[GTALUG] Securely wiping SSDs

[Giles Orr via talk]

I have, for many years, used "Darik's Boot and Nuke" on a USB stick to
securely wipe spinning hard disks.  It takes a long time, but I mostly
understand and trust the process.

I'm now at the point that I have to wipe and dispose of SSDs, and I'm
feeling a bit shaky on the methodology.  Here's what I did:

# hdparm -I /dev/sdX

Looked for enabled/locked/frozen in the output ... I won't go into
making sure those are toggled correctly, but that appears to be
needed.  This also lists what appears to be info about doing a wipe on
the drive:

6min for SECURITY ERASE UNIT, 60min for ENHANCED SECURITY ERASE UNIT

Then, set a password (why? but seems to be needed):

# hdparm --user-master u --security-set-pass foobar /dev/sdX

Last, run the wipe:

# hdparm --user-master u --security-erase-enhanced foobar /dev/sdX

Doing something like `dd if=/dev/sdX bs=5M count=5 | strings` (or
sending it to `less`) definitely shows that it's changed from
something organized to something full of identical characters.  But
I've never seen this wipe process take more than 60 seconds, which
makes me wonder about the `hdparm` declaration about the time required
for a secure wipe.

So I guess the big question is: should I trust this process?  Do we
really think it's securely wiped?  Or should I be taking a hammer to
the chips on the SSD because that's the only way to ensure it's fully
wiped?

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Local-only mail server

[Giles Orr via talk]

I figure this is probably a solved problem, but it's one I've never
looked at.  And I admit to doing no research yet: I thought pointers
from this group would be the place to start.

I find I have a lot of mail generated by cron and the like on several
local machines, and I'd like to A) centralize that mail on a local
server, and B) be able to view that with Thunderbird.  Platform is
Debian.  It's not my intention to handle my Gmail or anything like
that: this is meant exclusively for handling mail from local
computers.  Which would also mean the ability to send mail isn't
important.

Currently if I want to look at local administrative mail, I have to
use the nasty `mail` command from the "bsd-mailx" package.  (Yes, I
could install mutt - I don't like it much better.  I'm well aware it's
much more powerful.)  But do I remember to look at these emails?  No.
If instead each host forwarded these mails to a single host, and I
pointed Thunderbird at that ...  I'd have some hope of managing this
problem better.

"exim4" is usually (not always) installed on Debian systems.  It
claims to be an MTA: my knowledge of mail is so bad I don't even know
if that means it just "pulls" or is also capable of forwarding local
mail to another machine?  And what would be the recommended server
software?

Another possible solution would be to send the mail directly to a
folder on the host that has Thunderbird without using a server?  With
the snag that the host in question isn't always on, and I wouldn't
want it to run mail server software.  Is that possible?

Yes, I know this is probably bigger than I think.  I'd still like to
start poking around and finding out about it.  Suggestions welcomed,
thanks.

--
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Debian 12 takes too long to boot and login

[Giles Orr via talk]

On Sun, 17 Dec 2023 at 17:52, Ron / BCLUG via talk  wrote:
>
> Kevin Cozens via talk wrote on 2023-12-17 14:43:
>
> > Any clues as to why Debian takes 4 to 5 times as long? I'm hoping there
> > is some bad configuration out of the box causing Debian 12 to be acting
> > so slow. Anyone have any ideas where I should start looking? If I can't
> > get to the bottom of the problem I will be staying with Linux Mint.
>
> Try `dmesg | less` and see if anything eye-catching is in there.
>
>
> Also, `journalctl` with some option like `--boot=-1` for *previous* boot
> logs, or for running in Mint with separate partitioning, `journalctl
> --directory=/path/to/debian/logs`.
>
>
> Note that `journalctl -k` or `journalctl --dmesg` will show only kernel
> messages, like `dmesg`...
>
>
> Hope this helps somehow.

One of my favourite recent discoveries are the abilities of
`systemd-analyze`.  `systemd-analyze blame` shows how long every step
in the boot process took.  It's worth looking at, but doesn't account
for the fact that many of these things run in parallel.  So to better
understand what's causing the _real_ delays, try `systemd-analyze
critical-chain`.  The output is fascinating and I hope helpful ... but
please keep in mind that I never got further than that so I can't tell
you how to speed up the problem areas ...  But the last command may at
least focus your search in the `dmesg` output.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Protonmail WAS: Re: Linux Unicorns

[Giles Orr via talk]

On Tue, 28 Nov 2023 at 12:13, Evan Leibovitch via talk  wrote:
> Both the choice of OS and where hosted is infrastructure, operational issues, 
> and rarely part of a startup's pitch. If there is a privacy or security 
> component to the proposal's added value the hosting may be a factor (ie, not 
> hosted in a Five  Eyes country) but I think even that is less of a factor 
> with revelations that much claimed protection is an illusion.

( 
https://www.forbes.com/sites/thomasbrewster/2023/08/08/protonmail-fbi-search-led-to-a-suspect-threatening-a-2020-election-official/
)

You're unhappy with ProtonMail?  You thought they should protect all
anonymity in defiance of the law?  I think these quotes (from the
article you cited) are worth reading:

"Proton did say, however, that by Swiss law, for U.S. agencies to get
information on Proton accounts, they have to go through Swiss
authorities."

"The FBI didn’t get much back from Proton, but it did receive the
recovery and associated email addresses linked to the ProtonMail user.
They proved crucial, enabling the FBI to find more information about
the person online and conduct a sweep across the suspect's internet
accounts, including on Amazon, Apple, Coinbase, Google, PayPal and
Spotify."

I'm fond of the law.  I like to see it enforced.  Protonmail lives in
Switzerland, under Swiss law ... I'm okay with them complying with
those laws.  Sure Protonmail isn't perfect - but it's a hell of a lot
better than Gmail.

(Yes, I do use Protonmail.  I pay them - they don't pay me to say this.)

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] labwc is unhappy with Debian 12 and LightDM ...

[Giles Orr via talk]

I'm trying to get `labwc` running on a Debian 12 system using LightDM
as the login manager.  I chose labwc because I'm a long-time Openbox
user, and labwc is supposed to be almost a drop-in replacement for
Wayland.  Unfortunately, it's not available as a package so I had to
build it myself.  Which went reasonably well - although it demands a
newer version of wlroots than Debian has (Debian has 0.15.1, labwc
wants 0.16), so it downloaded and built that itself.  And at the end
of this, I ran `meson install --skip-subprojects -C build/` and it put
stuff in the /usr/local/ tree.  This dropped a
/usr/local/share/wayland-sessions/labwc.desktop file which LightDM has
ignored.  Probably fixable, but I instead copied it to
/usr/share/wayland-sessions/ alongside "weston.desktop".  This worked,
in the sense that I now have a "labwc" option at the LightDM login
screen, but didn't work in that attempting to use it blacks the screen
for 5-10 seconds and then returns to LightDM.  I should add that the
"weston" session works as expected.  Also, running `labwc` inside an X
session produces a window with a functioning labwc session inside of
it.  But of course I want to log directly into labwc.

Any thoughts or suggestions appreciated.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Old (mostly) O'Reilly books available free

[Giles Orr via talk]

I have a number of O'Reilly books that I haven't used in years (and a
couple other tech titles as well).  It's time I got rid of them, and
it seems a waste to toss them in the recycling.  If any of you are
interested in these, pick up will be in North York: either North York
Central Library or nearby.  Please see the image below for the titles
- and note that if there are modern versions of these titles, the
copies you see here are almost certainly NOT the most recent version.

https://gilesorr.com/forOthers/books.jpg

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Debian has suddenly become unstable

[Giles Orr via talk]

On Thu, 12 Oct 2023 at 11:09, D. Hugh Redelmeier via talk
 wrote:
> | From: Giles Orr via talk 
>
> | I have a Debian 12 system that's my daily driver.  In the last two
> | days, it crashed twice when I was away from the keyboard and nothing
> | was happening (around the same time of day now that I think about it).
>
> | I'm not great at debugging Linux crashes.  The `dmesg` command is
> | useless, as it only shows the log since the last boot.
>
> I think that everything in dmesg goes to the Journal.
>
> Have you compared journal entries from just before each crash to see if
> there is a common theme?
>
> Good luck!

Thanks to BCLUG for `journalctl --boot -1` (and I assume `-2` etc.).
That's a blessing.  I ended up running `systemctl disable cups.service
cups.socket cups.path` and `systemctl stop cups.service cups.socket
cups.path` - so more or less what you were trying to suggest.  :-)
The system didn't crash yesterday, so that's good.

Hugh seems to be correct: I think everything in `dmesg` ends up in the
journal.  But what I find interesting is that not everything in
/var/log/systemlog is in the journal.  I was comparing the systemlog
entries, and that's how I concluded CUPS/printer drivers were the
problem.  The loading of those modules weren't mentioned in the
journal.  In this case, it seems that the information I most needed
(ie. "this is a printer driver problem") came from syslog.

This is a negative test case - ie. I don't know it's solved, and won't
ever be certain.  Unless it crashes again, then I know it's not
solved.  Ugh.

Thanks everyone.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Debian has suddenly become unstable

[Giles Orr via talk]

I have a Debian 12 system that's my daily driver.  In the last two
days, it crashed twice when I was away from the keyboard and nothing
was happening (around the same time of day now that I think about it).
The system has previously been very stable, usually up for a month at
a time with reboots only to pick up new kernels.  I should note that
when I turned it on and ran upgrades on Monday after a week away, it
upgraded a lot of packages for Debian release 12.2.

I'm not great at debugging Linux crashes.  The `dmesg` command is
useless, as it only shows the log since the last boot.  So I turned to
/var/log/syslog.  What I noticed was this, the only line of
consequence about a millisecond before the reboot:

2023-10-10T11:36:23.839046-04:00 sli7d systemd-modules-load[399]:
Inserted module 'lp'

I don't have a printer, and I hadn't just done a "print-to-PDF" or
anything like that - the machine had been idle for a couple hours.
This morning it crashed again, and milliseconds before the crash I
found these (again, the machine was idle when this happened):

2023-10-11T12:20:54.647048-04:00 sli7d systemd-modules-load[382]:
Inserted module 'lp'
2023-10-11T12:20:54.647254-04:00 sli7d systemd-modules-load[382]:
Inserted module 'ppdev'
2023-10-11T12:20:54.647280-04:00 sli7d systemd-modules-load[382]:
Inserted module 'parport_pc'
2023-10-11T12:20:54.647290-04:00 sli7d lvm[372]:   3 logical
volume(s) in volume group "primary" monitored
2023-10-11T12:20:54.647302-04:00 sli7d systemd[1]: Starting
systemd-journal-flush.service - Flush Journal to Persistent Storage...
2023-10-11T12:20:54.647312-04:00 sli7d systemd-udevd[398]: Using
default interface naming scheme 'v252'.

I just ran `apt full-upgrade` (right now) and watched it upgrade
Samba.  Is it possible that Samba was triggering "lp"-related stuff
which was causing the crash?  Although why it would cause a crash I
don't know.  No new kernel (and thus no new modules).  I suppose I
could reboot and select and older kernel and see if that was stable
...

Suggestions on how to better debug this would be most welcome.  Does
blacklisting the "lp" module sound like a good idea?  Any other ideas?

Re-installing would be ... unpleasant.  This is my primary machine and
heavily tweaked-up.  But I guess I'll do that if I have to.  Keeping
it as a last resort though (daily crashes would get me there!).

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] looking for a general purpose boot media to keep in case of distasters

[Giles Orr via talk]

On Wed, 20 Sept 2023 at 15:16, BCLUG via talk  wrote:
>
> bitmap via talk wrote on 2023-09-20 10:02:
>
> > Does this OpenSUSE image come with a broad range of tools? The debian,
> > ubuntu, manjaro and others I have lying around are just the very basics.
> > Am often missing something needed.
>
> I seem to recall that Knoppix was a popular distro for repairing broken
> systems. This may be out-dated though.
>
>
> Also, ventoy is a tool that allows multiple bootable distros on a single
> USB stick, and I think allows some persistent storage.
>
> i.e. Use ventoy to install knoppix / whatever on a stick, boot the stick
> and install any missing tools, configure sshd, add scripts, etc., then
> next boot time everything ought to be available.
>
>
> I haven't yet tried this myself, just giving my 2¢ on how I'd approach
> things.

I started using Knoppix about 25 years ago.  Development has slowed,
but there are still occasional releases - the most recent being 9.1
from 2021-01.  It will boot on just about anything x86 or amd64 (at 4G
or less memory it switches to a simpler interface ... it may even
choke down to text mode eventually?), and come up to a fairly simple
GUI with pretty much all the tools you could hope for.  It remains my
go-to for recovery and repartitioning.  I'm still using v7.7.1 (2016 -
LXDE interface) because that's what's on my bootable USB stick: it has
thousands of tools, but ones I checked for today include 'inxi',
'gparted', and 'lshw'.  Keep in mind this is the DVD version which has
more tools than the CD version.  Most of the command line tools are
still on the CD, the bulk of the DVD space is for less interesting
stuff like Libre Office.

Ventoy sounds great, haven't tried it yet.  I suspect raw Knoppix will
boot on more machines than Ventoy, but that's idle speculation and a
quarter century of success with Knoppix talking.

http://knopper.net/knoppix-mirrors/index-en.html
http://knoppix.net/get.php

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] installing Windows is uselessly worse than installing Linux

[Giles Orr via talk]

On Fri, 15 Sept 2023 at 14:35, D. Hugh Redelmeier via talk
 wrote:
>
> [Always do backups before such drastic measures.]
>
> The best way to install Windows is to go to another Windows box and use
> the the Window Media Creation Tool to create a bootable USB stick
> 
> Alternatively you could download the WIndows 11 Disk Image.
>
> During installation, you have to answer a lot of questions.  It is
> extremely hard to evade linking the installation with a global Microsoft
> account (google to learn the ever-evolving tricks).
>
> Good news: Windows can let other partitions on the disk remain.  In
> earlier years, it would just grab the whole disk.
>
> Once it is installed, Windows will spend a lot of time installing updates
> and rebooting.  Installing updates is quite slow (perhaps due to
> anti-malware software).
>
> And it lies about whether there are more updates.  When it says there are
> none, ask it to look again.
>
> You also need to separately run the Microsoft app store app and ask it to
> do updates.
>
> You also need to run you system vendor's update thingee too for some
> driver and firmware updates.  Even then, it is useful to see what firmware
> is available on the vendor's web site
>
> This whole process involves multipler reboots, each requiring babysitting.

Regular readers of this list may recall that I recently bought a
mini-mini-computer: the slim instruction manual basically said that if
you wanted to avoid trouble (which I take to mean "having to connect
to a global M$ account"), I should boot WITHOUT a network connection.
Which I did, and it seemed to create a local account without
hesitation.  At which point I shut the machine down, imaged the drive,
wiped it, and installed Linux.  Windows might have fought me hard the
second I booted with a network connection, but that really seems to be
a great way to win that first fight.  How long they'll leave that door
open I don't know - I expect Windows 12 to say "I have no network
connection, therefore this is not a real computer.  Shutting down
now."

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] USB to Ethernet Dongles WAS: Debian Linux as-a-router Guide

[Giles Orr via talk]

On Fri, 8 Sept 2023 at 14:49, D. Hugh Redelmeier via talk
 wrote:
> More to your point, it may be convenient for your router to have more than
> two ethernet ports.  Giles' box only has two, yours and mine have four.
> (Giles's box sure is cute.)
>
> Common wisdom has it that USB ethernet dongles are not always stable 24/7.---

This is a subject I know way too much about - which means you've
triggered a story that can be skipped by those not interested.  The
main point: I've never found a totally stable USB-to-Ethernet dongle -
and I've tried quite a few.

I'm fairly hardcore about having my computers connected to the network
via wires rather than WiFi.  I do use WiFi occasionally, to sit on the
balcony or the couch - but the wireless router is on a physical switch
which is off most of the time.  Which means I've become very familiar
with USB-to-Ethernet dongles and their quirky behaviour.  The most
stable I've ever dealt with are the Apple-branded ones attached to
Apple computers (they're pretty good attached to non-Apple computers
as well, although I rarely use them that way).  But Linux computers
with USB-to-Ethernet dongles are never totally stable.  I would say I
get a couple minutes of network drop-outs per day (across multiple
brands).  Which isn't a problem when I'm not in front of the computer
(they're clients, not servers), but when you're using Barrier to use
one computer's mouse and keyboard to control another computer, a
network outage knocks you off the second computer completely.

Sometimes these outages resolve themselves, occasionally (rarely) I
have to run `dhclient` by hand on the machine with the dongle.  I've
never dug into the logs to figure out why.

I have multiple USB3-to-Ethernet dongles:
- one "Amazon Basics" - fairly good, but probably the most drop-outs?
- two Anker "Unibody Aluminum" - slightly better
- three Orico 3-port USB hub + Ethernet - these are noticeably more
stable (still not perfect) and, because of the added USB ports, more
useful
- absolute worst: Belkin USB-C "docking bay" thingy: the Ethernet port
on this bounced every 30 seconds to 1 minute, totally unusable (and of
course being a "docking bay" it cost much more than the others).  I
ended up plugging one of the Ankers into a USB port on the docking
bay!

I've never seen this instability with built-in Ethernet ports: they
work or they don't, end of story.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Debian Linux as-a-router Guide

[Giles Orr via talk]

As per my previous post, I just purchased a mini-PC which I intend to
turn into a router.  Is anyone aware of a guide for turning a Debian
PC into a _home_ router?  I'd like to be running probably DNSmasq,
using a blocklist, stuff like that.  I've found webpages that tell me
how to turn on network forwarding, or maybe configure DNSmasq, but not
the whole process.

Please don't suggest pfsense: I'm well aware of it, and it may well be
better.  But I'm very adept at managing Debian, and initially at least
I intend to try to set this up.  If it turns out to be direly
difficult, pfsense may happen later.

Thanks.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Cheap possible router PC?

[Giles Orr via talk]

I'm going to post twice about this ... but this email is simply about
a good deal and possible router PC:

https://www.amazon.ca/dp/B0C2HLRV41

Tiny (9x9x5cm), colourful (LED strip at the base, could definitely
live without), 8GB RAM, 256GB SSD, Intel 12th Gen 4-core N95, 3x HDMI,
2x Ethernet - $188 all in with free shipping (notice the large $84
coupon you need to apply).  It's also available with more RAM and/or a
bigger SSD.  Seemed like a good deal to me - feel free to comment ...

Mine has arrived, and I had no trouble at all installing Debian 12
headless on it (it comes with Windows 11 if that's your thing).

