quirks and prefer it to Firefox, Chrome, and
Chromium so I'll continue to triple click the 'close' box when invoking the
application.

Thanks for your suggestion,

Rich

Re: [PLUG] scp between two local (LAN) hosts

[Rich Shepard]

On Mon, 8 Apr 2024, wes wrote:


from the context, it seems to me that you're taking a different meaning of
the word "remote" than what is meant in the text. from the point of view
of scp, any host that is not the machine it's running on is "remote". it
seems to me that you're interpreting "remote" to mean not on the same
network as the machine on which you're running scp.


Wes,

I considered that the non-source host was remote even on the LAN, but didn't
get the destination address correct.


scp ./ salmo:
should get you started.



Thank you,

Rich

Re: [PLUG] Brave browser: asks for password when loading

[Rich Shepard]

On Mon, 8 Apr 2024, MC_Sequoia wrote:


A quick Google on "brave browser asks for password every time", returns a
few recent results regarding Brave & whatever the resident password
manager app is.

This might be the most relevant result from the Brave Community Forum,
https://community.brave.com/t/bug-never-fixed-save-password-default-toggle-unaccessible-cant-turn-off-keeps-asking-save-password/510180


Mike,

I apologize for not looking on the web before posting this thread.

Thank you,

Rich

[PLUG] scp between two local (LAN) hosts

[Rich Shepard]

It's been a very long time since I last used scp to copy files between
desktops and laptops on the local LAN. Searching the web for scp examples I
find only pages with commands for transferring a file from local to remote
host, from remote to local hosts. between two remote hosts, and transferring
multiple files to/from local/remote. Feh!

To transfer a file from the laptop to the desktop tried `scp ./
salmo'. That failed. So did adding the port option.

What am I missing?

TIA,

Rich

[PLUG] Brave browser: asks for password when loading

[Rich Shepard]

Haven't see this before. When I open the Brave browser on the laptop it
asks for a password. Clicking the "Cancel" button 3 times opens the browser.

Has anyone using Brave seen this before?

Rich

Re: [PLUG] Secure shell from/to laptop [RESOLVED--Partially]

[Rich Shepard]

On Wed, 3 Apr 2024, wes wrote:


the macs listed in ssh -Q mac are the macs the software is capable of using.
the macs listed in sshd_config are the macs _enabled_ for use.

macs are disabled by default when they are deemed to be below a
desirable security threshold. sometimes this happens because an algorithm
is actually broken, and sometimes it happens simply because of age.

algorithms are designed to maximize security with as little cpu overhead
as possible. as the average cpu power available to the general public
increases, algorithms designed for lower cpu power environments become
easier to crack. so after a certain age, an algo is simply assumed to be
broken without any actual evidence of that happening.

I don't agree with it, but that's the way it works.


wes,

Thank you.

Rich

Re: [PLUG] Secure shell from/to laptop [RESOLVED--Partially]

[Rich Shepard]

On Wed, 3 Apr 2024, Russell Senior wrote:


All Slackware why they aren't compatible and the best way to deal with it.


Russell,

It occurred to me that while the name might be the same the specifices might
change between releases. Although, why there are more macs reported with
`ssh -Q mac' on the laptop with the newer OS version than there are in that
host's /etc/ssh/ssh_config doesn't make sense to me. And the results of `ssh
-Q mac' are the same names on both hosts. Shrug.


Protocols get deprecated from time to time because of discovered or
perceived weaknesses. You can usually override the deprecations for a
while. If you don't like typing the override everytime, you can add a host
specific override to your .ssh/config file.


Yes. Until I'm able to complete setting up 15.0 on the NvME on the desktop
and boot into that version rather than the 14.2 on the SSD, I'll just add
the mac name on the command line when on the laptop and needing files from
the desktop.

Many thanks,

Rich

Re: [PLUG] Secure shell from/to laptop [RESOLVED--Partially]

[Rich Shepard]

On Wed, 3 Apr 2024, Russell Senior wrote:


Both sides.


The server's an earlier release and up-to-date for that one.

Regards,

Rich

Re: [PLUG] Secure shell from/to laptop [RESOLVED--Partially]

[Rich Shepard]

On Wed, 3 Apr 2024, Russell Senior wrote:


You should update openssh, I think. Ask someone who knows about slacker for
suitable advice.


Russell,

I updated Slackware64-15.0 last Friday. If openssh was patched it would have
been installed.

I can update again and see if that makes a difference.

Thanks,

Rich

Re: [PLUG] Secure shell from/to laptop [RESOLVED--Partially]

[Rich Shepard]

On Wed, 3 Apr 2024, Rich Shepard wrote:


The mac issue remains: I need to specify one (and I used hmac-sha2-512),
entered my password (not the passphrase) and was in. I can live with this
if I must, but it was not an issue before now.


Interesting. From the desktop (salmo) I can now ssh to the laptop (caddis)
without specifying a mac; the other way requires one.

The command `ssh -Q mac' returns more content on the server than on the
client.

In the client's /etc/ssh/ssh_config:
MACs hmac-md5,hmac-sha1,umac...@openssh.com

But,
$ ssh -Q mac
hmac-sha1
hmac-sha1-96
hmac-sha2-256
hmac-sha2-512
hmac-md5
hmac-md5-96
umac...@openssh.com
umac-...@openssh.com
hmac-sha1-...@openssh.com
hmac-sha1-96-...@openssh.com
hmac-sha2-256-...@openssh.com
hmac-sha2-512-...@openssh.com
hmac-md5-...@openssh.com
hmac-md5-96-...@openssh.com
umac-64-...@openssh.com
umac-128-...@openssh.com

Should I manually add those in the bottom list to ssh_config?

TIA,

Rich

Re: [PLUG] Secure shell from/to laptop [RESOLVED--Partially]

[Rich Shepard]

On Wed, 3 Apr 2024, Rich Shepard wrote:


and the same macs are on the client. But, if I don't specify, e.g.,
hmac-sha2-512 I get the missing mac error. The perms on both the server and
client .ssh/* match and those in
<https://linuxhandbook.com/fix-permission-denied-publickey/>.

Had not before had the mis-mac perms issues so I'm open to learning what
else I missed.


Fixed the public key denied issue: on both hosts in /etc/ssh/sshd_config
passwordauthentification were set to 'no'; changed that to 'yes' and
restarted sshd.

The mac issue remains: I need to specify one (and I used hmac-sha2-512),
entered my password (not the passphrase) and was in. I can live with this if
I must, but it was not an issue before now.

Regards,

Rich

Re: [PLUG] Secure shell from/to laptop

[Rich Shepard]

On Wed, 3 Apr 2024, Michael Ewan wrote:


It looks like the MAC (message authentication code) algorithm list does
not match between the two hosts. The destination host is offering one that
does not exist on the client.


Michael,

That's what I found when I searched for the error message. However, for
example, on the client I try `ssh hmac-sha2-512' I'm now told there's a
public key issue.

This is on the server:
$ ssh -Q mac
hmac-sha1
hmac-sha1-96
hmac-sha2-256
hmac-sha2-512
hmac-md5
hmac-md5-96
hmac-ripemd160
hmac-ripemd...@openssh.com
umac...@openssh.com
umac-...@openssh.com
hmac-sha1-...@openssh.com
hmac-sha1-96-...@openssh.com
hmac-sha2-256-...@openssh.com
hmac-sha2-512-...@openssh.com
hmac-md5-...@openssh.com
hmac-md5-96-...@openssh.com
hmac-ripemd160-...@openssh.com
umac-64-...@openssh.com
umac-128-...@openssh.com

and the same macs are on the client. But, if I don't specify, e.g.,
hmac-sha2-512 I get the missing mac error. The perms on both the server and
client .ssh/* match and those in
.

Had not before had the mis-mac perms issues so I'm open to learning what
else I missed.

Thanks,

Rich

[PLUG] Secure shell from/to laptop

[Rich Shepard]

I thought I had /etc/ssh_config and /etc/sshd_config properly set up on the
T430 laptop, as well as ~/.ssh/*.

But, when I tried to ssh from the laptop (caddis) to the server (salmo) it
failed because ssh could not find the proper MAC. As attachments are
stripped from this mail list the screen shot can be downloaded from
.

Should I add all the MACs in that message to salmo's /etc/ssh_config?

Advice needed.

TIA,

Rich

Re: [PLUG] Death Notice / Unsubscribe

[Rich Shepard]

On Tue, 2 Apr 2024, Russell Johnson wrote:


"If you can get physical access to a computer, it's not secure. It's only
a matter of time." That goes with the joke, "The only secure computer is
one that is in a bunker with no windows or door, no power, and no
connection to the outside world. And we're not sure about that one."


I was in the Army (counterintelligence and security) in (West) Germany in
the mid- late-1960s. We learned that no lock is secure given sufficient time
to pick/manipulate it. Even the Sargent-Greenlee padlocks could be opened
with enough time. Lock security was based on how much time it was not
observed. Computers fall into the same category.

Rich

Re: [PLUG] Why multiple ssh ports?

[Rich Shepard]

On Sat, 30 Mar 2024, wes wrote:


they can, but all of those hosts share the same IP address to the outside
world. so if you're connecting from outside to inside, you have to use a
different port for each host you want to connect in to.


Wes,

Okay, now I understand and that makes complete sense. When I'm away from the
office I'll use ssh to log into my server to check emails or, on occasion to
download a file.

Another good lesson learned. Thanks again.

Best regards,

Rich

Re: [PLUG] Why multiple ssh ports?

[Rich Shepard]

On Sat, 30 Mar 2024, wes wrote:


it's to enable you to have 3 machines inside your network you can reach via
ssh.


wes,

Interesting. I thought all hosts on the LAN can use the same ssh port.

Thank you,

Rich

Re: [PLUG] Why multiple ssh ports?

[Rich Shepard]

On Sat, 30 Mar 2024, Robert Citek wrote:


Have you tried running the script to verify that the comments align with
what the script does?


Robert,

No. I want to learn why I should have three port numbers rather than only
the one I use.

Regards,

Rich

Re: [PLUG] Why multiple ssh ports?

[Rich Shepard]

On Sat, 30 Mar 2024, MC_Sequoia wrote:


I'll attempt to demystify and sum up the techno-jargon.

This is common firewall/network security practice. This is an iptables, Linux 
Firewall script. The idea is not only accept packets inbound to your network on 
a non-default port such as the 3 destination ports in the the example, 
99,88,8889. Only tcp/ip packets with one of those dest. ports will be accepted 
and then re-routed out of the firewall to a machine on the network that is 
setup for SSH access on the default port of 22.

The idea is that only you, the sys-admin, developers, etc will know to setup 
ssh access to your internal pc's/servers, etc on those non-default ports that 
you've chosen. Which hopefully aren't the same port numbers in the example 
script!!! =)

Slackware Docs explains this. Reference - 
https://docs.slackware.com/howtos:security:ssh

/usr/sbin/iptables -t filter -A INPUT -p tcp -m multiport --dports 99,88,8889 
-j REJECT --reject-with tcp-reset

##Then, pick a number between 1 and 4294967295 Ill use 0x13F ()
##Were going to tell iptables to reject anything without this mark coming 
into port 22.

/usr/sbin/iptables -t filter -A INPUT -p tcp -m tcp --dport 22 -m connmark ! 
--mark 0x13F -j REJECT --reject-with tcp-reset

##Now well tell iptables what ports we will accept for ssh.

/usr/sbin/iptables -t filter -A FORWARD -p tcp -m multiport --dports 99,88,8889 
-j ACCEPT

##In the mangle? table we slap our mark on these packets.

/usr/sbin/iptables -t mangle -A PREROUTING -p tcp -m multiport --dports 
99,88,8889 -j CONNMARK --set-mark 0x13F

##Finally in the nat table we tell iptables to send the marked packets back 
to port 22

/usr/sbin/iptables -t nat -A PREROUTING -p tcp -m multiport --dport 99,88,8889 
-j REDIRECT --to-ports 22


Mike,

I understand the purpose and I don't know the value of repeating the script
I presented. Do six hashmarks have benefits over a single one as marking a
comment.

My question remains.

Regards,

Rich

[PLUG] Why multiple ssh ports?

[Rich Shepard]

Setting up the ThinkPad T430 brings me to reviewing my network's ssh
configuration. I found a Slackware script that hides the accepted ssh port
number while rejecting all others, including 22.

The example script includes:
# First, set SSHD back to the default port 22.
# Next, figure out what port or ports you want to do SSH over.
# Were going to use 99, 88, and 8889 here.
# Now we take care of the Hypothetical Evil Unprivileged User
# by not accepting anything over those ports in the first place.
# This is only effective for port 8889 but well do all three ports for the
# sake of completeness.  
/usr/sbin/iptables -t filter -A INPUT -p tcp -m multiport --dports 99,88,8889 
-j REJECT --reject-with tcp-reset

Why use three port numbers if only the last one will be valid?

TIA,

Rich

Re: [PLUG] Deleting directory in $HOME

[Rich Shepard]

On Fri, 29 Mar 2024, David Bridges wrote:


I do not use xfce but more than likely it is using xdg-user* stuff

Take a look in ~/.config/user-dirs.dirs

I modified the values to be what I wanted a long time ago on my system
because I the directories listed in that file were recreated after I
deleted them.


David,

Thank you. I had no idea that user-dir.dirs existed. Sure enough, all the
default Slackware installation directories in $HOME are there. I commented
out all of them.

Much appreciated lesson,

Rich

Re: [PLUG] Deleting directory in $HOME

[Rich Shepard]

On Fri, 29 Mar 2024, King Beowulf wrote:


Anything Xfce and X.org creates in $HOME as a default will just get
recreated if you remove/delete.  Just because YOU don't sue a config
does not mean the DE (Xfce or X) does not use it.


Ed,

The Desktop directory is not a configuration file; it's a user directory.


If you insist:
https://unix.stackexchange.com/questions/206594/why-cant-i-delete-this-file-as-root


Interesting. On the laptop ~/Desktop has only one attribute: e. That's not
shown in the man page and it's neither i or a. Perhaps removing the e
(whatever that is -- I'll look for an explanation) will help.

Thanks,

Rich

Re: [PLUG] Loading ~/.Xmodmap on laptop [FIXED]

[Rich Shepard]

On Fri, 29 Mar 2024, King Beowulf wrote:


That doesn't look right for Slackware.  Use the default one for 15.0.