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] debian: adduser doen't add user to many useful groups

[Giles Orr via talk]

On Sun, 13 Aug 2023 at 18:43, D. Hugh Redelmeier via talk
 wrote:
>
> | From: Znoteer via talk 
>
> Thanks for your reply.  It seems that I didn't miss anything obvious.
>
> It would be good to have a list of the powers granted by each group
> measurement.  This belongs in debian documentation.  Unfortunately there
> are several reasonable places.  group(5)?
>
> Interesting/inconvenient change to EXTRA_GROUPS.  I wonder why?  Probably
> security.
>
> | On Sun, Aug 13, 2023 at 04:12:18PM -0400, D. Hugh Redelmeier via talk wrote:
>
> | > Is there a magic shortcut to getting hugh added to all these groups?
> |
> | Don't know about a magic shortcut, but here is the command to add groups to 
> a user:
> |
> | $ sudo usermod -a -G "group1,group2,group3,,groupN" hugh
>
> I found a hack easier.  Just use a text editor's global search and
> replace silly$ with silly,hugh.  Except for group silly.

Thanks for walking through the details everyone.  It's something I'd
wondered about and never pursued - good to know the ins-and-outs.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] OLD Thinkpads, free

[Giles Orr via talk]

Hi, it's too late.  I gave it a week and ripped them apart.  Sorry.

On Wed, 2 Aug 2023 at 05:06, D. Hugh Redelmeier  wrote:
>
> Is it too late?
> These might be good for Karen.
>
> | From: Giles Orr via talk 
> | To: GTALUG Talk 
> | Date: Sat, 22 Jul 2023 10:56:57 -0400
> | Subject: [GTALUG] OLD Thinkpads, free
> |
> | 2 x T60, 1 x T42
> |
> | I have somewhat involuntarily acquired these three exceptionally old
> | Thinkpads.  None of them are fully operational although all do
> | something resembling booting.  None have HDs, and they have other
> | minor ailments as well.  No power supplies.  My inclination is to rip
> | them apart (I collect old CPUs), but if someone has an actual use for
> | them, we can talk ...  You would probably have to come to North York
> | to collect them: I nearly broke my back carrying all three home in my
> | backpack on the TTC.  They are not light.
> |
> | These should be considered for parts: they sorta-maybe run, but when I
> | give up on installing Linux on them, you can rest assured they're
> | either badly underpowered or problematic or both.
> |
> | I only ask because I know a lot of people on this list (myself
> | included) really like Thinkpads.  If no answer in a week, I'll gut
> | them (and that's fine too).
> |
> | --
> | Giles
> | https://www.gilesorr.com/
> | giles...@gmail.com
> | ---
> | Post to this mailing list talk@gtalug.org
> | Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
> |



-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Favorite desktop manager?

[Giles Orr via talk]

On Tue, 25 Jul 2023 at 14:19, mwilson--- via talk  wrote:
>
> I just installed Debian 12 on a new laptop and took Gnome as
> desktop/window manager.
>
> I really insist on something more Xerox PARC style.
>
> I had been using Xfce, and liked it until (following Debian 10 software
> updates) it began to act strangely.  A few months ago it started to kick
> up a huge fuss about difficulties starting the Panel after a restart.
> About 2 weeks back, following an upgrade, the Desktop area stopped
> displaying anything -- no wallpaper, no icons from the Desktop folder --
> just gray.  That prompted this move to Debian 12.
>
> I might go back to Xfce, and assume my troubles just resulted from being 2
> major releases back, but before that, what do other people use and like?

I personally use Openbox.  I loathe GNOME.  On the rare occasion I
want to install a "desktop" rather than just a Window Manager, I
usually use LXDE which is the lightest weight of the lot.  Its
relative simplicity might work for you.

If you're interested in window managers (not "desktop environments,"
per se) I maintain as complete a list as I can manage here:

https://gilesorr.com/wm/table.html

Apologies for the self-promotion, but it's something I'm interested
in.  The page isn't well suited to discovery-by-features though, it's
more just a list.

One odd possibility no one has mentioned is the Trinity Desktop (
http://trinitydesktop.org/ ), which is the continuation of KDE3.  I
haven't used it, but it's still maintained and it might be of interest
in this conversation?

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] OLD Thinkpads, free

[Giles Orr via talk]

2 x T60, 1 x T42

I have somewhat involuntarily acquired these three exceptionally old
Thinkpads.  None of them are fully operational although all do
something resembling booting.  None have HDs, and they have other
minor ailments as well.  No power supplies.  My inclination is to rip
them apart (I collect old CPUs), but if someone has an actual use for
them, we can talk ...  You would probably have to come to North York
to collect them: I nearly broke my back carrying all three home in my
backpack on the TTC.  They are not light.

These should be considered for parts: they sorta-maybe run, but when I
give up on installing Linux on them, you can rest assured they're
either badly underpowered or problematic or both.

I only ask because I know a lot of people on this list (myself
included) really like Thinkpads.  If no answer in a week, I'll gut
them (and that's fine too).

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Debian releases

[Giles Orr via talk]

As I mentioned in a previous email, I'm kind of intrigued by which toy
Debian releases will be named after.  To quote Wikipedia: "Debian
distribution codenames are based on the names of characters from the
Toy Story films."  The current release is "bookworm," and I was amused
to see that the current testing branch is called "trixie."  According
to Wikipedia's page on Debian's version history,
https://en.wikipedia.org/wiki/Debian_version_history, the release
after "trixie" will be "forky."  If we've progressed to "Toy Story 4,"
I look forward to Key and Peele's characters "Ducky" and "Bunny" -
although the names aren't the most exciting.  And of course we can
hope for "Caboom" - named after Keanu Reeves' decidedly Canadian
character ("for you, Rejean!").

--
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Red Hat Paywall...

[Giles Orr via talk]

On Tue, 4 Jul 2023 at 23:01, D. Hugh Redelmeier via talk
 wrote:
>
> | From: Giles Orr via talk 
>
> Thank you and others for telling us about your experience.
>
> |   If I need something newer, I can "pin" a
> | package, getting it either from testing or backports.  This is
> | admittedly a PITA to set up, and I hardly ever use it because of that
> | ... but I _have_ used it, and it proved reasonably reliable in
> | practice.  Although getting your head around pinning weights is ...
> | nasty.
>
> Can you use that to get the newer FireFox that you need?

Uhhh ...  I had never considered it, which seems more than a little
blind.  I haven't used pinning in a couple years, and when I did it
was usually to get a piece of software that wasn't in the current
distro (but was in the upcoming release).

However, on closer examination (I used this page:
https://www.debian.org/distrib/packages - you should become close
friends with this if using Debian) "testing" also includes Firefox
ESR, so the answer would appear to be no.  Makes sense in a depressing
way: Debian wants a _stable_ version for stable, and testing is for
testing things that will become the stable distro ...

Huh - testing is called "trixie."  I always wonder which toy is coming next ...

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Red Hat Paywall...

[Giles Orr via talk]

A couple of points following on to 01bigtenor's reply:

According to Wikipedia: "Devuan is a fork of the Debian Linux
distribution that uses sysvinit, runit or OpenRC instead of systemd.
Devuan aims to avoid 'lock-in' by projects like systemd and aims to
maintain compatibility with other init systems to avoid detaching
Linux from other Unix systems."  When systemd was initially a horrible
mess, I was very much in favour of this.  But for the last several
years systemd has been rock-steady, and once you wrap your head around
the basics, it's a LOT easier to use than maintaining those damn
/etc/rc.N/ folders.  At least that's been my experience.

As for Debian stable vs. testing: 20 years ago Debian testing was
great ... about a decade ago I found it to be noticeably less stable,
and the constant volume of updates got on my nerves.  So I switched to
stable.  Even if the stability of testing has improved, I don't want
to put up with the updates.  If I need something newer, I can "pin" a
package, getting it either from testing or backports.  This is
admittedly a PITA to set up, and I hardly ever use it because of that
... but I _have_ used it, and it proved reasonably reliable in
practice.  Although getting your head around pinning weights is ...
nasty.

On Tue, 4 Jul 2023 at 08:24, o1bigtenor via talk  wrote:
>
> On Tue, Jun 27, 2023 at 10:22 AM D. Hugh Redelmeier via talk
>  wrote:
> >
> > This Red Hat change concerns me.
> >snip
>
> >
> >   It feels as if RH steers the future of Linux by making so many
> >   contributions.
> >
> > - Ubuntu LTS + fresh Ubuntu has been pretty good.  I've had more
> >   problems with package updates on Ubuntu than of Fedora, but it has
> >   been pretty good.  Distro version upgrades have been good but not
> >   perfect in my modest experience.
> >
> >   Canonical has repeatedly acted in ways that offend or scare me.  So
> >   Ubuntu, although easy, feels like a potential trap.
>
> Found this to be true - - - - spent a lot of time a number of years ago
> looking into LXD  - - - the snap environment is one that NEVER will
> be seen here again if I can help it.
>
> Canonical is, imo, desperately looking for ways to monetize their
> brand - - - - - someone's just have to have more $.
> >
> > - debian Stable + Testing + Unstable.  I don't have much experience
> >   with debian.  I fear that the lack of full-time paid engineers might
> >   reduce the safety relative to RH (that could easily just be FUD).
> >   debian's goals are good by me.
> >
> > So: I'm thinking of switching to debian.
>
> I was with Debian for over 10 years - - - have now switched to
> Devuan - - - thereby getting rid of another item of 'control'.
> >
> > I'd like to learn from others.  How do you choose to solve these
> > problems?  Maybe some of them are non-problems.
>
> The problem is that the small encroachments don't tend to isolate
> - - they tend to grow - - - somewhat like microbes! (With similar results
> in my experience!).
> >
> > 
> >
> > Giles has a problem with needing a stable distro with a more recent
> > FireFox.  I suggested, against my preferences, that this might be a
> > perfect use for Snaps/Flatpacks.
>
> Snaps are a system controlling adventure - - - be aware of this BEFORE
> starting down that road. Haven't used flatpacks.
>
> I find that browsers are needing updates almost on a daily basis.
> Am wondering if there is a way of reducing the 'encroachment' of
> the nefarious bits of cruft adhering to all browsers (at least as far
> as I see).
> >
> > I wonder if I should be using a stable distro everywhere but with
> > containerized upgraded packages where they matter.  I yet don't think
> > so.
> >
> > The rest of my family uses Fedora on their workstations.  But they
> > hate applying updates (even when I do the work).  They are way behind
> > most of the time.  Maybe a stable distro + a fresh FireFox would be
> > best for them too.
>
> I've already been informed that if I weren't doing updates windows would be
> applied post haste - - - something about the devil they know (Work systems
> are all M$ Win!)
> >
> > How many other packages would I need to have fresher-than-stable?
> >
> > - support for newer hardware
> >
> > - compilers etc.
> >
> > - more pain-points would be discovered.
> >
> > 
> >
> > A fundamental problem is that feature changes and bug fixes are
> > usually mingled in upstream.  In some cases, it is a false
> > distinction.  Few developers want to maintain a bunch of old releases.
> > It is very hard for a distro to correctly separate these two, and yet
> > that is required to maintain a stable distro.
>
> I tend to run in debian's equivalent of 'testing'. Found over the last more
> than 10 years that that was a reasonable compromise to stability  and
> currentness.
>
> HTH
> ---
> Post to this mailing list talk@gtalug.org
> Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk



-- 
Giles

Re: [GTALUG] Red Hat Paywall...

[Giles Orr via talk]

I was just reading about it - this is an interesting (depressing) take
from Bradley Kuhn:

https://sfconservancy.org/blog/2023/jun/23/rhel-gpl-analysis/

On Tue, 4 Jul 2023 at 06:40, Evan Leibovitch via talk  wrote:
>
> My take:
>
> I wouldn't go as far as "I saw this coming", but I have long suspected that 
> the IBMification of Red Hat was far from complete -- layoffs, CentOS Stream, 
> now this. RH employees that I know describe a hard shift in corporate 
> culture. And I don't think they're done. It wouldn't surprise me the least if 
> they even change the name of the product to "IBM Red Hat Enterprise Linux" or 
> even just "IBM Linux". Now that they've effectively (and knowingly) destroyed 
> the community goodwill that was formed over more than a decade of Linux Expos 
> and Bob Young roadshows, I don't even see much added value in the RH brand to 
> IBM; the Red Hat we've known for decades no longer exists. Come to think of 
> it, the IBM that helped start LPI and championed Open Source against the SCO 
> and Java assaults of a decade ago is also long gone.
>
> IBM doesn't really give a damn about Alma and Rocky, they're just incidental 
> casualties. The #1 and maybe only target of the subscription-wall action is 
> IBM's longtime arch-enemy Oracle, which may now be forced to actually 
> maintain its own distro and will no longer be able to claim bug-for-bug 
> compatibility with RHEL (or whatever it will be called). They've calculated 
> that the value of the harm this causes Oracle exceeds the lost value of 
> community rejection.
>
> This unfortunate momentum could be stopped (or at least slowed) by a Fedora 
> developer revolt but I don't see that happening.
>
> I see an opportunity for SUSE which maintains both an enterprise-Linux focus 
> and good community relations. Are they up to it? As a longshot maybe even 
> Oracle could try to seize the moment and try a charm offensive to attract a 
> community... but that's unlikely considered its many burned bridges (Solaris, 
> OpenOffice, Java)
>
> Evan Leibovitch, Toronto Canada
> @evanleibovitch / @el56
>
>
> On Tue, Jun 27, 2023 at 11:23 AM D. Hugh Redelmeier via talk 
>  wrote:
>>
>> This Red Hat change concerns me.
>>
>> LONG: Some thoughts on what my "go to" distro pair should be.
>>
>> | From: Alvin Starr via talk 
>>
>> | On 2023-06-27 08:19, Lennart Sorensen via talk wrote:
>>
>> | > Yeah I am happy I switched to debian 25 years ago because Red Hat's
>> | > quality was so poor at the time.  Debian having a better designed
>> | > packaging system was a bonus.
>> | >
>> | Strangely around the same time I switched to RedHat because I got tired of
>> | having to apply my own security patches to the kernel and applications 
>> because
>> | the distribution was shipping with largely unmodified sources.
>> | Like many things in life "your mileage may vary".
>>
>> The RHEL / CentOS / clones drama is certainly unsettling.
>>
>> I don't think that Fedora is directly affected but it is hard to judge
>> whether there will be secondary effects.
>>
>> One upcoming GTALUG talk will be from a Rocky Linux guy.  That should
>> be interesting.
>>
>> I've already been struggling with where I want to go for a stable
>> system.  Besides the drama, I just don't think that RHEL's pinning
>> versions for 5 years is a good strategy.  Backporting for that long
>> feels like a wasted effort, prone to errors.
>>
>> Why do I care about the effort RH puts into backporting?
>>
>> - it creates a RH "moat": it prevents others from competing with them.
>>   Rocky, Alma, Oracle feel like clones, not creative competitors.
>>   That may be unfair to Oracle but that's not a company that I want
>>   a relationship with.
>>
>> - it is labour that feels wasted.  Perhaps that labour could be used
>>   for more constructive purposes.
>>
>> On the other hand RH has added a lot to the community and does do a
>> good job of beating back bugs.
>>
>> I do think that I need a pair of distros: one that is up to date, and
>> one that is low-drama.  If they are in the same family, that cuts down
>> on redundant learning on my part.
>>
>> - CentOS + Fedora has been a good pair for me.  TBH, CentOS has left
>>   me with technical debt: I get stuck on obsolete versions because the
>>   upgrade paths have been disrupted (twice!).  Fedora release updates
>>   have been good for some years.
>>
>>   For my workstation (desktop and laptop) use, I've been very happy
>>   with Fedora, but it sure has a firehose of updates.  I don't think
>>   that it is affected directly by any of this.  But if a lot of people
>>   migrate away from Red Hat stuff, it won't likely be good for Fedora.
>>
>>   It feels as if RH steers the future of Linux by making so many
>>   contributions.
>>
>> - Ubuntu LTS + fresh Ubuntu has been pretty good.  I've had more
>>   problems with package updates on Ubuntu than of Fedora, but it has
>>   been pretty good.  Distro version upgrades have been good but not
>>   perfect in 

Re: [GTALUG] Debian 12

[Giles Orr via talk]

On Tue, 20 Jun 2023 at 23:04, D. Hugh Redelmeier via talk
 wrote:
>
> | From: Giles Orr via talk 
>
> | Long release cycles are a real mixed blessing ...  
>
> Thanks for your note on debian 12 / bookworm.
>
> I'm personally interested in debian as a replacement for CentOS.
> (GTALUG is  going to have a speaker from Rocky Linux in the next few
> months.)
>
> I'm not enculturated in the debian world, but my impression is:
>
> - debian stable is about the same as RHEL.  Very stable, very old.
>   Suitable for those who value stability.
>
> - debian testing is pretty reliable.  Perfectly fine on ones desktop.

I used to use testing on my desktops, up until ... five years ago?  I
found it had become less stable, less reliable at that point, and my
appetite for new software wasn't as great anymore.  I haven't tested
recently.

> - debian unstable is more of an adventure

If you wait a month between updates, unstable will routinely download
over a Gigabyte of updated packages.  I found it unusable on the
desktop.

> Ideologically, isn't FF ESR a match for debian stable?
>
> If you want firefox, isn't that an indication that you are a candidate
> for "testing".

Stability is great, and I mostly agree.  But this (and their ancient
version of Strapi) are the only two places where that dedication to
old software has caused me personal grief.  Of course, I'm using my
own needs as the measuring scale ... it's the one I care about.  :-)

> I don't like snaps / flatpacks much.  For reasons that we don't need
> to go over.  But your situation might be a great use: you want a
> stable OS but need very select exceptions.
>
> ==
>
> We (GTALUG) run a debian stretch server that has fallen out of support.
> It falls on me (among others) to kick it forward.
> I was under the impression that the automated updating process is more
> recent then that.
>
> Is there a royal road to bookworm from stretch?
>
> My guess is that it gets complicated by out-of-distro things that we
> have installed.

The big question is, is this a cloud VM or real iron?  Because my
response would be to spin up another VM and re-install everything on
Bookworm to see if it worked.  This also gives you time because you
can keep the old machine running in parallel.  Time you don't have
when you're working on your own box because you want that back up now
...

https://en.wikipedia.org/wiki/Debian_version_history

Stretch is apparently still supported, and will be for a while.  But I
would encourage you to upgrade ASAP.  In place if you have to (as
Znoteer says, don't skip versions), separate VM if you can.  And yes,
out-of-distro stuff is likely to break because of the changing set of
libraries and compiler builds in the OS around it.  I have to rebuild
my Python venv every time I re-install Debian - happily only for one
project, it would be nasty for multiple projects.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Debian 12

[Giles Orr via talk]

On Sun, 18 Jun 2023 at 16:16, Colin McGregor via talk  wrote:
>
> I assume I'm not the only Debian GNU/Linux fan here. So, just of note
> Debian 12 (code name: bookworm (for the non-Debian fans let me note
> that major Debian releases get code names taken from characters in the
> "Toy Story" series of movies)) was released earlier this month. I have
> upgraded my main desktop and laptop computers from Debian 11 (code
> name : bullseye) to Debian 12 without any significant issues.
>
> A review of Debian 12 can be seen here :
>
> https://youtu.be/YRaJa-VjLGE

I've upgraded something like six of about 10 personal (Debian 11)
machines.  The upgrade process is the easiest and smoothest that
Debian's managed yet.  I haven't tried a new install yet, but if you
have a Debian 11 system, my experience so far suggests that the
upgrade process will go smoothly.

I was annoyed to find that this doesn't - exactly - bump the Firefox
version.  You remain trapped in the ESR version, and even though it's
a newer ESR release, it's still FF v102 which Slack will be disabling
in September.  We use Slack heavily at work - I could survive without
it running in the browser on my Linux machines, but I'd much rather
not.  Further research yielded the suggestion that version 114 will
become ESR in August ... I hope Debian will let that out the gate
before Slack's September deadline, but I wonder if they will.  They
don't like big version changes in the middle of a release.  I guess
I'll be peering into the backports repository if that's what happens
...  (don't suggest flatpak or snaps, thanks - I avoid those when
possible).

The only other thing I was really concerned about with Debian's
versioning was Strapi, another work thing.  Debian 11 had the ancient
version 12 which the developers at work refused to work with.  As
their systems administrator, that caused me major headaches.  Debian
has now jumped to version 18 of Strapi.

Long release cycles are a real mixed blessing ...  

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Toshiba Satellite L500 rejects Linux

[Giles Orr via talk]

On Mon, 22 May 2023 at 21:48, Lennart Sorensen
 wrote:
>
> On Mon, May 22, 2023 at 08:34:34AM -0400, Giles Orr via talk wrote:
> > I've recently acquired (through a friend who stopped using it) a
> > Toshiba Satellite L500 - Core i3 (3rd gen?), 4G RAM.  I'm determined
> > to get Linux onto it (preferably Debian).  I thought I had succeeded:
> > I booted from a Debian USB stick, installed to the HD.  All appeared
> > to go well, but the system won't boot.  It returns to the Boot Menu
> > and says "HD has failed."  What the search engines are telling me
> > is that with this generation of Toshibas, the problem is generally
> > Secure Boot / CSM etc.  Which makes sense, but ... there is absolutely
> > zero mention in the BIOS/UEFI ("Phoenix SecureCore Tiano Setup") of
> > "Secure Boot," "CSM," "Legacy," or "UEFI."  Acccording to notes I
> > found online, "SecureCore Tiano" has "full support" for legacy
> > booting.
> >
> > Another issue with this machine is my mixed success booting from USB
> > sticks: I have an old-ish USB stick I built myself that has GRUB and a
> > large menu of ISOs: works great on most systems, won't boot on this
> > thing - probably because it's an old-style BIOS-boot only(?).
> >
> > One of my ideas was to upgrade the BIOS: it appears there's a newer
> > version available, but it's NOT available from Toshiba, which is the
> > only place I'd want to download it from.  The rest look like dubious
> > secondary download sites (if you know one you consider reliable, let
> > me know).
> >
> > What I read online said that Fedora's installer puts an EFI partition
> > on the HD as part of the install, while Debian doesn't.  And that
> > may(?) be why I can't boot from my Debian install?  So ... I
> > downloaded the Fedora installer, put it on a USB stick ... and no joy:
> > the Toshiba doesn't recognize the Fedora USB stick as a bootable item.
> > Would this be because I burned it on a "Legacy" system?  Is there a
> > fix for that?  Except ... I'm about 99% sure the Debian Installer USB
> > stick was created on the same machine.
> >
> > Worst case, I can stick the HD from the Toshiba into another machine,
> > install Fedora on it, repartition to make room for Debian, put the HD
> > back into the Toshiba ... but that's getting damn complicated and
> > annoying.
> >
> > As always - any suggestions welcomed.
>
> Debian can definitely be installed in EFI mode, but you must boot the
> installer in EFI mode to do it, not legacy mode.  Usually on UEFI systems
> the boot meny gives you a choice of booting in legacy or UEFI mode.
>
> Of course if the system is set to legacy mode instead and you install
> in UEFI mode, then when it goes to boot later you will get a boot error
> (I think something like BBS HD error (BBS being Bios Boot Specification
> apparently)).
>
> It does appear those machines are a disaster and hence the unofficial
> BIOS versions out there trying to fix the complete disaster toshiba sold.

The problem is solved.  I think it's worth reporting here in the
manner of Hugh's "War Stories" because it was so weird.  Although it's
probably an edge case that others are unlikely to encounter.

As far as I can determine, this system will only boot from a USB stick
if the USB stick is willing to boot in "Legacy" mode - and even then
maybe only if that stick has an EFI folder.  I'm not kidding: it
wouldn't boot from a Fedora 38 installer stick, it wouldn't boot from
my old multiboot stick (pure legacy, mostly used for Knoppix), and it
would only boot from the Debian 11 installer stick if the BIOS was set
for "Legacy" even though the Debian stick should work fine in either
EFI or Legacy mode.  But because the Debian installer booted in Legacy
mode, my initial install (see above) wouldn't boot because it
installed Legacy ... and this is an EFI-only system (sort of).

Yesterday I booted from the Debian 11 installer and used the
"Advanced" install (which I hate - so very many steps, with the added
bonus of multiple opportunities for foot-gun).  I strayed from the
beaten path three times: once to say "yes, I'd like 'https://' in
/etc/apt/sources.list", once to create an encrypted VG, and the final
time - the reason I used "Advanced" - to say "force EFI install."
That option didn't show up at the point I expected it to.  I thought
it would be with the HD formatting steps, but instead it comes much
later, around the GRUB installation.  (In hindsight this makes sense,
but it wasn't what I expected.)  The install process then failed on
further package installat

[GTALUG] Anybody using rclone?

[Giles Orr via talk]

I saw a tech news article about a cloud storage provider reducing
their rates ( 
https://techcrunch.com/2023/06/02/dropbox-like-cloud-storage-service-shadow-drive-lowers-its-price/
) and this reminded me that I've been thinking about using cloud
storage as backup ... so long as it's encrypted with a key that's on
my side (only).  A quick search pointed me to "rclone" (
https://rclone.org/ ) which sounds outstanding if their self-promotion
is to be believed.  The fact that rclone is included as a standard
package in Debian stable goes a long way to convincing me.
Ironically, "Shadow Drive" isn't one of the providers that rclone
lists themselves as working with.

This seems like a good starting guide (and makes rclone look fairly
straight-forward).  I haven't used it so I'm uncertain of its
accuracy:
https://www.linuxuprising.com/2020/05/how-to-encrypt-cloud-storage-files-with.html

My current backup systems live and die by 'rsync', so I'm quite
familiar with a program rclone seems to be partially based on.  I've
been trying/hoping to move to 'rsnapshot', although it's kind of a
PITA (but good).  I use Fedora occasionally, but mostly Debian.

This has left me with so many questions, to which I would happily take
any and all answers:
- is rclone good?
- is rclone easy to use?
- does rclone handle encryption of remotes (mostly) transparently?
- in particular, is mounting remote encrypted cloud drives as local
drives fairly easy?
- what cloud storage providers have you used rclone with?
- do you recommend a particular cloud storage provider?  Why?
- do you disrecommend a particular cloud storage provider?  Why?

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Toshiba Satellite L500 rejects Linux

[Giles Orr via talk]

Hi Don.

Probably a good suggestion, but I don't think it will work for me: the
Toshiba in question does have an optical drive, but even if I can find
a CD burner, I'm not sure I have media I can burn to anymore (I have a
stack of blank CDs ... but they're 15+ years old).

On Mon, 22 May 2023 at 08:38, Don Tai  wrote:
>
> When I find an old computer that won't boot USB I go back to a 32 bit CD 
> install, then upgrade. Some old PCs simply won't reliably boot with USB.
>
> On Mon, 22 May 2023 at 08:34, Giles Orr via talk  wrote:
>>
>> I've recently acquired (through a friend who stopped using it) a
>> Toshiba Satellite L500 - Core i3 (3rd gen?), 4G RAM.  I'm determined
>> to get Linux onto it (preferably Debian).  I thought I had succeeded:
>> I booted from a Debian USB stick, installed to the HD.  All appeared
>> to go well, but the system won't boot.  It returns to the Boot Menu
>> and says "HD has failed."  What the search engines are telling me
>> is that with this generation of Toshibas, the problem is generally
>> Secure Boot / CSM etc.  Which makes sense, but ... there is absolutely
>> zero mention in the BIOS/UEFI ("Phoenix SecureCore Tiano Setup") of
>> "Secure Boot," "CSM," "Legacy," or "UEFI."  Acccording to notes I
>> found online, "SecureCore Tiano" has "full support" for legacy
>> booting.
>>
>> Another issue with this machine is my mixed success booting from USB
>> sticks: I have an old-ish USB stick I built myself that has GRUB and a
>> large menu of ISOs: works great on most systems, won't boot on this
>> thing - probably because it's an old-style BIOS-boot only(?).
>>
>> One of my ideas was to upgrade the BIOS: it appears there's a newer
>> version available, but it's NOT available from Toshiba, which is the
>> only place I'd want to download it from.  The rest look like dubious
>> secondary download sites (if you know one you consider reliable, let
>> me know).
>>
>> What I read online said that Fedora's installer puts an EFI partition
>> on the HD as part of the install, while Debian doesn't.  And that
>> may(?) be why I can't boot from my Debian install?  So ... I
>> downloaded the Fedora installer, put it on a USB stick ... and no joy:
>> the Toshiba doesn't recognize the Fedora USB stick as a bootable item.
>> Would this be because I burned it on a "Legacy" system?  Is there a
>> fix for that?  Except ... I'm about 99% sure the Debian Installer USB
>> stick was created on the same machine.
>>
>> Worst case, I can stick the HD from the Toshiba into another machine,
>> install Fedora on it, repartition to make room for Debian, put the HD
>> back into the Toshiba ... but that's getting damn complicated and
>> annoying.
>>
>> As always - any suggestions welcomed.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Toshiba Satellite L500 rejects Linux

[Giles Orr via talk]

I've recently acquired (through a friend who stopped using it) a
Toshiba Satellite L500 - Core i3 (3rd gen?), 4G RAM.  I'm determined
to get Linux onto it (preferably Debian).  I thought I had succeeded:
I booted from a Debian USB stick, installed to the HD.  All appeared
to go well, but the system won't boot.  It returns to the Boot Menu
and says "HD has failed."  What the search engines are telling me
is that with this generation of Toshibas, the problem is generally
Secure Boot / CSM etc.  Which makes sense, but ... there is absolutely
zero mention in the BIOS/UEFI ("Phoenix SecureCore Tiano Setup") of
"Secure Boot," "CSM," "Legacy," or "UEFI."  Acccording to notes I
found online, "SecureCore Tiano" has "full support" for legacy
booting.

Another issue with this machine is my mixed success booting from USB
sticks: I have an old-ish USB stick I built myself that has GRUB and a
large menu of ISOs: works great on most systems, won't boot on this
thing - probably because it's an old-style BIOS-boot only(?).

One of my ideas was to upgrade the BIOS: it appears there's a newer
version available, but it's NOT available from Toshiba, which is the
only place I'd want to download it from.  The rest look like dubious
secondary download sites (if you know one you consider reliable, let
me know).

What I read online said that Fedora's installer puts an EFI partition
on the HD as part of the install, while Debian doesn't.  And that
may(?) be why I can't boot from my Debian install?  So ... I
downloaded the Fedora installer, put it on a USB stick ... and no joy:
the Toshiba doesn't recognize the Fedora USB stick as a bootable item.
Would this be because I burned it on a "Legacy" system?  Is there a
fix for that?  Except ... I'm about 99% sure the Debian Installer USB
stick was created on the same machine.

Worst case, I can stick the HD from the Toshiba into another machine,
install Fedora on it, repartition to make room for Debian, put the HD
back into the Toshiba ... but that's getting damn complicated and
annoying.

As always - any suggestions welcomed.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Checking TLS certificates across multiple machines

[Giles Orr via talk]

Hi Alex.

Thanks!  That's what every OSS author hopes for.  It's great to hear.

On Thu, 27 Apr 2023 at 09:35, Alex Kink via talk  wrote:
>
> Hi Giles.
> I find your script very useful and have added it to my list of tools.
>
> Thank you!
>
> > On Apr 26, 2023, at 08:57, Giles Orr via talk  wrote:
> >
> > Three weeks ago I mentioned to a couple people on this mailing list
> > that I'd written a script at work that live-checks TLS certificates
> > across multiple machines.  This is a very useful thing to have when
> > one of my jobs is maintaining multiple small websites.  One of these
> > people responded that I should give a short talk about this project,
> > and added "Most of us do a little sys-admining and can learn from the
> > pros."  At which point I burst out laughing ... and realized Imposter
> > Syndrome had reared its ugly head once again.  A "professional" is
> > someone who gets paid to do a job, and I am that.  Doesn't mean I feel
> > like a "professional sys-admin," even though I am - more of a
> > professional jack-of-all-trades ...  But "Imposter Syndrome" is
> > another huge discussion ... we now return you to our irregularly
> > scheduled programming.
> >
> > That message prompted me to talk to my bosses, one of whom not only
> > endorsed my idea of publicly releasing this code, but did so
> > enthusiastically.  My thanks to Toronto Public Library for that.  And
> > so I give you:
> >
> > https://github.com/gilesorr/robotface-utils
> >
> > (The naming is a weird story - explanation in the README if you're curious.)
> >
> > Given a list of fully qualified domain names, the 'chkcertexpiry'
> > script goes to each one, retrieves the TLS certificate, and lists the
> > expiry date and issuer.  If the expiry date is soon (number of days
> > settable by a command line switch) the date is colour-highlighted.  If
> > there are connection problems, those are noted.
> >
> > At work, I run this once a week against all the sites I'm responsible
> > for.  I don't think we've had an expired cert on any of those domains
> > in the 5-7 years I've been using this script.
> >
> > I hope people will use this.  I welcome comments and suggestions.  I'm
> > a little scared of PRs because I've never had to deal with them before
> > ...
> >
> > Please feel free to share as you think appropriate.
> >
> > --
> > Giles
> > https://www.gilesorr.com/
> > giles...@gmail.com
> > ---
> > Post to this mailing list talk@gtalug.org
> > Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
>
> ---
> Post to this mailing list talk@gtalug.org
> Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk



-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Checking TLS certificates across multiple machines

[Giles Orr via talk]

Three weeks ago I mentioned to a couple people on this mailing list
that I'd written a script at work that live-checks TLS certificates
across multiple machines.  This is a very useful thing to have when
one of my jobs is maintaining multiple small websites.  One of these
people responded that I should give a short talk about this project,
and added "Most of us do a little sys-admining and can learn from the
pros."  At which point I burst out laughing ... and realized Imposter
Syndrome had reared its ugly head once again.  A "professional" is
someone who gets paid to do a job, and I am that.  Doesn't mean I feel
like a "professional sys-admin," even though I am - more of a
professional jack-of-all-trades ...  But "Imposter Syndrome" is
another huge discussion ... we now return you to our irregularly
scheduled programming.

That message prompted me to talk to my bosses, one of whom not only
endorsed my idea of publicly releasing this code, but did so
enthusiastically.  My thanks to Toronto Public Library for that.  And
so I give you:

https://github.com/gilesorr/robotface-utils

(The naming is a weird story - explanation in the README if you're curious.)

Given a list of fully qualified domain names, the 'chkcertexpiry'
script goes to each one, retrieves the TLS certificate, and lists the
expiry date and issuer.  If the expiry date is soon (number of days
settable by a command line switch) the date is colour-highlighted.  If
there are connection problems, those are noted.

At work, I run this once a week against all the sites I'm responsible
for.  I don't think we've had an expired cert on any of those domains
in the 5-7 years I've been using this script.

I hope people will use this.  I welcome comments and suggestions.  I'm
a little scared of PRs because I've never had to deal with them before
...

Please feel free to share as you think appropriate.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Canadian hosting?

[Giles Orr via talk]

On Mon, 10 Apr 2023 at 15:16, Alex Kink via talk  wrote:
>
> Digital Ocean, Vultr, OVHcloud have data centres in Toronto or Montreal.
>
> > On Apr 10, 2023, at 15:14, Stewart C. Russell via talk  
> > wrote:
> >
> > I know that Akamai Cloud (formerly Linode) has dedicated root Linux servers 
> > in Toronto, but is there anyone else non-terrible*? There's a potential 
> > client who absolutely must have all data and processing hosted in Canada. I 
> > realize there's probably a 3x (or more) premium for in-country hosting.
> >
> > Any suggestions/caveats greatly appreciated.

I second the Digital Ocean suggestion: I use them heavily both at work
and personally (gilesorr.com mentioned in my email signature is hosted
there), with all machines in their Toronto data centre.  We've used
them for approximately seven years (the Toronto presence is only about
five years old), and they've been excellent.  I think we should point
out one caveat: D.O. is American-owned.  Although if you were
considering Akamai/Linode, I guess that's okay.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] war story: gtalug.org's filled up

[Giles Orr via talk]

On Tue, 4 Apr 2023 at 17:50, D. Hugh Redelmeier via talk
 wrote:
>
> (This post is also a test of whether the mailing list is working.)
>
> GTALUG's server's filesystem filled up: little disk space left.
>
> I discovered this when I tried to do "apt update; apt full-upgrade".
> The second step failed, saying that /var/cache/apt/archives had no
> room.  This means that / has no room because / contains that
> directory.
>
> Tip: "df /some/path" will tell you how much space is used on the
> filesystem containing /some/path and it will tell you the mount point
> of that file system.
>
> I wandered around the filesystem, doing:
> sudo df -s * | sort -n
>
> This command lists things in the current directory, and their sizes,
> largest last.  (It skips things with names starting with ".".)
>
> Pretty soon, I found that most of the space was taken up by
> /var/cache/apt/archives after all.  This kind of surprised me.
>
> Googling got me to others with this problem.  The advice:
> sudo apt autoclean
> That gave back (only) 3% of the disk.
>
> That directory was still way too big.  Most of the space was taken by
> 35 versions of gitlab-runner.  I have no idea why we need multiple
> versions.
>
> Violence is sometimes the answer.
> sudo apt clean
>
> That left only 32% of / used.
>
> Note /var/cache/apt/archives is only a cache.  If the system wants any
> of these, it should be able to find them in a repo.