1. exit Xfce
2. make sure logged as your normal user, no GUI or X running, and not as
root
3. run "xwmconfig"
4. choose any OTHER desktop, hit OK
5. run "xwmconfig" again, choose Xfce, Ok


Ed,

Ah. my bad. I hadn't considered that xwmconfig changes with the versions. Of
course, it's been a long time since I upgraded to 14.2 and I've forgotten a
lot of details.


to start X and Xfce you simply
$ startx
 DO NOT RUN startxfce4 DIRECTLY!!! 


I don't. startx is an alias in ~/.bash_profile.


Make sure that the new X and Xfce stuff from 15.0 is installed in /etc
and you did NOT copy ANY old settings from 14.2.


Nope. All 15.0.

Much thanks,

Rich

[PLUG] Deleting directory in $HOME

[Rich Shepard]

Xfce4 adds a few Capitalized directories in a user's ~/ when it's installed.
I've never used them and had no problems deleting them until now: the
Desktop directory (755 perms) will not be removed either by me or by root.
It has 2 hardlinks: the current and parent directories.

The xfce mail list is very quiet; no posts since I rejoined yesterday. My
web searches haven't found a reason why a directory owned by me and in the
users group cannot be removed.

A clue stick will be appreciated.

TIA,

Rich

[PLUG] Loading ~/.Xmodmap on laptop

[Rich Shepard]

This has been an issue with me on all laptops I've owned: despite ~/.xinitrc
containing:

#!/bin/sh
# $XConsortium: xinitrc.cpp,v 1.4 91/08/22 11:41:34 rws Exp $

usermodmap=$HOME/.Xmodmap
/usr/bin/xmodmap $HOME/.Xmodmap
userresources=$HOME/.Xresources
sysmodmap=$HOME/.Xmodmap

#if [ -f $userresources ]; then
#xrdb -merge $userresources
#fi

virtual consoles do not load it when I startxfce4 and open v.t.s.

What am I missing?

TIA,

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Tue, 26 Mar 2024, Ben Koenig wrote:


We also have a standard udev.d setup in lib/udev/rules.d/ and
/etc/udev/rules.d/. The only difference is that we
use eudev rather than the systemd nonsense.


Ben,

Thank you. I thought I had looked in /lib, but obviously I didn't. Mea
culpa!


Rich, the files in /dev/disk/by-path are all symlinks, use ls -l to view
them. They should link to the normal drive names, however by-path names
don't change per device, they change per port. So all USB drives plugged
into a given port will use the same name in by-path, even if the drive
letter changes.


Ah, so. I saw they were symlinks but didn't follow them to their hard links.

This topic is not an issue on a desktop that's infrequently rebooted or laptop
that has only one drive. I'm developing this desktop to boot into either
Slackware64-14.2 or -15.0, I think that whether the USB port is /dev/sdg or
/dev/sdh upon rebooting depends on whether it's raining.

Your Slackware lessons are always valuable and appreciated.

Best regards,

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Tue, 26 Mar 2024, Russell Senior wrote:


I don't know anything about Slackware, but ...
 https://refspecs.linuxfoundation.org/FHS_3.0/fhs/ch03s15.html
I'm going to take a wild guess and suggest if you run:
 $ mount | grep '/run '
you are going to see /run is a tmpfs.


Then I'll find where to put udev rules so they remain.

Thanks, Russell,

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Tue, 26 Mar 2024, Johnathan Mantey wrote:


Do you have a /lib/udev/rules.d subdir?


Nope. On Slackware rules are in /run/udev/rules.d.

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Johnathan Mantey wrote:


People who are very skilled at making udev rules have done so in my
Fedora, and I assume *other* Linux distros, in the */lib/udev/rules.d*
subdirectory. I expect the */run/udev* subdirectory to be empty because it
is not a persistent file-system.


I have /etc/udev/rules.d/ with one entry from 2018. /run/udev/ is more
current.

When the kernel keeps the same device name for portable USB drives they're
loaded because that mount point is in fstab.


I'm not certain what you mean by "*Generic for any portable USB drive*".
You *want* each drive to have some unique label. If you don't then you
will most likely get some kind of access collision that isn't going to end
well.


What I mean is that my thumb drives come from different vendors.

Thanks,

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Johnathan Mantey wrote:


Notice that it makes a 'by-id' entry. It also makes a 'by-uuid', which may
be more interesting for this use case.


Johnathan,

Yes, using the UUID would be better, but my understanding is that is unique
to each device so it's not generic for any portalble USB drive.

Thanks,

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Johnathan Mantey wrote:


No you don't.
The /lib/udev/rules.d has a rule that fully populates the /dev/disk
subdirectory with the same name and same information on each boot.
This should be everything you need to use fstab directly.


Johnathan,

Then I'm not understanding. Here, /run/udev/rules.d/ was empty until I added
some. When I reboot this desktop and insert a thumb drive the kernel might
see it as /dev/sdg (which is what's in /etc/fstab) or /dev/sdh (which isn't
in /etc/fstab.)

Thanks,

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Johnathan Mantey wrote:


I found the rules I was after
/lib/udev/rules.d
These rules create a device for my USB drive here:
/dev/disk/by-id/usb-JetFlash_Transcend_8GB_B101484C-0:0


Johnathan,

Yes, I can create a udev rule file that has the vendor and ID for each USB
pen drive I have, Currently about a dozen. Which means that each time I
acquire a new one, or use one from someone else, I need to enter a new line
in that .rules file. That's no different from tailing /var/log/messages and
having root mount the drive.

Regards,

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Rich Shepard wrote:


Nothing useful there. I can plug in a drive and have fdisk -l identify the


Actually, that's `tail -f /var/log/messages.'

And rebooting the machine returned the USB port to /dev/sdg.

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, MC_Sequoia wrote:


Sounds like you want to create a custom mount point for your USB thumb
drives?

If so, This should walk you through the steps.
https://linuxconfig.org/howto-mount-usb-drive-in-linux


Mike,

Nothing useful there. I can plug in a drive and have fdisk -l identify the
drive numer then have root mount it. No different from what I've done for
years. The point of creating a udev rule is to have the kernel recognize the
drive regardless of its loading order and using that information to mount it
on /media/thumb/.

Regards,

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Rich Shepard wrote:


Now the last step: How do I modify /etc/fstab so it lets me read and write
to any pen drive inserted in a USB port?


Clarification: how to mount any inserted USB pen drive on /media/thumb.

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Rich Shepard wrote:


But, inserting and removing a USB pen drive does not add new lines to
/tmp/udev.log.


Well! The web pages on udev rules were incomplete. The correct command is
udevadm control --reload-rules
This works.

I wrote 41-usb-permissions.rules: SUBSYSTEM=="usb", MODE:="0666"

When I insert and remove different pen drives the action is recorded in
/tmp/udev.log.

Now the last step: How do I modify /etc/fstab so it lets me read and write
to any pen drive inserted in a USB port?

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Rich Shepard wrote:


But, inserting and removing a USB pen drive does not add new lines to
/tmp/udev.log.


Very interesting update. Advice on udev web sites I read say to reboot the
host rather than running `udevadm control --reload'. So I rebooted. The two
rules in /run/udev/rules.d/ are gone!

WTF?

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Johnathan Mantey wrote:


My initial reaction would be to look at creating UDEV rules for those
devices.


Reading several udev-related web pages I created two shell scripts installed
in /usr/local/bin and rule 80-test.rules in /run/udev/rules.d

- device_added.sh -
#!/usr/bin/bash

echo "USB device added at $(date)" >>/tmp/udev.log
--

- device_removed.sh -
#!/usr/bin/bash

echo "USB device removed  at $(date)" >>/tmp/udev.log
--

- 80-test.rules -
SUBSYSTEM=="usb", ACTION=="add", ENV{DEVTYPE}=="usb_device",  
RUN+="/bin/device_added.sh"
SUBSYSTEM=="usb", ACTION=="remove", ENV{DEVTYPE}=="usb_device", 
RUN+="/bin/device_removed.sh"
--

Inserting a USB pen drive no /tmp/scripts.log was created. I touched that
filename and reran `udevadm control --reload' but /tmp/scripts.log was
empty. As root, directly running the two device files in /usr/local/bin
added the proper content.

But, inserting and removing a USB pen drive does not add new lines to
/tmp/udev.log.

Advice needed.

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Johnathan Mantey wrote:


I can't remember if the results are enumerated in the /lib directory. You
may want to spend a few web searches seeing if you can find where they
live.


Johnathan,

On my Slackware hosts they're in /run/udev, but the rules.d/ subdirectory is
empty. Huh! Earlier releases had content there.

Regards,

Rich

Re: [PLUG] Assigning labels to USB pen drives

[Rich Shepard]

On Mon, 25 Mar 2024, Johnathan Mantey wrote:


My initial reaction would be to look at creating UDEV rules for those
devices. Assuming they are well behaved and have unique USB serial numbers
you should be able to craft rules that will create consistent /dev
symbolic links to wherever the kernel places them. There's also a very
extensive set of existing UDEV rules that enumerate USB devices as fully
as they can. I can't remember if the results are enumerated in the /lib
directory. You may want to spend a few web searches seeing if you can find
where they live.


Thanks, Johnathan.

Rich

[PLUG] Assigning labels to USB pen drives

[Rich Shepard]

For permanently attached drives I've assigned their PARTUUID to mount them
in /etc/fstab. With 6 continuosly attached drives (/dev/sda-/dev/sdf)
portable USB drives were seen as either /dev/sdg or /dev/sdg1 and that's how
they're mounted in /etc/fstab.

Yesterday afternoon the kernel decided to see these pen drives as /dev/sdh
and /dev/sdh1.

How can I assign a generic pen drive label in /etc/fstab so any pen drive
inserted in a USB port on a host so it can be mounted on /media/thumb or
/media/stick depending on how that brand of pen drive is seen?

TIA,

Rich

Re: [PLUG] SWAP size based on host's installed memory

[Rich Shepard]

On Sat, 23 Mar 2024, Johnathan Mantey wrote:


Did I miss the definition of SAR?


System Activity Report. On my Slackware desktop it's in /user/bin/sar.

man page available, too.

HTH,

Rich

Re: [PLUG] Making a bootable thumb drive [FIXED]

[Rich Shepard]

On Sat, 23 Mar 2024, Rich Shepard wrote:


Why sdg has two partitions I don't know. My question is to which device
name I point dd's of=? /dev/sdg or /dev/sdg1. Or should I use cfdisk to
re-partition the drive?


Never mind. I ran `dd if=usbboot.img of=/dev/sdg and I now again have a
bootable drive.

Rich

[PLUG] Making a bootable thumb drive

[Rich Shepard]

For some reason the bootable thumb drive I made (for Slackware64-15.0) now
asks to load PXE to install over the network rather than from the flash
drive itself.

I have the usbboot.img file and the flash drive is seen by the kernel:
Mar 23 11:04:35 salmo kernel: [246888.820945]  sdg: sdg1 sdg2
Mar 23 11:04:35 salmo kernel: [246888.827498] sd 10:0:0:0: [sdg] Attached SCSI 
removable disk
Mar 23 11:04:55 salmo kernel: [246909.135961] FAT-fs (sdg1): Can't find a valid 
FAT filesystem
Mar 23 11:04:59 salmo kernel: [246913.165744] FAT-fs (sdg): Can't find a valid 
FAT filesystem

Why sdg has two partitions I don't know. My question is to which device name
I point dd's of=? /dev/sdg or /dev/sdg1. Or should I use cfdisk to
re-partition the drive?

TIA,

Rich

Re: [PLUG] SWAP size based on host's installed memory

[Rich Shepard]

On Sat, 23 Mar 2024, Michael Ewan wrote:


Some swap is necessary but not the RAM+2GB or other calculations, those
kinds of recommendations are based on very old memory management routines.

You can start with a swap file for safety and then measure actual memory
use under actual work loads over a period of time, you can add/remove swap
files as necessary rather than creating a swap partition (look at the
swapon man page).

Using SAR you can get all the information you need.  Ages ago when we were
sizing some big iron we ran SAR over a period of time and sized the RAM
(very expensive at the time) so the machine NEVER paged out.  Paging stats
are far more important than swap stats, a machine pages all the time but
should never swap, swapping is very expensive and indicates a machine that
has pathetically small RAM.  You will probably see some page outs (page in
is fine since starting a process is a page in), so you can size your swap
space based on the paging needs of the current configuration.


Michael,

Thanks for makeing me aware of SAR.

Germene to all swap partition size justifications might the solution be
relative to use and memory size alone? If a host can support 16GiB (e.g., on
a laptop) or 64GiB (on a desktop) but there's only one or a few concurrent
users I wonder whether any swap partition would be necessary. Perhaps
running a spatio-temporal statistical or multi-dimentional hydraulic model
might require a lot of memory (unless it supports threads, which most do
nowadays) might benefit from a swap partition. And then, most storage
devices (hdds and ssds) are large enough that a swap partition of installed
RAM plus 2 GiB would not hinder application and data storage space.

Regards,

Rich

Re: [PLUG] something I am considering doing...

[Rich Shepard]

On Fri, 22 Mar 2024, Russell Senior wrote:


To me, the only open question is whether humans get stupider faster than
the machines.


Isn't that why the search for extraterrestrial intelligence has been going
on for so many decades?

Rich

Re: [PLUG] rsync: not all files copied

[Rich Shepard]

On Fri, 22 Mar 2024, Tomas Kuchta wrote:


Errors with rsync usually mean that the file system changed - such as logs,
proc, sys, var, etc.


Tomas,

I thought this might be the cause. While I did nothing in the partitions
being sync'd the kernel or some other application might have made a change.

Thanks,

Rich

Re: [PLUG] comment on rsync

[Rich Shepard]

On Thu, 21 Mar 2024, American Citizen wrote:


I run rsync and occasionally get files not copied error, but the output
stream usually lets me know what did NOT get copied.

I think if you look back on your log files, you might find them.


Randall,

I didn't see an rsync logfile. Next run I'll tee the output or log only
errors.

Thanks,

Rich

Re: [PLUG] SWAP size based on host's installed memory

[Rich Shepard]

On Thu, 21 Mar 2024, Jeffrey Borcean wrote:


Are you wishing to use suspend to disk hibernation? If so, the RAM + 2 GB
as Mike suggested is great.


Jeffrey,

I only rarely hibernate the laptop and never the desktops.