I used to use a method very similar to yours - I presume you meant
'du' not 'df'.  I used 'du -sh * | sort -h' which I found easier to
read and understand.

But then - "Hallelujah" (let's make this a religious discovery) I
found 'ncdu', which is a TUI that does exactly the same thing, but
makes it explorable by selecting and entering any directory to see the
same listing inside that new dir.  Give it a try: it makes the whole
process of finding space-hogging directories immensely faster.  And
because it's a TUI, not a GUI, it works on remote hosts over SSH even
if they don't have X/Wayland.  I don't know how long it's been in the
Debian repos: it's in the current Debian, it may not be in older ones.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Not all kernel fixes backported to CentOS

[Giles Orr via talk]

I think we have a few CentOS users on this list, so this may be a bit alarming:

https://www.neowin.net/news/google-discloses-centos-linux-kernel-vulnerabilities-following-failure-to-issue-timely-fixes/

Here are a couple of the important chunks:

"As detailed in the technical document here, Google Project Zero's
security researcher Jann Horn learned that kernel fixes made to stable
trees are not backported to many enterprise versions of Linux. To
validate this hypothesis, Horn compared the CentOS Stream 9 kernel to
the stable linux-5.15.y stable tree. For those unaware, CentOS is a
Linux distro closest to Red Hat Enterprise Linux (RHEL) and its
version 9 is based on the linux-5.14 release."

"As expected, it turned out that several kernel fixes have not been
made deployed in older, but supported versions of CentOS Stream/RHEL.
Horn further noted that for this case, Project Zero is giving a 90-day
deadline to release a fix, but in the future, it may allot even
stricter deadlines for missing backports ..."


-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Mail server Problems?: Re: Tonight's meeting announcement?

[Giles Orr via talk]

I received multiple messages yesterday from the list, many of which
appear to be up to a couple months old.  The one that follows is a
fine example: Colin and Stewart had a conversation in September, which
for some reason was released to me 2022-11-13-11:11.  Along with about
another ten threads, not included here.  This is very weird behaviour,
and makes it hard for me (and anyone else seeing this behaviour) to
participate in the conversations.  And yet I've seen other emails to
the list in the two months in between.  Very strange indeed.

On Sun, 13 Nov 2022 at 11:11, Colin McGregor via talk  wrote:
>
> On Thu, Sep 15, 2022 at 6:48 PM Stewart C. Russell via talk  
> wrote:
>>
>> On 2022-09-14 23:33, Colin McGregor wrote:
>> >
>> > That was Tuesday evening. Any event, it wasn't a very good talk (for
>> > several reasons), so you didn't miss much.
>>
>> I'm sure it was great, Colin. I'm sorry I wasn't able to attend. I
>> originally sent the message Tuesday morning, but it bounced around for a
>> day with a "FAILED_PRECONDITION: connect error (111): Connection
>> refused" issue.
>>
>> On the subject of boating and Linux, I did just see this:
>>
>> RCgmbh/PiLot: dotnet core based projects for the PiLot navigation and
>> logbook program for raspberry pi — https://github.com/RCgmbh/PiLot/
>
>
> Interesting, I will have to check out how this compares to "AvNav" another 
> web based navigation program that can run on a Raspberry Pi (and is one of 
> the MANY packages included with Bareboat Necessities ( 
> https://bareboat-necessities.github.io/ ) (I did mention Bareboat Necessities 
> in my talk, but not AvNav)).
>
> Thanks.
>
>
> Colin.
>
>>
>> via the Raspberry Pi Forum:
>> https://forums.raspberrypi.com/viewtopic.php?t=340217
>>
>>   Stewart
>> ---
>> Post to this mailing list talk@gtalug.org
>> Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk
>
> ---
> Post to this mailing list talk@gtalug.org
> Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk



-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] wired headset suggestions?

[Giles Orr via talk]

Hi Kare.

On Mon, 10 Oct 2022 at 02:27, Karen Lewellen via talk  wrote:
>
> Hi folks and happy thanksgiving if you are celebrating.
> Given others might spend far more time in  on line meetings than myself,
> plus the ideas here tend to be workable, I thought I would toss this
> question out here.
> There are some firm details for me, rather a few  strictly digital, or
> wireless headphones actually can make me dizzy.  part of my auditory
> processing  condition.
> so, seeking wired headset ideas,  with a boom mic, think for  a laptop,
> just plug in no
> pairing required.  I understand plantronics now Poly?, are a fine brand,
> but realize there may be others.
> Even if older, especially since my search is not producing options in a
> safe sensitivity range, either too low, as in less than 100, or too high,
> well above 110.
> I prefer impedance in the 32 range as well.
> Any model ideas?

These may not suit you for a number of possible reasons - not least of
which is that they plug into a USB port rather than the headphone
port, and I don't fully understand your requirements.  But take a look
if that's not a deal-breaker:

https://www.amazon.ca/gp/product/B000UXZQ42/ - Logitech H390

I mention them because I bought them recently and I've been very happy
with them.  I've had to do a lot of audio chats at work and was tired
of using the microphone in the computer, so I found these.  I chose
them because they're well reviewed and surprisingly cheap (currently
$36, although it fluctuates).  Some comments from personal experience:

- they "just worked" when plugged into a Mac, and also Linux (Fedora).
- the volume control on the chord works well on Mac - not sure if it
works on Linux although I think it does
- the ear pads sit ON you ears, rather than AROUND them: I find they
start to get a bit uncomfortable around the hour mark (which isn't too
bad, considering)
- given the price, the sound quality is quite good - everyone I've
asked on the other end of my calls says I sound better
- on pretty much every OS, plugging in a headset audio jack
automatically over-rides the built-in speakers ... this wasn't
initially true with the USB headset for either Mac or Linux, although
after some futzing both now do so (just saying: setup may be
problematic)

I hope this helps.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Data archaeology: archive KDE distortion from 2004ish

[Giles Orr via talk]

On Wed, 14 Sept 2022 at 23:35, Jamon Camisso via talk  wrote:
>
> On 2022-09-14 6:39 p.m., D. Hugh Redelmeier via talk wrote:
> > | From: Stewart Russell via talk 
> >
> > | Any suggestions for which ancient distro and where it might still live 
> > gratefully received.
> >
> > Knoppix?  Perhaps 3.5, which had a DVD version?  There must be copies
> > around.  See this for a list of versions:
> > 
> > Not sure where to find it now.  Perhaps
> > 
> >
> > This would surely require BIOS/MBR for booting.  And new fangled devices
> > won't be understood.
> >
> > There are many other possible distros.
>
> Kubuntu 6.06 might ship with DigiKam pre-installed?
>
> https://old-releases.ubuntu.com/releases/kubuntu/6.06/
>
> Otherwise, it should be available in the packages archive.
>
> I'd boot that ISO in a virtual machine, mount the database files into it
> as a shared folder, and then try starting the program.
>
> Good luck!

Hi Stewart.

Just back from vacation, explaining this delayed response (although
it's not really a helpful one).  This seems like a solvable problem,
and the suggested approaches from Hugh and Jamon are exactly where I
would have gone.  Note that I said "seems like ..."  I'd love to hear
how this turned out, if you're willing to share.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] ThinkPad T60s available?

[Giles Orr via talk]

Apologies: this is not directly Linux-related.  Think of it as
greening the world by giving new life to old hardware ...  And I know
a lot of people on this list are ThinkPad users.

I'm looking for a functional ThinkPad T60 (or two) preferably with a
copy of Windows XP installed.  XP isn't required - I think I can
handle that part if needed.  This is for a friend who needs to run an
exceptionally old piece of music software - and for this application,
real hardware makes more sense than emulation.  The computer won't be
connected to the Internet.

Similar machines (ie. approximately that generation) would be
considered, but we know the T60 works well in this context so it's
preferred.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Linux on Chromebook

[Giles Orr via talk]

I used to go to Detroit a lot with geeky friends, and we pretty much
always visited Microcenter.  Although I haven't gone in years.
Stewart's message suggests there isn't stock of this particular item
as close as Detroit, but ... for most people on this list, I think
you'd enjoy a visit to Microcenter if you get the chance.  Fun place -
imagine, if you will, a bigger and even geekier Canada Computers, with
the added bonus of (mostly) cheaper American prices.

On Fri, 9 Sept 2022 at 19:53, Stewart C. Russell via talk
 wrote:
>
> Someone at the Raspberry Pi meetup last night mentioned the Evolve III
> Maestro 11.6" educational laptop, if anyone's looking for a *really*
> cheap laptop. I don't know of anyone selling it in Canada at a
> reasonable price, but Micro Center - a brick & mortar electronics store
> in the USA - is selling them for US $60 (~ $78). Unfortunately, the
> nearest store with stock is likely a day and half's drive away, plus
> border crossing and all that faff. User reviews on Reddit are not super
> positive.
>
> The machine is incredibly barebones, but the fact it even works at all
> is quite impressive at this price:
> https://www.microcenter.com/product/649971/evolve-iii-maestro-116-laptop-computer-dark-grey
>
> This was in response to me introducing the frankly overpriced pi-top [4]
> Raspberry Pi 4-based "tablet": https://www.pi-top.com/products/pi-top-4
> It might have a glorious touchscreen, but a tablet it ain't.
>
>   Stewart
> ---
> Post to this mailing list talk@gtalug.org
> Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk



-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Ubuntu Studio ver 22.04 Desktop (KDE) does not seem to allow for images spanning multiple screens

[Giles Orr via talk]

> --- Original Message ---
> On Thursday, August 18th, 2022 at 5:33 AM, sciguy via talk  
> wrote:
>
>
> >
> >
> > Hi
> >
> > For the record I am using Ubuntu Studio 22.04 with a KDE desktop (Plasma
> > version 3.0).
> >
> > Unless I am missing something major, I don't see a way that KDE Plasma
> > allows for expansion of a desktop graphic across two screens. Graphic
> > dimensions I have in mind are something like: 3840 x 1080. When I load
> > the graphic for the desktop in the settings, I just get the graphic on
> > one screen. If I select the jpg from the file manager (Dolphin) to set
> > as wallpaper, I get duplicate graphics on each monitor, rather than the
> > same graphic spanning both monitors, which is what I want.
> >
> > If anyone can be of help I would appreciate it.

On Thu, 18 Aug 2022 at 11:27, gs via talk  wrote:
>
> Unfortunately there isn't a way to do this built into kde.
>
> There is a 3rd party solution at:
> https://github.com/hhannine/superpaper

Probably not what you were looking for, but maybe better than
installing an unknown third party piece of software: you can probably
use a photo editor to slice your image into two pieces and use half as
the background on one screen and half as the background on the other
screen.  (I'm not on a KDE machine right now - can't confirm it can do
this, but I'd assume it can display different backgrounds on different
monitors.)

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Making RPM from directory tree?

[Giles Orr via talk]

On Sat, 11 Jun 2022 at 18:09, William Park via talk  wrote:
>
> To RPM guys,
>
> I have a directory tree, created from "make install DESTDIR=...".  Now,
> I want to make RPM package from the tree.  Online docs are too verbose
> and confusing.
> 1. Is there GUI app with appropriate fill in boxes?
> 2. Is there command line utility?  In Slackware, it's just "makepkg".
>
> Background:
>
> I use "netpipes-4.2" a lot.  Debian and Slackware have it, but Redhat
> variants don't.  I can use netcat and pipe (and I do), but there are so
> many variations (ie. nc, netcat, ncat), each with different behaviour
> and options.  I like to use what I'm familiar with.
>
> So, I downloaded the source, compiled it, and installed to a directory.
> Now what?

I used to use 'checkinstall' a lot.  Whenever I did the "./configure ;
make ; make install" dance, I used 'checkinstall' on my predominantly
Debian systems.  It tracks the created files, and creates a system
package.  And if I didn't like the product, I could then 'apt purge X'
to remove the package I'd made.  I rarely used it to move the package
to another system, but that usually worked if the other system was the
same architecture and same OS.  I think 'checkinstall' works on Fedora
(it claims to make RPMs), but it doesn't seem to be packaged for it.

http://checkinstall.izto.org/

Last bug fix 2016, spectacularly retro website.  But it's still
packaged in Debian.  Not sure this is a path you want to pursue, but
it's a program I found very useful for a whole lot of years.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Removing snapd from Ubuntu

[Giles Orr via talk]

On Wed, 11 May 2022 at 14:54, Lennart Sorensen via talk  wrote:
>
> On Wed, May 11, 2022 at 09:30:41AM -0400, Stewart Russell via talk wrote:
> > I just did the 22.04 upgrade thing, and it seems that Firefox will be held
> > at v 99 if you don't have snapd. So beware of old/held packages as you
> > update.
> >
> > Another delightful thing I found is that Ubuntu took its very own special
> > path in the "Sensible things to do in the Python 2 / Python 3" debacle:
> > remove Python 2, but don't link python3 to python. Move fast and break
> > stuff is very tiring when you're constantly getting beaten up like this.
>
> Upstream python says not to call python3 as python.  So Unbuntu did
> it correctly.
>
> Python 3 code calls /usr/bin/python3 (or probably better yet /usr/bin/env
> python3) and legacy python 2 code calls /usr/bin/python.

That actually makes sense.  Then the next one can be 'python4' without
causing problems.  But many distros - and many system administrators
will probably just make it 'python'.  

Yup.  Here's Fedora 35:

$ which python
/usr/bin/python
$ python --version
Python 3.10.4

Debian is guilty of the same thing - which is interesting, because
Ubuntu is based on Debian and would have had to take a detour to "do
the right thing."

Debian and Fedora both also have /usr/bin/python3.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Ethernet alias triggering network security warnings

[Giles Orr via talk]

On Wed, 27 Apr 2022 at 22:26, Kevin Cozens via talk  wrote:
> I am using an embedded system that is running a version of an Ubuntu distro.
> The embedded system is being set up so that the physical Ethernet port will
> have two aliases. One is eth0:0 and the other is eth0:1.
>
> eth0:0 is set to get an IP address via DHCP when a network cable is plugged
> in (assuming a DHCP server is available). The eth0:1 is set to have a static
> IP address. This allows an installer to connect a laptop to the system and
> do some onsite configuration (setting WiFi parameters).
>
> The network settings I'm using are:
>auto eth0:0
>iface eth0:0 inet static
>address 192.168.100.110
>netmask 255.255.255.0
>gateway 192.168.100.1
>
>auto eth0:1
>allow-hotplug eth0:1
>iface eth0:1 inet dhcp
>
> This was working at one point but lately a Windows machine on the network
> will start complaining either about possible ARP poisoning or it will say
> there is a duplicate IP address. The program on the Windows box reporting
> the problem is esset Internet Security. I know nothing about the program. It
> is on the machine of the person preparing boxes to be sent out in to the 
> field.
>
> When I check the system that seems to have triggered the duplicate IP
> address warning I see no duplication. I have checked all the MAC addresses,
> inet v4 and inet v6 IP addresses listed by the ifconfig program.
>
> Does anyone on this list know anything about that security program or why
> the person setting up the embedded systems would be seeing the duplicate IP
> warnings?

Most DHCP servers I've used on the 192.168.N.* tend to be set up to
assign numbers in the 100-254 range.  Choosing "110" as your number
puts it in that range, and unless you explicitly told the DHCP server
that 110 is off bounds, it may assign it to another machine as well.
That would cause collisions, and all kinds of unhappiness.  Notice
that 1-99 aren't usually used by the DHCP server, and so are available
for use as static addresses.  This is of course dependent on the setup
of your DHCP server, you should take a look at its setup.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] From Slackware to which distro?

[Giles Orr via talk]

On Wed, 27 Apr 2022 at 00:37, William Park via talk  wrote:
> I've been running Slackware since forever.  It's time to grow up and see
> the world.  Which distro would you recommend that I move to?  Yes, I
> know it's personal, and reasons will be varied and educational.
>
> - Ubuntu -- OK.  I use it at work in VM and in WSL (Windows Subsystem
> for Linux).  For me, Mint is another flavour, just like Kubuntu,
> Xubuntu, etc.
>
> - Oracle -- I use it at work too.  Was CentOS, but switched to Oracle
> because they said delivering end-of-life OS is bad marketing.
>
> - Fedora -- OK.  Doesn't seem to have its equivalent in Ubuntu side.
>
> - OpenSUSE -- Difficult to pin down.  It uses RPM but in their own way.
> It has rolling release (Tumbleweed) and versioned release (Leap).
>
> - Arch -- no.  I don't need/want to learn what they are trying to teach.
>   I run Slackware, so I already know all that.
>
> Thanks for any feedback.
> William

Something I tend to forget about is the GUI: like Chris, I install
whatever distro, then install OpenBox (which isn't very flexible, but
suits my way of working), drop in my own configuration file, and I'm
ready to go.  Many people are extremely picky about their GUIs: if
that's you, my advice isn't much use on that.  I'm extremely picky
about the setup of my terminal - but that can be easily managed in
almost any Linux distro in existence.

I use Debian most of the time, and Fedora on the machines that Debian
can't handle.  This is generally because Debian can be a bit slow with
the newest hardware drivers.  Both have extensive package repositories
and it's very rare for me to have to step outside those to get things
done.  I don't like Fedora's six month release cycle - and I
particularly don't like that they drop their old releases three months
after a new one comes out.  Debian's slower release cycle means you
can get stuck with older software, but they're great about
back-porting security fixes and "stable" is indeed very stable.  I
also prefer 'apt' to 'dnf' as a package manager, but in practice that
hasn't made a lot of difference (although I got to say ... they could
both learn a lot from the impressive speed of Arch's package manager
'pacman' ...).

I personally avoid Devuan: during SystemD's initial year or two of
existence, it was indeed a horror show.  But now that they've forced
users of multiple distros to find the problems with it, it's been well
debugged.  As Lennart says, it's now better than the old Init system.
(Debian and Fedora both use SystemD.)

I also have to agree that you'll save yourself some grief using
whatever distro(s) you're made to use at work.  Even if you don't love
that distro: using it at home gives you more familiarity, and less
work overall.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Package Manager upgrades on Linux

[Giles Orr via talk]

This is curiosity on my part and not meant to start a massive debate.
I apologize should it do so.