Otherwise consider no disk based swap, and instead use zram as swap. This
greatly increases responsiveness compared to any disk based swap
configuration. There is a zram-generator [1] that makes it easy to
configure.

[1] https://github.com/systemd/zram-generator


Interesting. Thank you.

Regards,

Rich

Re: [PLUG] SWAP size based on host's installed memory

[Rich Shepard]

On Thu, 21 Mar 2024, MC_Sequoia wrote:


That recommendation is for RAM < 2 GB for ram > 2 GB, the recommendation
is RAM + 2 GB.


Thanks, MC. I could not find this information in the SA book or online.

Regards,

Rich

[PLUG] SWAP size based on host's installed memory

[Rich Shepard]

I'm waiting for delivery of a 2T SSD to replace the 500G HDD on the ThinkPad
T430. The advice to create a SWAP partition sized at RAM x 2 comes from the
past when computers had very little memory.

When a host has 16G, 32G, or 64G RAM what size SWAP partition would be
appropriate?

TIA,

Rich

Re: [PLUG] rsync: not all files copied

[Rich Shepard]

On Thu, 21 Mar 2024, Michael Ewan wrote:


As others have said, capture the output, but instead of a pipe use
'--log-file=rsync.log' option.  Also, when you run rsync the second time it
should only log the errors since all the files have been copied already.


Michael,

Good point. Will do so.

Thanks,

Rich

Re: [PLUG] Partitioning new HDD; installing file system

[Rich Shepard]

On Thu, 21 Mar 2024, Michael Ewan wrote:


I can highly recommend that book. I used it as a textbook when teaching
college courses in Linux Administration. I also knew Evi personally, a
truly wonderful person and a great loss to our corner of the universe.


Michael,

I have a copy and read/referred to it in the past. Hadn't occurred to me to
look at it again this week. Thanks for the reminder.

Regards,

Rich

Re: [PLUG] rsync: not all files copied

[Rich Shepard]

On Thu, 21 Mar 2024, King Beowulf wrote:


you can tee stdout to a file


Ed,

True that. Didn't occur to me as I was focused on getting that replacement
drive set up.


$ rsync blah-blah 2>&1| | tee rsync.log


I'll do this the next new disk update.

Thanks,

Rich

Re: [PLUG] rsync: not all files copied

[Rich Shepard]

On Wed, 20 Mar 2024, wes wrote:


why wait a few days? you could run rsync again immediately and you can
look for errors.


wes,

Two reasons:

1) I'm waiting for the weekend to replace the desktop's internal hard drive.
I'll run rsync then to update what changed from today.

2) The list of copied files scrolls by very quickly. Even if I sat her
watching I'd probably not see an error message and wouldn't know how
to stop the process to read it.

3) The three partitions duplicated contain 39G, 145G, and 171G. The first
one took 21 minutes; I didn't time the others but they were much longer.

Best regards,

Rich

Re: [PLUG] Partitioning new HDD; installing file system

[Rich Shepard]

On Wed, 20 Mar 2024, MC_Sequoia wrote:


At this point, you might consider compiling some good crib notes or what
we used to call "Method of Procedure" or MOP in military parlance for all
these sys-admin type tasks that you inquire to the list about.


I save the threads to a linux file. Most of the time.

Rich

[PLUG] rsync: not all files copied

[Rich Shepard]

After running for about an hour rsync ended this way:
-
sent 181,880,754,804 bytes  received 1,336,778 bytes  43,642,973.39 bytes/sec
total size is 181,831,399,934  speedup is 1.00
rsync error: some files/attrs were not transferred (see previous errors) (code 
23) at main.c(1327) [sender=3.2.5]
-

Is there a way to find which files were not transferred?

Or, if I run rsync again in a few days might those files be transferred?

TIA,

Rich

Re: [PLUG] rsync between 2 hdds on the same host

[Rich Shepard]

On Wed, 20 Mar 2024, Michael Ewan wrote:


That should work.


Thanks, Michael.

Another linux lesson learned.

Regards,

Rich

[PLUG] rsync between 2 hdds on the same host

[Rich Shepard]

My experience with rsync has been between hosts on the LAN and backups from
a desktop to the MediaSonic Probox external drive. My web searches for the
appropriate syntax between two drives on the same host finds nothing
suitable.

If I mount the new drive on /media/hd0 I wonder if this syntax will work:
rsync -av /home /media/hd0/1
rsync -av /opt /media/hd0/2
rsync -av /data1 /media/hd0/3
Leaving off the source final slash will create the directory name and
contents on the target's partion.

Comments welcome,

Rich

Re: [PLUG] Partitioning new HDD; installing file system

[Rich Shepard]

On Wed, 20 Mar 2024, Russell Senior wrote:


You missed that you didn't format a partition, but instead formated the
whole drive. Create your partitions again, and format just the partition.


Russell,

Thanks. I do this so infrequently I need to remember, or relearn, the
sequence. And, a bit more web research reminded me to use cfdisk on the new
disk because it has an DOS disklable type by default.

Regards,

Rich

[PLUG] Partitioning new HDD; installing file system

[Rich Shepard]

I've a new Seagate FireCuda 4T HDD connected to the desktop via a SATA/USB
adapter.

Tailing /var/log/messages the kernel sees this disk as /dev/sdg.

I ran `cgdisk /dev/sdg' and it told me it did not have GPT as the
partition type, but would try to fix that. Assuming that to be the case, I
created 3 partitions, wrote them to disk, quit cgdisk, then ran `mke2fs -t ext4
/dev/sdg' to completion.

Now, `fdisk -l' shows:
Disk /dev/sdg: 3.7 TiB, 4000787030016 bytes, 7814037168 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes

with no disklabel type. And, `cgdisk /dev/sdg' shows it to be unparitioned.

What have I missed here?

TIA,

Rich

Re: [PLUG] UEFI and hardware replacement

[Rich Shepard]

On Wed, 20 Mar 2024, Vince Winter wrote:


UEFI is different firmware system on the motherboard. Some of the firmware
is backwards compatible with legacy boot loaders.


Thanks for confirming, Vince.

All my hosts use UEFI.

Regards,

Rich

[PLUG] UEFI and hardware replacement

[Rich Shepard]

Just to confirm: UEFI, like the old BIOS, is in a chip on the motherboard
and not affected by hard drive replacement. Yes?

TIA,

Rich

Re: [PLUG] ThinkPad T430 RAM replacement [RESOLVED]

[Rich Shepard]

On Tue, 19 Mar 2024, Michael Ewan wrote:


This video shows the clips being pressed outward.
https://support.lenovo.com/us/en/videos/pd027009-memory-modules-removal-movie-thinkpad-t430


Michael,

Didn't watch the video. Instead, I used two fingers on one hand to hold both
clips out from the board sides and used a very small screwdriver to gently
lift the long edge. It did pop up and allow me to wiggle the module out of
the socket.

This was the under-keyboard module. After replacing it and re-installing the
keyboard I'll do the in-base module.

Thanks,

Rich

Re: [PLUG] ThinkPad T430 RAM replacement

[Rich Shepard]

On Tue, 19 Mar 2024, Michael Ewan wrote:


I do not have direct experience with the T430, but the clips should move
out away from the DIMM (not up or down), and then the DIMM should pop up a
bit.


Michael,

On other laptops I've owned I've been able to rotate the clips away from the
module (same level), but that's not working here.

Thanks,

Rich

[PLUG] ThinkPad T430 RAM replacement

[Rich Shepard]

I'm upgrading the memory on my ThinkPad T430; the new modules should be here
tomorrow morning. Each module is held in place by a small clip on each end.
I've tried gently moving them out, but that doesn't work.

There are excellent instructions on this page

but he doesn't explain how to remove the modules once they're exposed.

How do I release the modules so I can remove them and insert new modules?

TIA,

Rich

Re: [PLUG] Celebration time, COME ON Apple SHUT THE F UP!!!

[Rich Shepard]

On Tue, 5 Mar 2024, Ted Mittelstaedt wrote:


I think this needs to be on the docket as a speakers topic once the bill
is signed. Do we know anyone in the Right to Repair movement?


Ted,

There's a company called ubreakifix
. I used the one on Jansen
Beach and they seem to know what they're doing.

HTH,

Rich

Re: [PLUG] Linux Software for Data Center Monitoring

[Rich Shepard]

On Sat, 2 Mar 2024, Ben Koenig wrote:


Does anyone here who works with SMB scale datacenter environments have any
tips or industry standard strategies for wrangling this type of setup? Are
there any good FOSS software tools to help organize and monitor a mess
like this? We have a software team that keeps and eye on the applications,
but they do not appear to be monitoring things like power consumption,
temperature, or even tracking parts as they get re-used. Our server "map"
is literally just a Google Sheets document that was formatted to look like
server rows with IP addresses listed by physical location. And I'm pretty
sure everyone hates it. So I'm basically looking for tools to help me set
up the following infrastructure:

- server documentation. Type, hardware configuration, and parts compatibility
- temp monitoring. Many of the servers are running CUDA applications on
  Dual/Quad GPU systems. They get toasty.
- power consumption monitoring. Our PDUs are able to report usage via a
  network interface, but nobody ever bothered to set it up. Would be nice
  to have a dashboard that shows when one of the servers freaks out and
  trips the breaker.
Thoughts? Solutions? Apps? I'm just looking for ideas at the moment.
Everything is running (or so I'm told) but we currently have a bus number
of 1 which is obviously a recipe for disaster. I don't mind piecing
together my own set of scripts and utilities but if something already
exists that does the work for me, even better :)


Ben,

As you well know, this is not in my wheelhouse, but it's very similar to
environmental permit management for large corporations, so I offer a couple
of suggestions for you to ponder.

1. First thing, is write a SOP (Standard Operating Procedures) that
encapsulates all needs of the data center and describes all hardware,
software, and personnel involved. This puts in place a standard that can be
used for new staff, hardware, and software.

2. Develop a database application with a web interface; Postgresql and Flask
or Django (the latter initially developed for on on-line newspaper that's
constantly being updated.) Since temperature and power monitoring are
continuous/continual the database will be updated in real time. There almost
certainly are tools that will automate measurements and insertion in the
appropriate database table.

The initial effort will pay for itself in a short time and make your life
(and that of others in the datacenter and company) much easier.

HTH,

Rich

Re: [PLUG] Replace ZiplyFiber VoIP router

[Rich Shepard]

On Thu, 29 Feb 2024, MC_Sequoia wrote:


Color me confused. Firstly, you said you didn't need WiFi support and
secondly the TP-Link isn't a "fiber optic gateway."


That's what I learned after more research. The terminology is brand new to
me. I found a completely different set of less expensive and more
appropriate choices when I searched the web for 'fiber optic routers.'


Why not just see if you can buy the Commscope Homevantage fiber gateway?
However, it seems to be marketed to Service Providers.
Maybe Zipley will sell it to you outright.


I'll ask them about selling it.

Thanks, again,

Rich

Re: [PLUG] Replace ZiplyFiber VoIP router

[Rich Shepard]

On Thu, 29 Feb 2024, Russell Senior wrote:


I went through this with my mother over the last year. At the beginning of
2024, we moved her voice service to Ooma. You need to buy an Ooma Telo.
You can find Telo's on ebay for about $50. Ooma also charges just south of
$50 to port your number to their service, but thereafter it's a little
under $20/month. Another gotcha was when we ported the number, Ziply had
to assign a new account number (because it is tied to the phone number).
When they converted my mom's account, they just deleted (or made
unavailable to us) her old account information, including billing history.
We shipped the router back to them with their shipping label, but they
continued to bill for it another month after they'd received it. Also the
paperless billing and autopay we had set up with the old account went
poof, so they happily billed for that as well for a month until we figured
out what was happening and asked them to stop. I *think* that has all
settled out now, and her bill will be $45/month for just 100/100 internet
service. Because there is no regulatory body that cares, they (and their
"competitors") are free to gouge, and we mere customers are without
practical recourse.


Russell,

Thanks for the detailed history. It might well be that having a business
account with them, rather than a personal account, making a change could be
more expensive with more of a hassle. I'm far less experienced than you in
telco matters as the phone is a tool for my business so I know much less
about the subject.


One regression associated with VOIP is you no longer have central office
battery to keep your service working in a power outage. When your internet
goes out, your voice service stops as well, unless you take your own steps
to power your own equipment (and even then, their intermediate equipment
might die for lack of power). You might be lucky and your cell phone still
works, but ... you can't count on it. Be prepared to communicate with
smoke signals during power outages.


Yep. We've had a couple of electrical outages the past few months and I used
my cell phone to contact PGE. Both times the company sent me a text message
that power was back on several hours after it actually was restored. :-)

I guess I should stick with the monthly rental because I have other uses for
my time than hassling with Ziply. Another example of the power of monopoly
to affect consumers.

Regards,

Rich

Re: [PLUG] Replace ZiplyFiber VoIP router

[Rich Shepard]

On Thu, 29 Feb 2024, MC_Sequoia wrote:


I don't have Ziply service and I don't know much about it, so I did some
quick Internet sleuthing and combined with my many years of telcom
networking experience, here's some maybe useful guidance I can offer.


I, too, looked on the 'Net, but I lack the many years of telcom networking
experience. :-)


1. Based on my understanding, what I think the install tech is talking
about is a "fiber optic router", not a voip router, that connects to the
ONT (Optical Network Terminal) that's installed on the outside of
house/building with a FTTP (Fiber To The Premises) service like Zipley
provides.


The installation tech upgraded the ONT (in place since Aracnet suddenly
died) with Ziply's current model.


Ref - https://www.broadbandsearch.net/blog/ziply-fiber-modem


Yep, the one on the left looks like my model NVG578LX. A reddit thread

suggests "Indeed, on fiber service the ONT fulfills the role of a modem so
you can use whatever router you want, as long as it can do DHCP, and can
connect to the twisted pair Ethernet port on the ONT!

"Raspberry Pi? sure. Simple single off the shelf router? sure. Mesh system?
sure. A full on rack mount server? sure."


To further support my logic about this, VOIP can be handled by any generic
off-the-shelf wifi router. because all it's doing is passing IP packets
that contain voice data.


I have a Meraki WAP running openWRT that works just fine.


If this is about VOIP router and not a fiber optic router, the other
reason to use Zipley's router, would be for VOIP Quality Of Service
configuration and settings that they've worked out with a particular
router and its OS that they know works and they can support.