I started using Linux with Slackware.  That wasn't really a "managed
package system" at the time - you just unpacked tarballs.  I moved to
RedHat, where "rpm" was a huge improvement in package management over
Slackware.  But rpm was subject to "dependency hell," which wasn't
much fun.  I switched to Debian, where I was impressed by the superior
"apt-get".  These days I use both Debian and Fedora.  Debian has moved
to "apt" (not terribly different from "apt-get"), while RedHat/Fedora
first came up with "yum" (painfully slow, but eliminated "dependency
hell"), and then moved on to "dnf" which I'd say is almost as fast as
"apt" - even though it's mostly an abstraction layer on top of "rpm"
(as was "yum").

Just recently I started using Arch on one of my machines.  Arch is a
rolling release, and has fairly high package turn-over.  If I leave
the machine alone for a couple weeks and tell it to do a full upgrade
(their package manager is called "pacman"), it usually comes back with
"have to download 650M packages."  My reaction the first couple times
was "Damn, I'd better go get a coffee!"  Only to find that by the time
I stood up and started heading for the door ... it was finished.
"apt" and "dnf" should hide their heads in shame.  Fedora and Debian
could learn a hell of a lot from whatever pacman is doing to fly
through package installations perhaps four times faster than their
package managers.

So my questions (yes, there are questions here): does anyone know of
other package management systems?  I suppose I'm thinking of
OS-specific ones rather than flatpak or snap.  How fast are those
other systems relative to those already mentioned?  And do we have any
full-time Arch users on the list?

Contrary to this discussion, my decision of what Linux distro to
install and use isn't primarily based on the package manager's speed.
Installing Arch in 2021 was a lot like installing Slackware in 1998: I
had to research and type dozens of commands in a shell just to get a
functional system, and at that I'm not entirely sure I did all the
security stuff right - which is more than a little alarming.  But I
think they have the best documentation of any distro in existence
(which made that tricky install possible), and a really excellent
package manager.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Positions at Toronto Public Library

[Giles Orr via talk]

Two positions are being advertised in my immediate department, should
anyone be interested.

https://www.linkedin.com/jobs/view/2926612537/ - ITS Lead, Digital Products

https://www.linkedin.com/jobs/view/2926616569/ - Dev/Ops Analyst

Feel free to contact me if you have questions.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Can anyone recommend a "health" smartwatch?

[Giles Orr via talk]

On Fri, 4 Feb 2022 at 08:32, David Collier-Brown via talk
 wrote:
> My fitbit with a heart-rate sensor died, so I bought a Spade and Co "Health 
> Smartwatch 2", which was absolutely beautiful, just what I wanted... and 
> didn't work.
>
> All three apps they use all demand I own a phone with a lage screen, so 
> you can fill in the forms. I have a small screen, and the app knows I 
> wouldn't want to scroll down, so it prohibits my doing so. (They should 
> terminate their UX person with extreme prejudice (:-))
>
> Any recommendations about good "health" watches?  If folks have had good 
> experiences with some, I'll try their apps, and if they work, order one.

Hi Dave.

My brother recently long-term loaned me his older Withings watch
because he bought the latest by Withings.  Stylistically quite
different from "smart watches," it's mostly meant for health and/or
sports.  It actually looks quite good.  I've been hugely impressed
with the battery life - I charged it when I got it, and a month later
it's at about 50% ... I haven't been using it much (I wear it when I
walk, and not even always then), but compared to "smart watches," it's
still astonishing.  It does steps and heart-rate (although in its
current mode you have to press a button to request heart-rate).  It
probably does a whole lot more I'm not aware of yet.  I wanted it
mostly for heart-rate ... and as a result of my privacy fanaticism, I
haven't even installed their app.  My brother is pretty hard-core on
privacy too: I'd say that his buying their watch(es) at all - and
using their app - is a vote of confidence.  Worth a look at least?

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] RaspberryPi won't automount USB memory stick

[Giles Orr via talk]

On Sun, 16 Jan 2022 at 23:55, Stewart C. Russell via talk
 wrote:
>
> On 2022-01-14 23:24, Kevin Cozens via talk wrote:
> >
> > Why is Raspbian not set up to automount a memory stick
>
> Raspberry Pi OS Lite (no desktop) doesn't automount USB devices.
> Raspberry Pi OS with Desktop does.

As a follow-up to Stewart's message - automounting is, I think, pretty
much exclusively the realm of GUIs.  It's handled by a GNOME package
or a KDE package or the PCManFM package which handles the desktop for
LXQt.  Automounting isn't usually desirable in non-GUI environments -
because of the assumption that they're server-like environments.  I
don't like automount even in a GUI environment - but usually manage to
avoid it by going old-school on my Window Manager.  I still run
OpenBox or LXQt.  The latter _does_ automount stuff, but if you dig
into its config files there's a toggle to turn the behaviour off.

All of which doesn't answer your question - although it does perhaps
offer an indirect solution.  Install a desktop environment (as opposed
to a "window manager") if you don't have one (this may not be
desirable).  You haven't said if you have one - although I'm guessing
that Stewart thought "embedded system" meant "Pi OS Lite" and thus no
automount.  I believe there is a non-GUI, background utility program
somewhere out there that does automounting.  You may have already
found it in the form of the "usbmount" package but perhaps there's a
setting you need to tweak.  Although ... searching the Raspberry Pi OS
package repos, I don't see a package called just "usbmount" so I'm not
sure what that fix might be.

This is very old:

https://unix.stackexchange.com/questions/11472/automount-usb-drives-with-no-gui-requirement-halevt-replacement

but may be worth a look for the udev rules, which probably haven't
changed, and because the second answer has a pointer to Archwiki about
automounting USB.  I consider Archwiki to be pretty much the single
best source of documentation on Linux: that article would be worth a
read.  I hope this helps.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] X10 - Gear you don't want? Or replacement suggestions?

[Giles Orr via talk]

X10 ( https://en.wikipedia.org/wiki/X10_(industry_standard) ) isn't
directly related to Linux, but this seems like the right group to ask
about it ...  I've been happily using X10 transmitters and receivers
to automate switching lights on and off around my place for about two
decades.  These days that requires a USB-to-serial converter cable,
but otherwise works just fine.  X10 has significant limitations
(unencrypted, local interference, very slow, one-way) but works well
for the simple things I want it for.  Which leads me to my two related
but rather different questions.

First Question: Does anyone have any X10 gear they'd like to get rid
of?  I'll happily take it off your hands.  If you'd prefer to sell it
rather than giving it away, we can talk.  I'm particularly looking for
a "Transceiver Module."  I can still buy these from x10.com - but
they're an American company without a Canadian customs broker.
They'll happily ship me stuff ... but whether or not it arrives, how
much I have to pay in duties, and whether or not I have to travel to
Niagara Falls to pick it up ... all my problem.

Second Question: What's a more modern replacement for X10?  Keeping in
mind that I have several limiting caveats.  The first of which is, no
"calling home."  All interactions should happen within the confines of
my computers and my residence.  I would also fairly strongly prefer
not to rely on a local wireless network: I have wireless, but it's
deliberately off most of the time and I prefer it that way (for
misguided security reasons, not misguided health reasons ... I like
wires).  I also don't want to have to do my own soldering (I'm capable
of it, barely, but would prefer plug-and-go).

Thanks.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Man and Info Pages

[Giles Orr via talk]

On Thu, 6 Jan 2022 at 02:28, Howard Gibson via talk  wrote:
>
>I am updating my UNIX Command Line HOWTO.  I have a remark in my text to 
> the effect that man pages contain text stating that man is obsolete, and that 
> you should use the info pages instead.
>
>I actually have not seen this lately.  What is the status of man and info 
> at the moment?

I think the intention (20 years ago?) was that 'info' should replace
'man'.  I can understand, even appreciate, the idea: it allows
linking, multiple documents per command, and is a "richer" format.
Unfortunately, the reality was that the uptake was awful because most
of us (I include myself in "us" here) hated 'info' and didn't use it
because we couldn't remember the multiple awkward commands required to
interact with it.  I open a man page, I scroll - done.  I opened the
man page because I wanted to learn about the command whose man page I
opened - I don't want to have to learn about 'info' before I can learn
about any other command ...  The people behind 'info' stuck to their
guns for a decade or more, but I think the whole thing has pretty much
given up the ghost at this point.

This is based on my own limited observation and opinion, not any
actual research.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Pi in-place upgrade with weird side-effects

[Giles Orr via talk]

I have two Raspberry Pi 4s.  Both had Raspbian 10 (buster) installed
on them initially, and both have now had in-place upgrades to
Raspberry Pi OS 11 (bullseye).  I'm aware that the original and new
OSes aren't quite the same beast (Raspbian vs. Raspberry Pi OS).  For
the most part, this went fine, but has left some unfortunate
artifacts.  The first (that I know of, there may be many others) is
/etc/apt/sources.list.d/raspi.list, which includes one active line:

deb http://archive.raspberrypi.org/debian/ buster main

Notice that this is pointing to buster.  Which made me do some digging:

# for package in $(dpkg-query -W --showformat='${Package}\n'); do if
apt info $package 2> /dev/null | grep -q buster ; then echo $package ;
fi ; done
chromium-browser
chromium-browser-l10n
chromium-codecs-ffmpeg-extra
libraspberrypi-bin libraspberrypi-dev
libraspberrypi-doc
libraspberrypi0
raspberrypi-bootloader
raspberrypi-kernel
rng-tools-debian

These are all packages that are installed from the buster repository,
and they seem like some seriously important packages (the kernel? the
bootloader?) that are being held back at an older version of the OS.

I would rather not re-install the OS, so any suggestions about what's
going on or how to fix it (or if that's necessary) would be
appreciated.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Booting linux from nvme disk?

[Giles Orr via talk]

On Fri, 19 Nov 2021 at 21:09, Peter King via talk  wrote:
>
> Anyone had success with getting linux to boot from an nvme disk?  I've
> been fighting with UEFI and Arch Linux all day now, trying to get a new
> computer to even recognize the nvme disk as a boot device.  (Last time I
> encountered this problem I gave up and installed an ordinary HD to boot
> from.)  For what it's worth, I have disabled Secure Boot, reformatted the
> nvme disk to have a new EFI partition without Windows, a swap partition,
> and a root partition; I'm trying to use efibootmgr (so no loading of a
> further bootloader), and, as far as I can tell, identified all the right
> partitions by device name or PARTUUID.  Still no go.  I even tried adding
> the parameter nvme_load=YES into the "root" part of the efibootmgr, and
> also adding nvme and vmd as modules in mkinitcpio.conf, all without any
> success.  If necessary I'll just punt and install a regular HD to boot
> from, but that rather takes away from having an nvme disk in the first
> place.
>
> (Over the years I've learning to approach installing Linux with fear and
> loathing, with almost all the problems being with the bootloader -- from
> LILO through GRUB and GRUB2 now down to UEFI.)
>
> Any suggestions welcome!  I didn't want to spend my weekend doing this.

Hi Peter.

I'm pretty sure I have Linux installed on an NVMe disk somewhere.  I
apologize that I'm not sure where: most of my machines are "older,"
but not all - however, I have enough of them that I'm not sure where
that NVMe disk is.  The point is: if it's in this house, I installed
Linux on it and it evidently wasn't a problem because I don't remember
it.

I would suggest downloading and installing a recent Fedora.  You've
chosen Arch: that's a hard row to hoe.  I'm not saying you should
change distros, but installing Fedora abstracts away a LOT of the
difficult work Arch insists you do by hand (I've installed both,
recently).  If installing Fedora works, A) you've proven this computer
can be made to work with Linux, and B) the hard drive may be formatted
in a useful way for another attempt at Arch.  Or you could (and I
think this might be wise initially) install Arch as a secondary OS to
Fedora.  With that in mind, you could use Fedora to format two 32G OS
partitions, and an all-the-rest-of-the-drive partition as /home/.

Just a thought.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Command doesn't work in script but works on command line?

[Giles Orr via talk]

On Sat, 6 Nov 2021 at 10:04, D. Hugh Redelmeier via talk
 wrote:
>
> | From: Jim Ruxton via talk 
>
> | I'm having trouble with a simple script. It adds a second sounccard. This
> | command works on the command line but not in a script. Any ideas why?
>
> How are you running that script?  Is it from a cron script or something
> like it?  Does your script work when invoked from your command line?
>
> I haven't paid attention to JACK, but I assume that each Linux user
> session has some kind of jack context -- plumbing done by one user should
> not affect the plumbing of another user.  I don't really see this issue
> addressed in the first few jack documents I read.
>
> My guess is that your script is modifying a different context from the one
> you are observing.  That's really vague and hand-wavy, but that's the
> direction I'd investigate.
> ---
> Post to this mailing list talk@gtalug.org
> Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Similar to Hugh's ideas, but more specific, I would check first the
PATH (mentioned by several other people) and then the environment as a
whole.  I'm not sure that what I'm going to suggest is the best way to
do this, but it's the one I know: run the command 'export' and look
for any settings that look to be specific to the program you're
running, then make sure those same settings are set in the script
you're working on.  You can look at the PATH this way as well.
Scripts often (especially under cron) run in different environments
from what you expect, and thus don't get the information they need to
function properly.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] keyboards

[Giles Orr via talk]

On Fri, 22 Oct 2021 at 12:44, Jason Shaw via talk  wrote:
> On Fri, Oct 22, 2021 at 12:24 PM Giles Orr via talk  wrote:
>> There appear to be a lot of definitions of what "ergonomic" means when
>> it comes to keyboards.  I tend to the most extreme example: I own
>> three (working) Kinesis Advantage keyboards.  The Advantage2 can be
>> had for $400-$450 Canadian - not cheap.  They come with Cherry brown
>> switches, but I have a habit of retrofitting them with Cherry blues -
>> which makes them noisier and more expensive.
>>
>> Everybody is recommending the keyboard(s) they love ...  I have a
>> CoolerMaster with Cherry Blues, and several IBM model Ms, and have had
>> many other types of "ergonomic" keyboards over the years.  I love the
>> feel of the switches in both the CoolerMaster and the model Ms, but I
>> find I need the two halves of the keyboard farther apart for comfort.
>> I also prefer vertical columns of keys - as opposed to the now totally
>> unnecessary leftward slant of key columns on almost all modern
>> keyboards.  I've experimented with keyboards a lot to end up where I
>> am: the Kinesis Advantage took a month to adjust to, but has been
>> worth it ... for me.  It's my daily driver both at home and at work.
>> You have to make your own call on these things.
>>
>> A note about gaming keyboards: gamers seem to tend to prefer low
>> activation force, linear keys (Cherry Blacks or Cherry Reds).  Most
>> people who type for a living (as opposed to gaming) seem to prefer
>> "tactile" keys, which is quite different from the "linear" keys which
>> don't have any feedback at all until they bottom out.  I hate linears
>> (but again - personal taste).  And then there's the "clicky" thing:
>> the noise the IBM model M makes can be enough to clear a small room.
>> Some people really really hate that noise.  I get that, but I still
>> love the feel of those things.
>>
>> Further reading (my intro to keyboard layouts and key types):
>> https://www.gilesorr.com/blog/computer-keyboards.html
>
>
> Thanks for that Giles.
>
> Keyboards are a very personal choice, and what's right for one user isn't 
> going to be right for another.
>
> Personally, I use a ThinkPad X1 Carbon quite regularly and have for ~10 
> years, and love the touchpoint (red nub), so even my external keyboard for a 
> long time was the ThinkPad TrackPoint kb 
> (https://www.lenovo.com/ca/en/accessories-and-monitors/keyboards-and-mice/keyboards/KBD-BO-TrackPoint-KBD-US-English/p/4Y40X49493).
>  Over time though, my posture was suffering and my wrists were sore at the 
> end of a workday. Combined with a personal accident that resulted in some 
> permanent fingertip damage on my dominant hand, I needed something with 
> better ergonomics, so I took the plunge on a split style, mechanical 
> keyboard, and can't imagine going back.  The added benefit of having the desk 
> area immediately in front of me wide open is great too.
>
> I would like to try an ortholinear split keyboard at some point, but have no 
> real desire to spend a month or two re-learning to type again.

Jason (and Dave),

if you can find an "ortholinear" keyboard (and/or split keyboard)
that's not a Kinesis or Maltron, you'll find the finger retraining
period is actually quite small.  The "month" I've been given to quote
as my adjustment period for the Kinesis has almost nothing to do with
the vertical key columns or the split.  What takes so long to adjust
to is the weird (but good!) key curve, the repositioned keys, and
especially the much-increased use of the thumb.  And I know these are
the causes because I had a vertical row keyboard before I moved to the
Kinesis: the key layout was otherwise traditional, and adjusting to it
took only a couple days.

I've also found that switching between keyboard layouts is easy now
that I know both.  I use a Kinesis as a daily driver, but I can switch
to a standard keyboard at the drop of a hat.  The worst thing I do is
occasionally hit the space bar with my left thumb because that's where
the Escape key should be (what can I say - a remapping for a hardcore
Vim user), and I usually get over that in about ten minutes.

I'm going to add a "Hallelujah!" for the Matias placement of the "6"
key on the right hand: the vast majority of split keyboards put this
on the left, which is incorrect for touch typing.  And again a matter
of personal taste ...

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] keyboards

[Giles Orr via talk]

On Thu, 21 Oct 2021 at 16:40, o1bigtenor via talk  wrote:
>
>
>
> On Thu, Oct 21, 2021 at 2:41 PM D. Hugh Redelmeier via talk  
> wrote:
>>
>> | From: Evan Leibovitch via talk 
>>
>> |   The emergence of high-end gaming on PCs has led to a quest for
>> | keyboards that are super-responsive and comfortable for long periods.
>>
>> Yeah.  That seems like the best place to look.
>>
>> Beware: I've found their goals are not completely aligned with mine.
>> Here are some.
>>
>> One feature that I don't care about is rainbow coloured lights for the
>> keys.  Benign, but you are paying for this.  On mine (a Razer Blackwidow
>> Ultimate -- love the names), I've got a green light per key, solidly on
>> unless I install a daemon to change that.
>>
>> Another is "tenkeyless" which means "without numeric keypad".  I use
>> the numeric keypad and don't want to lose it.
>>
>> Another is low-latency: I've never noticed keyboard latency.
>>
>> Another is: ugly seems to be valued.
>>
>> All support n-key rollover, for large n.  This requires sending multiple
>> USB packets per keystroke.  This in turn confuses the firmware on a couple
>> of our machines: if we wish to adjust firmware settings, we have to plug
>> in a different keyboard.  As you can imagine, that took some effort to
>> figure out.
>>
>> There are many web pages that describe the colours of switches.
>> 
>>
>> My wife likes "blue" keys (noisy and tactile).
>> I like "brown" keys (less noisy but tactile).
>>
>> If you care a lot, you may care about the company that produced the
>> switches.  I haven't bothered to sudy this aspect.
>>
>
> Interesting comments from all of the responders so far.
> Is anyone using an ergonomic keyboard from this group.
> The old rectangular keyboard makes my hands ache at the
> thought of using only this. I have some cheap keyboards gotten
> with machine purchases - - - my working keyboard for anything
> more than a few keystrokes is an ergonomic version. I'm about
> 60 cm across the shoulders so a keyboard that's some 35 cm
> just isn't comfortable.
>
> I also like my numeric keypad - - - - do a lot of entries on that
> for business use and would like to have it part of the keyboard
> if at all possible.
>
> I have used the mechanical keys in a very long time - - - think
> I would prefer less noise rather than more but for high quality
> would like give on that!
>
> Re: gaming - - - - I'm having too fun fun and use far too much time
> on my system without playing any games - - - tend to relax with a
> book (most often a physical copy too).
>
> Thanking one and all for their input!