Here's a link to a Reddit thread on Zipley VOIP that might be useful,
https://www.reddit.com/r/ZiplyFiber/comments/12fupvw/a_few_questions_about_voip_from_ziply_fiber/



First, I'd confirm whether the install tech and you are talking about the
same piece of equip. I suspect it's fiber optic router, he's talking
about, but you'll need to do some more research to confirm.


Apparently the eqipment I have is a WiFi-6 multigig fiber optic gateway made
by CommScope and is used by Ziply. So, might the TP-Link Deco XE-75 work as
a replacemen? 

Thanks,

Rich

[PLUG] Replace ZiplyFiber VoIP router

[Rich Shepard]

When I spoke with the ZiplyFiber business tech support about migrating to
VoIP they did not tell me that I had to 'rent' one of their modems @
$12/month just for the telephone. While I recognize that POTS on twisted
pair copper wires are no long supported I have a perfectly good Ubiquity
Edge router for the 'Net and asked the installation tech why I needed their
router. His explanation seemed weak to me, but I was unaware that I'd be
paying for it each month.

I'd appreciate learning of a replacement router (no need for WiFi or
internet access) just for the VoIP on my regular Panasonic phone with
portable handset.

TIA,

Rich

Re: [PLUG] Backup Solutions

[Rich Shepard]

On Wed, 21 Feb 2024, Charles Sliger wrote:


Looking for recommendations for Linux backups. Anyone use theirs for
actual restoring of files or disks?


Charles,

Dirvish . It's based on rsync. I've been using it
for years and have restored individual files, directories, and entire
partitions when a hard drive failed.

I have a root crontab that runs every night at 00:30 backing up the
server/workstation, incrementally (i.e., only changed files.)

If you need help ask me or sign up to the dirvish mail list.

HTH,

Rich

Re: [PLUG] gawk: a pattern that separates two similar strings

[Rich Shepard]

On Wed, 14 Feb 2024, wes wrote:


you can exclude lines containing a given value with grep.
grep -v Inactive input.txt
this could cause problems if one of the other fields contained the word
Inactive but it's a start.


wes,

That's a good idea. As I just responded to Reid, without printing the first
record (the column headers) the script works. That doesn't make sense to me.

Thanks,

Rich

Re: [PLUG] gawk: a pattern that separates two similar strings

[Rich Shepard]

On Wed, 14 Feb 2024, Reid wrote:


Okay, that's not my experience. Using your script and adding a row-number field 
to the end of each line:



$ cat /tmp/tmp.txt
410 Auto Wrecking,410 Auto Wrecking,So3011780,Inactive,Industrial SW GP,27865 
Sr 410 E,Buckley,Pierce,98321,ROW_1
410 Auto Wrecking,410 Auto Wrecking,WAR011780,Active,Industrial SW GP,27865 Sr 
410 E,Buckley,Pierce,98321,ROW_2
410 Auto Wrecking,410 Auto Wrecking,So3011780,Inactive,Industrial SW GP,27865 
Sr 410 E,Buckley,Pierce,98321,ROW_3
410 Auto Wrecking,410 Auto Wrecking,WAR011780,Active,Industrial SW GP,27865 Sr 
410 E,Buckley,Pierce,98321,ROW_4

$ gawk -f /tmp/tmp.gawk /tmp/tmp.txt
410 Auto Wrecking,410 Auto Wrecking,So3011780,Inactive,Industrial SW GP,27865 
Sr 410 E,Buckley,Pierce,98321,ROW_1
410 Auto Wrecking,410 Auto Wrecking,WAR011780,Active,Industrial SW GP,27865 Sr 
410 E,Buckley,Pierce,98321,ROW_2
410 Auto Wrecking,410 Auto Wrecking,WAR011780,Active,Industrial SW GP,27865 Sr 
410 E,Buckley,Pierce,98321,ROW_4

Note that the first "Inactive" line is printed because it's the first record. The second 
"Inactive" line is not printed.


Reid,

Removing the subscr() and using this script:
BEGIN { FS = OFS = "," }
#  NR ~ /1/ { print $0 }
  $4 == "Inactive" { next }
  $4 == "Active" { print $0 }

it now works; only Active rows are returned. But, when I uncomment the first
statement to print the header row I also get the inactive rows.

Why might this be?

Thanks,

Rich

Re: [PLUG] gawk: a pattern that separates two similar strings

[Rich Shepard]

On Wed, 14 Feb 2024, Reid wrote:


The `NR ~ /1/` line is causing the first line to be printed, regardless of its 
contents.


Reid,

With the full data set that's the column header row so I left it in the
script because I'm working with the full set.

However, commenting it out I still get the Inactive records (and the Draft
ones because I've not yet removed those.)

Thanks,

Rich

Re: [PLUG] gawk: a pattern that separates two similar strings

[Rich Shepard]

On Wed, 14 Feb 2024, Rich Shepard wrote:


Would the substr() function do the job? E.g,, substr(Active,1,6)


Apparently not:
BEGIN { FS = OFS = "," }
  NR ~ /1/ { print $0 }
  $4 == substr(Active,1,6) { print $0 }

still has Inactive in the output.

Sigh,

Rich

[PLUG] gawk: a pattern that separates two similar strings

[Rich Shepard]

I've downloaded a data file where the permit type can be Active, Inactive,
or Draft.

Removing all Draft records was easy. Removing all Inactive records has so
far escaped my gawk (and LO Calc) efforts because Inactive contains the
substring 'active'.

My script:
BEGIN { FS = OFS = "," }
  NR ~ /1/ { print $0 }
  $4 == "Inactive" { next }
  $4 == "Active" { print $0 }

I've also tried with matching:
  $4 ~ /Inactive/ { next }
  $4 ~ /Active/ { print $0 }

And, of course, specifying only field 4 containing Active also fails.

My AWK books and web searches haven't taught me how to write a pattern that
matches only 'active' and not 'inactive.'

Would the substr() function do the job? E.g,, substr(Active,1,6)

Here are two sample records:
410 Auto Wrecking,410 Auto Wrecking,So3011780,Inactive,Industrial SW GP,27865 
Sr 410 E,Buckley,Pierce,98321
410 Auto Wrecking,410 Auto Wrecking,WAR011780,Active,Industrial SW GP,27865 Sr 
410 E,Buckley,Pierce,98321

TIA,

Rich

Re: [PLUG] Testing mailx script [WORKING]

[Rich Shepard]

On Thu, 1 Feb 2024, Rich Shepard wrote:


I have a list of four test addresses, all individuals @example.com. Is there
a better domain I should use to test whether the script works as intended
when sent from my local MTA?


No need. I checked some temporary address mailers but they allow only single
recipients. Instead I sent it to a few individuals who know me (and to my
company's generic user) and it works like a charm using the suggested
while() loop.

My thanks to all for the suggestions.

Best regards,

Rich

Re: [PLUG] Weather underground no longer reached by browser

[Rich Shepard]

On Thu, 1 Feb 2024, MC_Sequoia wrote:


Rich -What happens if you search on weather underground and then click on
the search result?


MC,

It loads just as it has for years. Sonofagun!

I had tried entering wunderground.com in a new tab and that also failed.

Strange.

Thanks for the pointer.