There appear to be a lot of definitions of what "ergonomic" means when
it comes to keyboards.  I tend to the most extreme example: I own
three (working) Kinesis Advantage keyboards.  The Advantage2 can be
had for $400-$450 Canadian - not cheap.  They come with Cherry brown
switches, but I have a habit of retrofitting them with Cherry blues -
which makes them noisier and more expensive.

Everybody is recommending the keyboard(s) they love ...  I have a
CoolerMaster with Cherry Blues, and several IBM model Ms, and have had
many other types of "ergonomic" keyboards over the years.  I love the
feel of the switches in both the CoolerMaster and the model Ms, but I
find I need the two halves of the keyboard farther apart for comfort.
I also prefer vertical columns of keys - as opposed to the now totally
unnecessary leftward slant of key columns on almost all modern
keyboards.  I've experimented with keyboards a lot to end up where I
am: the Kinesis Advantage took a month to adjust to, but has been
worth it ... for me.  It's my daily driver both at home and at work.
You have to make your own call on these things.

A note about gaming keyboards: gamers seem to tend to prefer low
activation force, linear keys (Cherry Blacks or Cherry Reds).  Most
people who type for a living (as opposed to gaming) seem to prefer
"tactile" keys, which is quite different from the "linear" keys which
don't have any feedback at all until they bottom out.  I hate linears
(but again - personal taste).  And then there's the "clicky" thing:
the noise the IBM model M makes can be enough to clear a small room.
Some people really really hate that noise.  I get that, but I still
love the feel of those things.

Further reading (my intro to keyboard layouts and key types):
https://www.gilesorr.com/blog/computer-keyboards.html

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Heads up: Ubuntu 21.10 kills your desktop icons

[Giles Orr via talk]

On Wed, 20 Oct 2021 at 11:03, Lennart Sorensen via talk  wrote:
>
> On Wed, Oct 20, 2021 at 10:47:30AM -0400, Michael Hill via talk wrote:
> > I prefer the elegant design and centralized vision of the GNOME
> > desktop, but the whininess of the GNOME hatred here never gets old.
>
> When a desktop going from version 2 to 3 throws away everything users
> are used to, and the developers simply don't care, then it deserves any
> hate it receives.
>
> If they wanted to have a totally new vision (which is perfectly fine to
> do), they should have started a new project, not hijacked an existing one.

I think that Lennart has just reduced the whole problem to one
sentence.  A new vision is good - the history of Linux is littered
with "new visions," most of them long deceased ... however, the ones
that succeed move us forward.  But GNOME - as he points out - didn't
so much create a new vision as simultaneously hijacking and destroying
an old one.  Hugh sees this as a rewrite, but when almost nothing of
the former "vision" or paradigm remains, it's really a new project.
And yeah, if you spin it off as a new project you may be abandoning
the old one - but at least if you abandon it there's an opportunity
for someone to pick it up ...

Which brings up an interesting point.  As much as people scream about
GNOME3, I don't think anyone forked GNOME2, which they're more than
welcome to do.  Someone offended by KDE4 forked KDE3 and carried it on
as the Trinity Desktop Environment ( http://www.trinitydesktop.org/ ),
which is still going.  So I guess the GNOME3 haters just aren't that
motivated ...

If anyone is wondering where I stand on this (I'm sure most don't care
but one or two might) - I never liked GNOME2, and I really dislike the
behaviour of GNOME3.  But I don't have any skin in the game because I
still use Openbox on most machines (with 'fbpanel' and 'dmenu' or
'rofi' as a launcher).  If I feel the need for a full desktop, LXQt
isn't bad and lighter weight than the alternatives.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Do people have opinions about the Framework modular laptop

[Giles Orr via talk]

On Tue, 19 Oct 2021 at 07:57, Dave Collier-Brown via talk
 wrote:
>
> On 2021-10-18 16:12, Lennart Sorensen via talk wrote:
>
> > I would certainly also like to see an AMD option.
> >
> > Although for me the main issue is that a 13.5" laptop has zero interest
> > to me.  Too small.  Come with a 15" or 17" and I will be interested.
>
> My wife has a 15" Mac laptop, and never takes it on trips because it's
> so big.
>
> I have a 14" thinkpad and a tiny netbook, and the thinkpad is just a
> little large for my backpack. 13.5" sounds about right.  At home and in
> the office I have docking stations hooked to quite large screens, so the
> laptop screen usually only displays my email (;-))
>
> ---dave

If I'm at home, the laptop will be connected to an external screen -
so I don't care about screen size.  If I'm travelling ... anything
over 14" is murder on my back when I'm carrying it, so I'll just deal
with the small screen size.  I like a big screen, but for a laptop,
size (or more correctly, "weight") is the constraining factor.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Blank screen can't be refreshed

[Giles Orr via talk]

Hi folks.

My main machine is a Fedora 34 laptop that drives a 43" 4K Philips
monitor (it's actually a "monitor," not a "TV").  I've set
'xscreensaver' to blank the machine after five minutes.  The delay is
that short because the Philips (while otherwise a nice display) is
notorious for ghosting.  It's a Barrier client to another computer (a
NUC) - so the keyboard and mouse that control this machine are
actually attached to another machine.

Occasionally, the machine/screen can't be awoken from the blank screen
state.  The Philips backlight is on, I have a pointer I can move
(either by Barrier, or by the touchpad on the laptop itself) ... but
the screen remains blank.  The machine is up and running, and I can
SSH in and change stuff, but nothing non-disruptive I've found wakes
the screen.  I tried 'kill -HUP ' to no effect.  I tried
'openbox --restart' but it failed because it's a remote machine (I can
probably over-ride this with an appropriate environment variable(s),
but haven't got to that).  'kill ' - also had no effect.
But 'kill -9 ' brought me back to the login screen
(functioning correctly on the previously blank screen).  This (and
rebooting) are the only solutions I've found so far.  This problem has
usually occurred on wake from suspend-to-RAM, but has now happened a
couple times on wake from just screen blanking.

Having had to write that out, it makes me think that Openbox is the
problem.  Openbox is supposed to accept 'kill -HUP' and do something
sane with it, but it's clearly got itself in a state where it's not
listening to 'kill' at all (thus the need for '-9').

I've looked at the machine logs: I'm not great at reading them, but
I'm not seeing anything that looks like this problem, and I did spend
a fair amount of time looking ...  If it is Openbox, I should be
looking at the X logs, which I haven't done yet?  I've used Barrier
heavily for many years, and this problem hasn't manifested on any of
the several other computers that have been Barrier clients to the NUC
... but should I consider Barrier?  What about xscreensaver itself?

I could potentially compile Openbox rather than relying on the Fedora
package.  I don't want to do that, but if needed ...  Huh - I suppose
I could also switch to Fluxbox for a while, and see if it has the same
issue.

Any ideas appreciated, any thoughts on where to look or how to debug ...

--
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Do people have opinions about the Framework modular laptop

[Giles Orr via talk]

On Fri, 15 Oct 2021 at 18:42, GS via talk  wrote:
>
> I've had my eye on it since the reviewers first got their hands on it. So far 
> the reception has been overwhelmingly positive but finding opinions you can 
> trust can be difficult. I tend to trust Cory, Leo and Linus.
>
> https://doctorow.medium.com/the-framework-is-the-most-exciting-laptop-ive-ever-used-5415da0a46e5
>
> https://www.youtube.com/watch?v=TKTFBpEET0o
>
> https://www.youtube.com/watch?v=LSxbc1IN9Gg
>
> On 2021-10-15 17:59, Dave Collier-Brown via talk wrote:
>
> I see from Slashdot that it has an app store now:
>
> https://hardware.slashdot.org/story/21/10/14/2110258/modular-framework-laptop-gets-marketplace-for-all-those-modules
> https://www.theverge.com/2021/10/14/22725935/framework-laptop-modular-marketplace-launches-spare-parts-upgrades-expansions
>
> and I wonder if anyone has an opinion of it.

Cory is good Toronto content, and among his many achievements, I think
excessive honesty is one of them so yes, I'll trust his review.  On
the other hand, reading it is a bit more than a review of the
Framework, as he lays into Lenovo's ThinkPads and mentions that he
just had a hip replacement.  Wait, what?  I think of Cory as "young"
(we shared some social groups many years back), what's he doing
getting a hip replaced?  Okay, I'm getting slightly off track (but
then, so did he).

I'm enthusiastic about the Framework, and am a likely purchaser ...
but I have to admit I'm really, really hoping they come out with an
AMD motherboard for the thing before I need my next laptop.  That's
the only thing I thought was missing: I'm not happy with Intel.  Not
their market dominance, not their poor price-to-performance ratio, and
not their chip's security flaws.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] [GTALUG-Announce] Meeting Tomorrow

[Giles Orr via talk]

On Tue, 10 Aug 2021 at 12:22, David Mason via talk  wrote:
>
> I just thought I’d point out that the meeting page gets screwy if you look at 
> it with a window wider than about 80% of a 2560 screen (the left navigation 
> bar slides over the content block). Same on Firefox, Chrome, and  Safari.
>
> ../Dave
> On Aug 9, 2021, 10:03 PM -0400, hi--- via talk , wrote:
>
> https://gtalug.org/meeting/2021-08/

Dave's assessment perhaps doesn't need a +1, but since I do websites
for a living, I was curious.  And yes, the TLUG meeting page goes
quite wonky on wide screens with bad overlap between the two columns.
Ah - they're not columns: I've never used the  tag, but it
seems to be the issue here?  Works great at narrow widths, breaks down
with extra wide ones.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] ThinkPad P15 -- crap?

[Giles Orr via talk]

On Thu, 1 Jul 2021 at 08:59, Scott Sullivan via talk  wrote:
>
> On 2021-06-26 2:42 a.m., ac via talk wrote:
> > On Sat, 26 Jun 2021 01:06:23 -0400
> > William Park via talk  wrote:
> >> Hi guys,
> >> I'm using ThinkPad P15 Gen 1 at work.  The keyboard, touchpad, and
> >> trackpoint are not what I expected.  They feel cheap.  My old T450 is
> >> better.
> >> Is it just me?  Or, has anyone notice this also?
> >>
> > not just you, I also am of the opinion that it is cheap crap :(
> > personally I have now started using HP, the quality is just better
>
> Over the last three years my employer has issued me both an HP Elitebook
> G3 (Intel) and a G6 (AMD). Both have run linux well, both have had
> trackpoints. The build quality was decent on both, plastic on the G3,
> aluminum on the G6. The G3 machines had a common screen defect were one
> corner would develop a backlight hotspot, but it was readily ignorable.
>
> HP's dedicate dock for them also worked well, and the USB-C dock option
> has also been working well on the G6.
>
> Getting into the cases has always been easy with just a few screws on
> the back panel and no-nonesense access to RAM and M.2 slots.
>
> The BIOSes include a handy option to swap the fn and ctrl keys, which
> are reversed compare to thinkpad layouts. Handy for me as I'm used to
> the thinkpad layout. I do miss the dedicated middle click button that
> the HPs lack, but middle button emulation (holding buttons 1 & 3) works.

My experience with HP has mostly been with their consumer-grade
machines - which have generally been poor.  We ran into them more than
most at Repair Cafes, with busted hinges and cracked plastic as the
screws sunk in plastic aged and busted ...  There seems to be a big
difference between their consumer and their "professional" (or
whatever you want to call it) lines.  I wouldn't ever consider their
consumer products after what I've seen of them.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] ftp helper app, and how to screenshot on Ubuntu 21.04?

[Giles Orr via talk]

On Fri, 4 Jun 2021 at 19:28, Stewart C. Russell via talk
 wrote:
>
> also it seems this upgrade has borked xclip, which is a can't-live-without
>
> Did anyone think Wayland was a good idea? A desktop without a working
> clipboard?? And by working, I mean "the way it always did before"

When change makes things better, we love it.  When it interferes with
the way we work, we hate it.  But the reality is that Linux has to
evolve.  And you can either learn how to rewrite 'xclip' yourself or
find a replacement if it doesn't work anymore.  That's the wonder and
horror of open source.  Yes, Wayland is causing a lot of grief.  But
go talk to Scott about it sometime: he's got me convinced that it's a
very good idea ... we just have to struggle through the rough spots.

In other news: I really liked 'scrot' too.  I've switched to
'flameshot', which is a radically different tool.  It's graphical and
much more flexible - it may or may not work for you.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Looking for a can shield for an UNO or compatible

[Giles Orr via talk]

On Fri, 14 May 2021 at 10:29, Dave Cramer via talk  wrote:
>
> Is there a local source for something like MakerFocus CAN-Bus Shield V1.2
> ?
>
> Bonus points if it had GPS
>
> Dave Cramer

This is snark - not helpful, just amusing(?).

I love the fifth image where they convert the board length from 78mm
to 30.7 inches.  Our American compatriots who can't read metric are
going to be expecting the delivered board to be bigger than a letter
size sheet of paper.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] War Story: Fedora 33 update silently broke Dovecot (IMAP / POP3 server)

[Giles Orr via talk]

On Sun, 9 May 2021 at 08:30, Giles Orr  wrote:
>
> On Fri, 7 May 2021 at 22:43, D. Hugh Redelmeier via talk
>  wrote:
> >
> > Only one of us uses IMAP / POP3.  She stopped being able to pick up mail.
> > The message was obscure: Thunderbird reported getting resets from the
> > server.
> >
> > Detective work (leaving out the blind alleys):
> >
> > Look for funny messages in the server log.  The right command is arcane:
> > $ journalctl -b _SYSTEMD_UNIT=dovecot.service
> >
> > - -b means "since last boot".  Very handy because the log can go back a
> >   long time.
> >
> > - _SYSTEMD_UNIT=dovecot.service: only messages about dovecot.
> >   Such an intuitive form.  It won't work without the .service.
> >   It turns out that there is a shortform, -u or --unit=.
> >
> >
> > May 04 14:07:12 redhop-mimosa-com dovecot[977]: config: Warning: please set 
> > ssl_dh= > May 04 14:07:12 redhop-mimosa-com dovecot[977]: config: Warning: You can 
> > generate it with: dd if=/var/lib/dovecot/ssl-parameters.dat bs=1 skip=88 | 
> > openssl dhparam -inform der > /etc/dovecot/dh.pem
> > May 04 14:07:12 redhop-mimosa-com dovecot[977]: pop3-login: Error: Failed 
> > to initialize SSL server context: Can't load DH parameters (ssl_dh 
> > setting): error:1408518A:SSL routines:ssl3_ctx_ctrl:dh key too small: 
> > user=<>, rip=192.139.70.95, lip=192.139.70.82, session=
> > May 04 15:30:28 redhop-mimosa-com dovecot[977]: pop3-login: Error: Failed 
> > to initialize SSL server context: Can't load DH parameters (ssl_dh 
> > setting): error:1408518A:SSL routines:ssl3_ctx_ctrl:dh key too small: 
> > user=<>, rip=192.139.70.95, lip=192.139.70.82, session=
> >
> > The first two messages were almost impossible to read because the were in 
> > yellow.
> >
> > It turns out that the first means that you have to edit 
> > /etc/dovecot/conf.d/10-ssl.conf
> > and change
> > #ssl_dh =  > to
> > ssl_dh =  >
> > /etc/dovecot/dh.pem was already there.
> >
> > OPTIONAL:
> >
> > /etc/dovecot/dh.pem specified a really weak Diffie-Hellman group.
> > Your should probably strengthen it.  I chose ffdhe2048, a weak one (2048 
> > bits) that is still "OK".
> > See:
> > 
> >
> > You need it in PEM format.  First save the old one.  Then grab a new one:
> > curl https://ssl-config.mozilla.org/ffdhe2048.txt > 
> > /etc/dovecot/dh.pem
> >
> > Then adjust ownership and permissions.
>
> This sent me on a bit of a chase.  Nginx uses a DH parameters file
> that's used in the same way.  And it's considered a bad idea to use a
> widely known DH parameter (like the one that ships with the software,
> or that sits on a Mozilla server).  This is a semi-useful read:
> https://security.stackexchange.com/questions/94390/whats-the-purpose-of-dh-parameters.
> I say "semi-useful" because honestly some of it was over my head.  And
> that leads to reading about Logjam ( https://weakdh.org/ ).
>
> With Ansible I've automated the generation of a new DH parameter file
> on each new server:
>
>  openssl dhparam -out  
>
> Generating this file takes a significant amount of time (minutes) if
> "size" is reasonably large (2096, although I would recommend 4192)
> even on a modern machine.  But it's a one-off for any given machine,
> so it's a small imposition that I'd recommend.  If this breaks the
> mail setup, you can obviously return to the known value from Mozilla.

"2096" "4192" ...   They're just numbers, but I did mean to
say "2048" and "4096" as regular power-of-2 numbers ...

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] War Story: Fedora 33 update silently broke Dovecot (IMAP / POP3 server)

[Giles Orr via talk]

On Fri, 7 May 2021 at 22:43, D. Hugh Redelmeier via talk
 wrote:
>
> Only one of us uses IMAP / POP3.  She stopped being able to pick up mail.
> The message was obscure: Thunderbird reported getting resets from the
> server.
>
> Detective work (leaving out the blind alleys):
>
> Look for funny messages in the server log.  The right command is arcane:
> $ journalctl -b _SYSTEMD_UNIT=dovecot.service
>
> - -b means "since last boot".  Very handy because the log can go back a
>   long time.
>
> - _SYSTEMD_UNIT=dovecot.service: only messages about dovecot.
>   Such an intuitive form.  It won't work without the .service.
>   It turns out that there is a shortform, -u or --unit=.
>
>
> May 04 14:07:12 redhop-mimosa-com dovecot[977]: config: Warning: please set 
> ssl_dh= May 04 14:07:12 redhop-mimosa-com dovecot[977]: config: Warning: You can 
> generate it with: dd if=/var/lib/dovecot/ssl-parameters.dat bs=1 skip=88 | 
> openssl dhparam -inform der > /etc/dovecot/dh.pem
> May 04 14:07:12 redhop-mimosa-com dovecot[977]: pop3-login: Error: Failed to 
> initialize SSL server context: Can't load DH parameters (ssl_dh setting): 
> error:1408518A:SSL routines:ssl3_ctx_ctrl:dh key too small: user=<>, 
> rip=192.139.70.95, lip=192.139.70.82, session=
> May 04 15:30:28 redhop-mimosa-com dovecot[977]: pop3-login: Error: Failed to 
> initialize SSL server context: Can't load DH parameters (ssl_dh setting): 
> error:1408518A:SSL routines:ssl3_ctx_ctrl:dh key too small: user=<>, 
> rip=192.139.70.95, lip=192.139.70.82, session=
>
> The first two messages were almost impossible to read because the were in 
> yellow.
>
> It turns out that the first means that you have to edit 
> /etc/dovecot/conf.d/10-ssl.conf
> and change
> #ssl_dh =  to
> ssl_dh = 
> /etc/dovecot/dh.pem was already there.
>
> OPTIONAL:
>
> /etc/dovecot/dh.pem specified a really weak Diffie-Hellman group.
> Your should probably strengthen it.  I chose ffdhe2048, a weak one (2048 
> bits) that is still "OK".
> See:
> 
>
> You need it in PEM format.  First save the old one.  Then grab a new one:
> curl https://ssl-config.mozilla.org/ffdhe2048.txt > 
> /etc/dovecot/dh.pem
>
> Then adjust ownership and permissions.

This sent me on a bit of a chase.  Nginx uses a DH parameters file
that's used in the same way.  And it's considered a bad idea to use a
widely known DH parameter (like the one that ships with the software,
or that sits on a Mozilla server).  This is a semi-useful read:
https://security.stackexchange.com/questions/94390/whats-the-purpose-of-dh-parameters.
I say "semi-useful" because honestly some of it was over my head.  And
that leads to reading about Logjam ( https://weakdh.org/ ).

With Ansible I've automated the generation of a new DH parameter file
on each new server:

 openssl dhparam -out  

Generating this file takes a significant amount of time (minutes) if
"size" is reasonably large (2096, although I would recommend 4192)
even on a modern machine.  But it's a one-off for any given machine,
so it's a small imposition that I'd recommend.  If this breaks the
mail setup, you can obviously return to the known value from Mozilla.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Fwd: Developer job postings at TPL

[Giles Orr via talk]

Toronto Public Library has posted three developer positions (and some
other tech jobs!):

https://www.torontopubliclibrary.ca/about-the-library/library-jobs/

You'll find them titled "ITS Senior Specialist (Web & Mobile
Development)."  I've heard the pay may not look competitive, but there
are other benefits: a strong union means when we tell you you'll be
working a 35 hour week, we mean it.  Equally important, I really like
our team.

Feel free to mail me off list if you have questions (I only really
know about the developer positions, the others are in another area).

--
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Docker, NodeJS and complexity

[Giles Orr via talk]

On Thu, 22 Apr 2021 at 09:56, William Witteman via talk  wrote:
>
> I have some existing projects to contribute to that use NodeJS and Docker, 
> and I am finding that getting set up to work on them is extremely daunting.
>
> I have never really used Docker.  I have never really used NodeJS/npm/etc.
>
> I do know Javascript, but I don't build projects like this - it reminds me of 
> Ruby-on-Rails, which I found to be a huge snarl of complexity.
>
> I know that the initial development was done on Macs, but in theory at least, 
> I should be able to set up the Docker container and use that to get me into 
> an equivalent development environment, but I don't really know how.
>
> Does anyone have a Debian-flavoured tutorial for setting up Docker/NodeJS so 
> that I can build and run someone else's code?

Hi William.

Another option to consider is VirtualBox (or something like it).  I
suppose this depends on your familiarity with VB: I'm very familiar
with it, so spinning up a new machine in VB is quite straightforward
for me.  I'm not claiming VB is better (it definitely uses more
resources), just that you could find it easier if you're comfortable
with it.  I also found Docker a struggle to get started with, and
throwing a new language on top of that makes it really daunting.  Best
of luck.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] You may already have a Windows licence and not know it …

[Giles Orr via talk]

On Tue, 13 Apr 2021 at 19:48, Znoteer via talk  wrote:
>
> On Tue, Apr 13, 2021 at 07:42:47PM -0400, Stewart Russell via talk wrote:
> > If the file /sys/firmware/acpi/tables/MSDM exists (it's read-only to root),
> > there's probably a key embedded in there you can pull out with strings
> >
>
> I don't seem to have that file, and I'm not surprised.  Under what
> circumstances might one have that file on a linux system?

When you installed Linux on a Windows computer?  I wiped Windows, but
the firmware data is still there:

# cat /sys/firmware/acpi/tables/MSDM
MSDMU:_ASUS_Notebook ASUS-K--K-K#

(License string obfuscated.)  'strings' not required.

Fascinating: thank you Stewart, this may make it possible to install
Windows 10 in VirtualBox - of course I still have to lay hands on a
matching version of Windows, although those are easier to find than
legitimate license keys.  Hmm - I wonder if the "ASUS" at the
beginning of the key means it only works with Asus versions of W10 ...

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] TIL: rtcwake / wakealarm

[Giles Orr via talk]

On Fri, 9 Apr 2021 at 22:23, Stewart C. Russell via talk
 wrote:
> This may be old news to many of you, but today I learned you can have the 
> computer's real time clock boot your machine at a specific time. The more 
> proper way seems to be to use the 'rtcwake' command, but you can also do it 
> by writing the timestamp of the startup time to /sys/class/rtc/rtc0/wakealarm
>
> Some more details of the /sys method - 
> https://www.raspberrypi.org/forums/viewtopic.php?f=63=309093=1849326=ca7a14f7d160f929378be4691b1aed9e#p1849291

This sounded pretty useful to me too, so I thought I'd try it out on
an old netbook.  I decided to use the netbook before messing about
with power events on a computer I cared about.

# rtcwake --date +5min --mode mem

I chose "--mode mem" because this machine has often and successfully
been suspended to RAM.  The machine started flashing the power button
orange as soon as I ran the command, which means "suspended-to-RAM."
Five minutes later the power button turned steady blue again.  But the
screen remained off, and didn't wake on many keypresses.  Including
even on the power button.  The machine was also not on the network,
un-pingable.  In the end, it was the five second press on the power
button to kill it completely.

This has left me reluctant to try this on any other machine.
Presumably it would work fine on a Raspberry Pi as it came from a Pi
forum and they're a fairly homogeneous form of hardware.  But my one
attempt doesn't recommend this ...

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Surveillance Capitalism [was another thread]

[Giles Orr via talk]

On Fri, 2 Apr 2021 at 16:38, Dhaval Giani via talk  wrote:
>
> >
> >> are all aware of. Many women were uncomfortable around RMS and avoided
> >> him. Many refused to participate in our community because of
> >> interactions with him. Do you think RMS is more important than a
> >> community of developers he is pushing away?
> >
> >
> > See all the stuff you say we are all aware in this message is just rumors 
> > and innuendo to me.
> >
>
> Wait, so all these women saying those words are rumours and innuendo?
> You choose to disbelieve them? After a pattern of behaviour that
> multiple people have confirmed and talked about?
>
> > What are not rumors and innuendo are the historical facts on IBM, their 
> > influence, their power and powerful friends and most importantly their big 
> > ball of money which they spend on influenceing the influencers.
> >
> >>
> >> I want to explicitly state this. RMS is a major reason free software
> >> is where it is. RMS's contributions to free software are gigantic.
> >> However, RMS cannot be a leader of our community if he continues to
> >> isolate a significant population of prospective developers. RMS the
> >> contributor - YES. RMS the leader - NO.
> >>
> >> RMS cannot be the poster child of our community if it is going to be
> >> relevant in the future.
> >
> >
> > This is where being the willing poster child of a charitable institution, 
> > used to raise funds, diverges from the science of truth and innovation.
> >
> > In the legal science of truth, a person is innocent until proven guilty in 
> > a court of law. However, media and the media barons in conteol, crucify 
> > persons and their personas daily, just to make a buck.
> >
>
> And no one has charged RMS with a crime. All we are saying is, he is
> not representative of a majority of us, and we don't want him to
> represent us. Some of us are minorities who have heard racist
> statements being made by prominent folks in the community and have
> made us feel our contributions are not valued. It is not hard to
> believe after that experience that other prominent folks can be
> sexist. RMS has not stepped up and owned up to his actions and
> apologized. I have no problem with people growing. We all make
> mistakes. But doubling down like this, well I don't want to be a part
> of that community. And the reality is, there are tons of "other"
> people who will not join in and we will never know. So yes, if the
> choice is between thousands of those people, having a diverse
> community, growing and being relevant to the world, I would rather RMS
> step down than us lose this community. And I would rather you leave
> the community if you think being more welcoming to other voices is not
> important. We don't need your contributions at the risk of alienating
> many more people.
>
> Again, I restate this. RMS as a contributor - yes. RMS as a leader -
> no. He doesn't represent me, and he certainly doesn't represent the
> community of foss developers. This is a discussion about RMS, not the
> conspiracy theories you are throwing about.

I've spent a lot of time thinking and reading about RMS since the
first thread started in TLUG.  I don't know if anyone has mentioned
this: Stallman founded the FSF (maybe most people know that, but I'd
forgotten).  Sometimes organizations grow beyond their founders.  As
Dhaval says: free software, and particularly the GNU project, would
never have achieved what it has without him.  That doesn't make him a
nice guy: he's influential, intelligent, and a good coder.  I've dealt
with him myself - 20 years ago, but it was a memorable experience and
not in a good way.  His utter inflexibility was on full display.  He
believes what he believes, and no compromises are acceptable.  That
kind of focus drove free software forward in its early years, but now
organizations like FSF are more about outreach than bulldozing a path
for a new way of doing things.  And I think multiple articles already
mentioned have shown that if there's one thing RMS is not, it's a
diplomat.  RMS should rightly be honoured for his contributions, but
he's not the person I want as the face of an organization that
represents my beliefs.  His attitudes and rigidity make him
unwelcoming to newcomers.  No, he hasn't done anything illegal: but he
doesn't have to to be a poor choice as a public representative.  What
Linux and free software need right now are leaders who can usher
people in rather than driving them away.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Has the graphics-card world gone mad?

[Giles Orr via talk]

On Tue, 30 Mar 2021 at 13:08, Anthony de Boer via talk  wrote:
> Evan Leibovitch via talk wrote:
> > I have two 4K screens (Samsung U28E590D) that I want to use in day to day
> > work. Maybe some streaming, but the most intensive game I would play is
> > Cities: Skylines
> > Any suggestions on what is the minimum GPU that will reasonably drive them?
> > I've backordered a GTX 1660 at $315. Would a 1650 Super be enough?
>
> If all else fails, a Raspberry Pi 4 will drive a pair of high-end
> monitors, probably not with the grunt that any modern game would want,
> but enough to be the new X Terminal.

That just shifts the problem.  He has a good computer with a
not-entirely-working video card.  You're suggesting something that
would put him in a situation where he'd have a passable video card
with a not-entirely-working computer attached to it.

Don't get me wrong: I use Raspberry Pis for a lot of things.  But I
don't use them as my daily driver, because they don't have the
horsepower.  Shifting to ARM would also break most games.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] what MUA do people like? Use?

[Giles Orr via talk]

On Sun, 21 Mar 2021 at 11:57, D. Hugh Redelmeier via talk
 wrote:
>
> | From: William Park via talk 
> | Date: Sat, 13 Feb 2021 13:52:21 -0500
>
> | I'm beginning to like Thunderbird.  It replaces fetchmail, procmail, 
> sendmail,
> | mutt, and vim for mailing.  I miss vim, though.
>
> I'm interested in improving my email routines.  I spend a lot of time
> with email so any improvement would be useful.
>
> So: what do you use?  Why?
>
> Here's my answer.  It is not a recommendationo.
>
> Most of us in this household use Alpine.  Thats a venerable text-based-GUI
> mail user agent.  I've been using Pine/Alpine since the early 1990s (when
> I switched from Berkeley mail (like mailx on Linux)).  As you can tell,
> I'm quite conservative.  I used to say that I changed MUAs every decade,
> but I'm behind now.
>
> Why do I like and stay with Pine?
> - inertia
> - modest subset of EMACS keystrokes
> - stable but well-maintained
> - works well through SSH
> - I'm very comfortable with it
> - has most features that I know that I want.
> - Alpine does not hold my mail hostage: ordinary UNIX text tools can get
>   at it.  (Alpine supports various formats but I use mbox.)
>
> Molly (my wife) uses Thunderbird.
> - she's used to WIMP GUIs
> - she does not use any advanced features
> - dislikes and avoids updates
>
> Thunderbird:
>
> - seems attractive
>
> - was busted by a Ubuntu update that I had to diagnose and back out of.
>   I pinned the version of Thunderbird and the library at fault.
>   Nothing said by Ubuntu folks convinces me that it is safe to unpin
>   (archived mail is very important).
>
> - I don't know how to export the Thunderbird mail archives (but I haven't
>   put my mind to the problem)
>
> - future looks precarious.  Mozilla seems to have cut the Thunderbird
>   project loose
>
> GMail:
>
> - seems to be taking over the world
>
> - I'm sometimes forced to use it.
>
> - runs well on smart phones
>
> - someone else does the maintenance
>
> - But: I want control over my mail.  I don't want it in the cloud.  I
>   don't want it to go through Google's hands (we run our own mailserver).
>
> - I want painless offline access to mail archives---
> Post to this mailing list talk@gtalug.org
> Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

I have two primary email accounts, one at Gmail and one at Protonmail
(Protonmail has free accounts: I like them enough that I'm paying for
their account, although on usability Gmail is definitely better ...).
I was intending to move entirely to Protonmail, I seem to have stalled
on that.

I have Thunderbird set up at home, primarily as a way to back up Gmail
and Protonmail to my hard drive: it's rare that I use Thunderbird as a
client, but it's always worked fine when I do.  Protonmail, because of
its encryption, requires a special "bridge" software to allow local
clients to access their mail server.  It's a lousy arrangement, but
does seem to work - and I bought into their service in part because of
that encryption.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] War Story: debugging remote port access

[Giles Orr via talk]

A couple quick details:
- I believe the TLUG list was down: I sent a message 12 March which
didn't appear, and GMail eventually informed me it couldn't deliver
it.  Two days later it told me it had been trying all along and had
given up (I emailed Alex and Scott)
- I'm borrowing Hugh's "War Story" format ... remember, imitation is
the sincerest form of flattery

This is the problem I emailed the list about, but I've now solved it.
It's slightly embarrassing, but may be helpful to others.

I have a local Samba and MPD server running on a Raspberry Pi 2.  This
has run far better than I had expected for nearly two years.  But
because of the USB2 ports on the Pi 2, data transfers to and from it
are slow.  I configured a Pi 4 as its replacement, but I couldn't
finish because after I installed MPD I couldn't access it remotely.
MPD is the "Music Player Daemon", but the only important part of this
is it offers access on port 6600.  On the Pi 2 (which has the older
GUI Raspbian on it), I just installed MPD and it worked - according to
my notes.  I didn't bother with a firewall.

On the new machine (Pi 4 with USB3 and Raspberry Pi OS Lite) with MPD
running, I could access MPD locally with:

$ mpc status
volume: n/a   repeat: off   random: off   single: off   consume: off

'mpc' is the command line client for MPD, and the response you see
above is the correct one when MPD is idle.  But as soon as I changed
to pseudo-remote, or actually remote:

$ mpc --host=pi44.local status
mpd error: Connection refused

That's without a firewall.  I thought perhaps the new Raspberry Pi OS
Lite might have helpfully provided some iptables/netfilter rules, and
I figured UFW would help:

# apt install ufw
# ufw enable
# ufw allow 6600

This changed nothing about the behaviour of 'mpc' (although it did
break Samba ... fixed by opening ports 137, 138, 139, and 445).  I
dragged out 'telnet' to connect to the local port: all that did was
prove that I can access MPD locally but not remotely (MPD responds "OK
MPD 0.21.4" when you're able to connect).

 -

Finally I figured it out.  When I told a co-worker, he got it right
away, so I suspect many of you will have too.  I said above that it
"just worked" on the original Pi, but I suspect that's not true.  I
take notes when I install software, but I failed to note one change
I'm fairly sure I made ...

What I finally remembered is that almost any service can be configured
to listen to "localhost" only.  Remembering that, I found the single
critical line in /etc/mpd.conf:

bind_to_address  "localhost"

I compared that to the older Pi, which says:

bind_to_address  "0.0.0.0"

For those not aware, "0.0.0.0" is short for "listen on all
interfaces," while "localhost" reasonably enough means "listen only to
connections coming from the local machine."  This is a common thing in
servers and I should have remembered sooner.  Restarted MPD and all is
good.  At least I have a firewall now ...

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Unanswered Question, WAS: Raspberry Pi

[Giles Orr via talk]

On Sat, 6 Mar 2021 at 07:57, Chris Aitken via talk  wrote:
> Am I on the wrong list for posing basic questions about Linux? I posted a 
> question a few weeks ago about about not being able to copy files because of 
> permissions and because filename had a '?' in it. I received no response. I 
> understand everyone is a volunteer. I'm not complaining - just want to know 
> if this is an appropriate forum for basic questions. I know it's wrong to 
> piggyback on this thread, but it is getting a lot of traffic, so I had no 
> choice. Do I keep posting, and just see which posts are of interest to 
> people, and which questions are not? Sorry, I'm not one to take a hint or 
> "know" when to leave a party. I need someone to say, "Chris, leave the  
> party." :)
> Chris

Do try to avoid hijacking threads.  However, I think your question
went unanswered because no one ever saw it.  I keep all the TLUG
messages, and I see no such question from you (unless you posted under
a different name: I searched for "Chris Aitken").  This may have been
a technical problem on your end, or it may have been a technical
problem with the TLUG server: as you may have noticed from some of the
messages, the server has had a rough few months.