Rich

Re: [PLUG] Testing mailx script

[Rich Shepard]

On Thu, 1 Feb 2024, MC_Sequoia wrote:


Rich - If you do a web search on temporary email addresses, you can
quickly and easily get an email address or addresses to use for testing
purposes.


That would not have occurred to me. So much to learn and so little time.

Many thanks!

Rich

Re: [PLUG] Weather underground no longer reached by browser

[Rich Shepard]

On Thu, 1 Feb 2024, Rich Shepard wrote:


I tried refreshing https://www.wunderground.com and both brave and
chromium tell me that URL cannot be reached: Atlanta (404) error code
displayed. Yet I can ping the URL. What might explain this?


Strange. I, too, can load it on the mobile phone.

Perhaps tomorrow morning it will have recovered from whatever ails it now.

Rich

[PLUG] Testing mailx script

[Rich Shepard]

I have a list of four test addresses, all individuals @example.com. Is there
a better domain I should use to test whether the script works as intended
when sent from my local MTA?

TIA,

Rich

[PLUG] Weather underground no longer reached by browser

[Rich Shepard]

I tried refreshing https://www.wunderground.com and both brave and chromium
tell me that URL cannot be reached: Atlanta (404) error code displayed. Yet
I can ping the URL. What might explain this?

TIA,

Rich

Re: [PLUG] mailx: use of -S option

[Rich Shepard]

On Wed, 31 Jan 2024, MC_Sequoia wrote:


Rich - See example 6:


Ah! I missed that.

Many thanks,

Rich

Re: [PLUG] mailx: use of -S option

[Rich Shepard]

On Wed, 31 Jan 2024, Reid wrote:


I don't know about `mailx -S`, but you could put a placeholder in your msg
file like "Hi NAME_GOES_HERE". Then something like:

#!/bin/sh

while read addr <&3 && read name <&4; do
   sed "s/NAME_GOES_HERE/$name" msg | mailx -s "subject here" $addr
done 3

Reid,

This looks promising. I was stuck on a placeholder at the top of the message
and putting the right name with the right address since both input files are
in the same sequence.

Thanks,

Rich

Re: [PLUG] mailx: use of -S option

[Rich Shepard]

On Wed, 31 Jan 2024, MC_Sequoia wrote:


Here are some examples on use of the mailx -s option:
https://www.binarytides.com/linux-mailx-command/


MC,

The '-s' option means subject; the '-S' option is a keyword:value pair.

Rich

[PLUG] mailx: use of -S option

[Rich Shepard]

The mailx man page describes the -S option:

-S variable[=value]
Sets  the  internal option variable and, in case of a string option,
assigns value to it.

I've not found an example of its use.

I have a working mailx script:
#!/bin/sh

for ADR in $(cat mailing-addr.list)
do
  cat msg | mailx -s "Subject here $ADR
done

Is there a way to use -S to add the first name salutation associated with
each email address in mailing-addr.list by putting that list of names in
another file?

If not, is there a way I would expand the shell script to add that
salutation to each recipient? For example, an associated for loop? (I'll
check my bash scripting book this evening.)

TIA,

Rich

Re: [PLUG] SSL/TSL only for web sites?

[Rich Shepard]

On Mon, 29 Jan 2024, Paul Heinlein wrote:


If you're sending via Postfix, no, you don't need them for SMTP stuff.

Postfix itself can be configured to use SSL/TLS, but that's completely 
separate from what you're trying to accomplish.


Paul,

Thank you. I just found that I don't need them when reading about the Python
modules.

Much appreciated,

Rich

[PLUG] SSL/TSL only for web sites?

[Rich Shepard]

I'm developing a Python script to send email messages. Researching SSL/TLS
on the web suggests that these are used with web applications (along with
HTTPS).

Since I'm sending messages from the command line (or Python script) out
through my local Postfix MTA, are these security layers necessary or needed?

I've not seen an answer in my web searches.

TIA,

Rich

Re: [PLUG] Cable tester

[Rich Shepard]

On Mon, 29 Jan 2024, mo wrote:


Hi. I need to buy a cat5 cable tester aka tone detector. There are so many!
How should I choose one? What features, brands, etc do you recommend?


mo,

I've used a fox & hound audio cable tester for phone lines with RJ-11 plugs
since I moved into this house 31 years ago. Without fetching it now I assume
it can work with computer data cables and RJ-45 plugs, perhaps needing an
adapter.

HTH,

Rich

Re: [PLUG] Edit PDF

[Rich Shepard]

On Mon, 29 Jan 2024, Charles Sliger wrote:


You basically have to do the layout and formatting manually. I use it
frequently to edit invoices and receipts that come in pdf form so that
they print on a single page. I change the page layout to landscape, use
snap lines, change font size, etc.


Charles,

Take a look at pdfjam. The man page begins with
pdfjam(1)   
 pdfjam(1)

NAME
   pdfjam - A shell script for manipulating PDF files

SYNOPSIS
   pdfjam [OPTION [OPTION] ...] [SRC [PAGESPEC] [SRC [PAGESPEC]] ...]

DESCRIPTION
   pdfjam  provides  a  front  end  to most capabilities of the "pdfpages" 
package (by
   Andreas Matthias) of pdflatex.  Detailed  information  can  be  found  via  
"pdfjam
   --help", and also in the web page mentioned below .

   A working installation of pdflatex, with the pdfpages package, is 
required.

   The  pdfjam  script  is  distributed as (the main) part of the pdfjam 
package.  The
   homepage of pdfjam is at https://github.com/rrthomas/pdfjam .

SETUP
   See https://github.com/rrthomas/pdfjam

I use it to resize PDF web page downloads to letter size, and a few other
things.

Regards,

Rich

Re: [PLUG] Edit PDF

[Rich Shepard]

On Sat, 27 Jan 2024, mo wrote:


Nothing beats Master PDF


+1 MasterPDFEditor 

I've been using the free version for years.

Rich

Re: [PLUG] Understanding postfix mail rejection message

[Rich Shepard]

On Wed, 24 Jan 2024, wes wrote:


next step would be to run
host example-com0i.mail.protection.outlook.com
I'm betting the problem is at the other end.


Wes,

Certainly appears that way:
$ host example-com0i.mail.protection.outlook.com
Host example-com0i.mail.protection.outlook.com not found: 3(NXDOMAIN)

I'll call and see what seems to be a general issue.

Thanks,

Rich

Re: [PLUG] Understanding postfix mail rejection message

[Rich Shepard]

On Wed, 24 Jan 2024, Michael Ewan wrote:


Correct.  Somewhere in your stack a gethostbyname() call is failing.  There
are many places to look for misconfiguration, start at the command line and
type 'host example.com' and see what you get.  My results (Linux Mint) were,

example.com has address 93.184.216.34
example.com has IPv6 address 2606:2800:220:1:248:1893:25c8:1946
example.com mail is handled by 0 .


Michael,

Here,
$ host example.com
example.com mail is handled by 0 example-com0i.mail.protection.outlook.com.

Am I stuck somewhere in M$-land?

Thanks,

Rich

[PLUG] Understanding postfix mail rejection message

[Rich Shepard]

Example (redacted):
: Host or domain name not found. Name service error for
name=example-com0i.mail.protection.outlook.com type=A: Host not found

My web search suggests that my postfix MTA is not able to obtain the proper
DNS record for example.com. Is this the correct interpretation?

TIA,

Rich