We're not even "volunteers:" we're a group of people interested in
Linux who have joined a mailing list.  We answer questions if we know
the answer, we have time, and we're interested.  That means it can be
hit-or-miss.  And like a lot of mailing lists, people may get mildly
annoyed if you ask a question that can be easily answered by Google so
I recommend doing your own research first before turning here.

With all that said, clearly this latest message from you has got
through so you should be good to try reposting the question.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Raspberry Pi

[Giles Orr via talk]

On Fri, 5 Mar 2021 at 19:11, Kevin Cozens via talk  wrote:
>
> On 2021-03-05 5:57 p.m., Aruna Hewapathirane via talk wrote:
> > Understood. I just want to try and build the smallest lightest kernel. Like
> > puppy linux or knoppix. I use Debian at home so this is good news.
>
> Other options for a small distro are DSL (Damn Small Linux) or Morphix.

While I agree that Damn Small was pretty good ... it was discontinued
nine years ago.  I know this because I was a fan back in the day, and
was very sorry to see it go.  DistroWatch tells me that Morphix was
discontinued in 2003.  I'm not sure either of these are good choices
...

Maybe try DistroWatch set to "Old Computers?":

https://distrowatch.com/search.php?category=Old+Computers#simple

Their first recommendation is Puppy, which is hardly surprising.  I
wasn't much of a fan.  Eighth on the list is Tiny Core.  Tiny Core is
 interesting, without doubt.  And light, again without doubt.  But
a real PITA because of the way it stores stuff.

If Raspberry Pi OS is in the running, I'd go with that.  I find it a
decent lightweight OS that acts a lot like other OSes that people are
used to.  This cannot be said of Tiny Core, and Puppy is a bit out
there too.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Raspberry Pi

[Giles Orr via talk]

Hi Aruna.

On Wed, 3 Mar 2021 at 21:56, Aruna Hewapathirane via talk
 wrote:
>
> Hi Mat, many thanks and the more ram it has on board the better. This is 
> strictly
> for my own experimenting and learn by doing stuff. So any Pi will do to start 
> off with I guess.
>
> I am hoping to go down the embedded linux rabbit hole. I have been messing 
> around
> with arduino for a while now and  I guess it is now time to move on to 
> something a little
> easier to compile and test a linux kernel on :-)
>
> I am also very interested in seeing if a Pi can replace my ancient desktop. I 
> simply can't
> afford the Pi-4 desktop version with the dual monitor setup so thought I will 
> ask and see if
> anyone has a spare.
>
> On Wed, Mar 3, 2021 at 9:38 PM Matt Price  wrote:
>>
>> I might have one, what generation do you need and what's the application?
>>
>> On Wed., Mar. 3, 2021, 9:06 p.m. Aruna Hewapathirane via talk, 
>>  wrote:
>>>
>>> Hello,
>>>
>>> Does anyone have a Raspberry Pi lying around you no longer need ?
>>>
>>> Thanks - Aruna

The Pi 2 and Pi 3 come with 1G of memory, end of story.  Don't use a
Pi 1 - there are too many issues and they (some or all, can't
remember) don't have 1G of memory.  The only ones with more than 1G
are the recent Pi 4 series - and the ones with more than 1G all cost
more money.  The Pi 4 is also the only series with USB3.  They're the
newest and the best and as such, people aren't likely to be giving
them away yet(?).

I'm using a Pi 4 with 8G as a secondary desktop, and have been finding
it quite useful.  Mind you, I'm not using it for Gimp photo editing -
but I do use it for web browsing (not a lightweight activity) and it
handles that well.  I think you would find any of the 1G models
unsatisfactory as a desktop replacement: I wouldn't think about it
until it had 4G RAM, and I'd prefer the 8G.

If you want to learn about using a Pi - go to it.  But if you want a
desktop replacement ... I'm afraid you need the one you've said you
can't afford.  Sorry.

Something worth noting is that you need a good USB power supply.  Most
USB chargers _don't_ cut it: the Pi will be constantly telling you
it's undervoltage.  That's a long conversation of itself, but the main
point is you may have another cost on your hands.  And of course
you'll need a microSD card for the OS.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] micro controllers

[Giles Orr via talk]

On Tue, 16 Feb 2021 at 11:23, o1bigtenor via talk  wrote:
> I'm wanting to use micro controllers in more of an industrial setting.
> Don't want to spend the $$$ to get the officially hardened etc
> etc etc models.
>
> Looking for information - - - - - there's piles on using Arduino,
> Raspberry Pi but they're not really developed for serious use (one
> point - - - - almost always have to add a cape or a shield to get real
> connectivity - - - - don't need the the hp waste that's RPi so I"d
> rather not use that . . .   .)
> Trying to find other ucontroller boards - - - - can find plenty of SoB
> - - - - but not boards.
> Hoping to use open source tools for the programming and control.
>
> Any suggestions as to where to look for this kind of stuff?
>
> Am I wasting my time? (I hope not!!)

I think when you said "Raspberry Pi" above you were referring to the
SBC, but they have very recently issued their own microcontroller
board, "the Raspberry Pi Pico:"

https://www.raspberrypi.org/documentation/pico/getting-started/

In fact they don't seem to have shipped yet, but they're going to be
selling for $5.25CA each.

I mention it as an alternative to the Arduinos, with the caveat that I
know very little about microcontrollers.  I own an Arduino Teensy, and
have done a little programming on it, but that's all.  And my
understanding is that the reason the Arduino made such an impression
on the market is because it's a lot easier to program than the older
microcontroller boards that existed before it.  And I suspect that
initially, the Pi Pico will suffer from much the same problem.  I say
"initially," because the Pi itself has been so successful that it may
drive sales of the Pico which would in turn lead to development of the
coding environment for the Pico.

Anyway, I thought I'd put that out there in case the Pico and/or its
MicroPython (or C/C++) development environment sounds good.

(This is of course NOT industrial grade.)

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] HP Z420 + rx5600 + MSI MAG272CQR ?

[Giles Orr via talk]

On Tue, 9 Feb 2021 at 00:06, D. Hugh Redelmeier via talk
 wrote:
>
> | From: steve--- via talk 
>
> | Subject: Re: [GTALUG] HP Z420 + rx5600 + MSI MAG272CQR ?
>
> Does HP support linux on its workstations?  That would be nice.
>
> For some reason AMD video support has seemed slow to settle.  Is the
> rx5600 support solid?
>
> I've never needed that much video horsepower.  But plenty of people
> do.
>
> | >  Just looking at some stuff.. looking to replace my DELL Vostro
> | > 220S 4G machine with
> | > something.
>
> I take it that it has an Intel Core 2 Duo processor.  Yeah, it's
> probably worth upgrading.  On the other hand, if you've lived with
> that, an middle-of-the-road current machine should be fine.
>
> The z420 is probably expensive overkill.  Is it old?  I saw a review
> from 2013.  Are you buying used?
>
> If you don't want to fuss, I kind of like the Lenovo ThinkCentre M75q
> Tiny computer (but I am having a problem with the WiFi 6 driver):
> 
> Or if you wish to choose things (like DVD burner)
> 
> This second thread mentions a deal for the m75q, a Small Form Factor
> variant that allows more expansion.
>
> The m75q has all mod cons.  Since you're old too, you probably
> remember that term.  For others: all modern conveniences.  WiFi6, some
> magic USB-c stuff, HDMI (dongle, supplied) USB-3.x, 1G ethernet,
> support for gobs of RAM.  NVMe.  I think that it has a 2.5" bay but
> I'm too lazy to check.
>
> Note that HP z series and Lenovo ThinkCentre are not cheap consumer
> junk.  (I don't mind cheap consumer junk.)
>
>
> Monitors are very personal.  For me, a decent but inexpensive 40"
> UltraHD TV beats a 27" monitor every time.  Gamers would disagree
> because the refresh rates can be better on monitors.
>
> | > Cant go to Movies anyMore.
> | >
> | > GoTa bring the Movie to Me.
>
> You will want quiet, then.  Not guaranteed with a workstation.
>
> You also need to take a little care with the video, depending on your
> monitor (the rx5600 would be fine).
>
> For our household, a separate, non-computer solution is better for
> watching movies.  Example device: a $65 Google ChromeCast with Google
> TV.  We do have a dedicated tiny PC connected to our TV for viewing
> the few things that our Android TV device does not support.
>
> |Ooops forgot Ubuntu 18.04 Bionic Beaver.
> |
> | This is my last upgrade ever, and the perfect place to stop.
>
> 18.04 is already 2 years stale.  Surely 20.04 LTS would be a wiser
> choice.  And more likely to support the rx5600 (just a guess).
>
> And don't say never.  Not a good way to think about computers.  They
> are kind of disposable (not quite as badly as mobile phones).
>
> Planning for the distant future is hard.  Perfection is the enemy of
> getting things done.  I know.

Going to add a few "me too" votes to a couple things Hugh said.

Are you a gamer?  Because if you're not, the card and the monitor
sound like overkill - expensive and not necessary for your declared
use case (movies).  Keep in mind that retro-gaming doesn't count as
"gaming" in this context: retro-games don't require a high-end video
card.  I haven't purchased a specialty video card in 15 years, and I
watch movies frequently on my built-in video cards - works fine up to
1080p, although I admit I see tearing and problems at 4K.  The
built-ins are also usually a lot easier to get running in Linux for a
lower cost.

However, like Hugh, I do most of my video watching either on a first
gen Apple TV (I got it free ...) that acts as a Netflix box, or on an
Asus O!Play Mini, both of which are hooked to my TV.

I'd also like to back Hugh's recommendation of a larger and cheaper
(or perhaps equally priced) 40" monitor over the 27", unless your
eyesight is a lot better than mine.  I'm finding larger monitors to be
a huge blessing as I get older, and the properties a gamer monitor
might provide aren't usually of any use to me.


--
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] grub2 on Fedora war story

[Giles Orr via talk]

On Mon, 8 Feb 2021 at 16:01, D. Hugh Redelmeier via talk
 wrote:
>
> Anything that can get more complicated will get more complicated.  Boot
> loaders seem to be an example.
>
> It used to be straightforward to read /boot/efi/EFI/fedora/grub.cfg.
> Now, building the list of kernels for the menu is farmed out to blscfg (a
> grub module).
>
> I needed to have a Fedora box default to booting a kernel that isn't the
> latest (because the latest cannot bring up the display on my computer).
>
> 1. I needed to make updates not delete the working kernel.  Normally
> updates keep only the last three kernels.  As of today, two are duds.
> Fix: change /etc/dnf/dnf.conf's installonly_limit from 3 to 0
>
> 2. Find the list of kernels known to grub:
> sudo ls /boot/loader/entries/*.conf
>
> 3. set the default to one of those.  Use the filename, without the
> directory and without the .conf
>   sudo grub2-set-default 
> 2733f1c892a5422c98bdb188c4f62737-5.10.9-201.fc33.x86_64
>
> I don't know how long this sticks.

A couple comments:

- I think raw GRUB still works from a single .conf file that can be
edited by hand.  The problem you're seeing is that most distros have
built out a complex system to construct and replace that .conf file
whenever a new kernel arrives.  The assumption is of course that mere
mortals shouldn't be touching that config.  I learned a lot about
GRUB2 configuration at one time, but I don't mess with Fedora's GRUB
config system.
- Nicholas mentioned the in-place replacement of kernels: I think(?)
I've seen that behaviour on Debian with same-version kernels with
security updates, but I don't think I've ever seen it for any reason
on Fedora.
- if you have a separate /boot/ partition (very likely these days) and
you're never deleting kernels, you stand a good chance of over-filling
that partition and getting into trouble, particularly on Fedora which
likes to push a new kernel every couple weeks.  You'll want to keep an
eye on the space remaining in /boot/ and delete some of the newer
kernels you don't need by hand.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] YAD-Zenity-GTK on Pi - assistance needed

[Giles Orr via talk]

On Tue, 11 Aug 2020 at 20:47, Giles Orr  wrote:
> YAD ("Yet Another Dialog") is a fork of Zenity - essentially a way to
> display GTK+ dialogues from shell scripts.  I've chosen YAD because it
> supports a couple things Zenity doesn't have, namely form buttons and
> tabs.  I've created an ugly but functional interface that's meant to
> go on a 3.5" touch screen on a Raspberry Pi.  But here's the problem:
> by the time it gets on a screen that small, the buttons are too small
> to reliably poke with a finger.  So I want to increase the font size.
> All that was needed to fix that on my Fedora desktop was:
>
> # file: ~/.configs/gtk-3.0/settings.ini
> [Settings]
> gtk-font-name = Sans 32
>
> And voila, 'yad' appears with a super-huge font.
>
> However - this behaviour isn't replicated on the Pi.  It's an old(ish)
> Debian stretch install on an old Pi B (not Pi 2 B, just "Pi B").  I'm
> pretty sure that ~/.configs/gtk-3.0/settings.ini is the right file,
> because when I mis-configured it, 'yad' complained loudly.  But it
> ignores the font setting that works fine on another host.  I tried
> 'yad --font' which throws up what amounts to a GTK font selection
> dialogue, and "Sans 32" is a valid font on the Pi.
>
> Anyone got any ideas on this?

I'm reviving this thread because I think I understand it a bit better
and the question is a bit different.  I've re-installed the OS
(Raspberry Pi 10 instead of Raspbian 9 last time) and the touch screen
drivers.  The behaviour is precisely the same.  I've also confirmed
that the problem isn't with Debian: font sizes increase as expected on
Debian.  But what I've noticed is this: the appearance of YAD on
Fedora and Debian is essentially identical, but on Raspberry Pi OS the
appearance of the tabs are different, and the buttons - which have
rounded corners on Fedora and Debian and occasionally icons - have
square corners and no icons.

What this suggests to me (I'm guessing here, hoping someone can help
me figure it out) is that YAD is compiled with GTK support on Debian
and Fedora, but some other graphical library on the Pi.  Although - as
I established above, YAD still reads the GTK3 configuration file on
the Pi - it just doesn't respect it.  How can I figure out if this
guess is correct?  Is recompiling YAD with GTK3 support sane or
possible on a Pi (if I'm correct)?  Is there another configuration
file that might affect the font size?

Buying a larger touch screen is a possibility, but the 3.5" screen
would be perfect for what I'm trying to achieve so I'd rather not.
(Cost isn't a problem ... but obstinance might be.)

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Odd VLC problem

[Giles Orr via talk]

On Mon, 25 Jan 2021 at 02:23, Evan Leibovitch via talk  wrote:
> Last week, in an attempt to make my desktop more readable to my ageing eyes, 
> I bought a $40 stand and set up my monitors vertically (see 1st pic).
>
> Changing the KDE setup to accommodate this (Kubuntu 20.04 LTS) was easy, and 
> all of my apps seem to work fine with the new setup ... except VLC.
>
> VLC no longer works the way it used to.
>
> Videos will now play only if I decouple the video window from the VLC 
> interface ("Integrate video in interface" must be unchecked). And now if I 
> try to resize the interface, it visually "stutters" (see second photo) and 
> remains stuttered in this form until I resize the window again.
>
> I have tried multiple options for video output with no change in behaviour. 
> Again, the videos display fine but only if I separate the video window from 
> the interface window.  Resizing the video window works fine. Also, videos 
> displayed inside the browser (ie Plex and YouTube) work fine.
>
> All suggestions are appreciated.

Hi Evan.

I apologize for not directly addressing the problem you're asking
about, but I'll offer a couple possible solutions.

If you're not tied to 'vlc', try 'mpv' as a movie player.  I have both
installed and used vlc as my main video program for many years, but
now mostly lean to mpv.  Can't promise it won't have issues with the
layout, but I think it's worth a try.

The "aging eyes" issue is one I'm familiar with: my desktop now has a
43"(!) primary monitor that's flanked by a couple vertical monitors
(those are different computers, and not always on).  Re-arranging the
login screens to be vertical can be a real PITA depending on which
login manager it is, otherwise it works quite well.  If money (and
desk real estate) aren't an issue, I would recommend seeing if you can
find a big-ass monitor in the 40" range - I love that thing.  Your
eyes will thank you.  I was lucky though: it's actually a "monitor"
rather than a repurposed TV, it's 4K, and I managed to get it for
under $600 a couple years ago.  (Looking at current prices of 40"+
monitors, I'm thinking I should be using the phrase "spectacularly
lucky.")  That's not a popular market segment, so such items are thin
on the ground and/or quite expensive.  If you're okay with a TV (I'm
not clear on the disadvantages - distorted colour palettes, high
contrast, limited ports?), those are available at lower prices.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] How do I demonstrate a crappy network?

[Giles Orr via talk]

On Tue, 19 Jan 2021 at 16:21, David Collier-Brown via talk
 wrote:
>
> On 2021-01-19 11:15 a.m., David Collier-Brown wrote:
> > On 2021-01-19 9:56 a.m., Stewart C. Russell via talk wrote:
> >
> >> On 2021-01-19 7:19 a.m., David Collier-Brown via talk wrote:
> >>> What is something that I can make fail, either audibly or visually?
> >>> Showing my grandmother ping statistics isn't going to work (;-))
> >> Video call over 2.4 GHz wifi, then start the microwave. Glitchtastic
> >> every time.
> >
> > I'm not sure the Linux networking changes help with that (;-))
> >
> > --dave
>
> Joking aside, is there a famously _bad_ video or audio site, one that
> everyone hates?
>
> Or a setting for a service that makes it fail?

Here's an idea - no clue if it's a good one.

Get yourself access to a low bandwidth connection: if the following
table is correct (dubious source and video compression always varies
depending on content) then you can easily saturate a 10Mbps DSL line
with a 4K video:

Resolutions Required   Bandwidth Required
  H264H265
1280×720(HD)  3Mbps1.5Mbps
1920X1080(FHD) 6Mbps 3Mbps
3840×2160 (UHD)   25Mbps   12Mbps
4096×2160 (4K)   32Mbps   15Mbps

(source: https://www.synopi.com/bandwidth-required-for-hd-fhd-4k-video/
in case the table got mangled into unreadability. )

Maybe you have a way to fractionalize your own network connection if
you don't have access to 10/1 DSL.

With YouTube, you can select the resolution (although it's not very
granular at the top end).  If you can toggle CAKE on/off, see if the
saturated line becomes unsaturated.  If not, reduce the video
resolution and repeat.  YouTube's "Stats for Nerds" may also be
helpful?

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Rust Frontend Funding in GCC News

[Giles Orr via talk]

On Wed, 13 Jan 2021 at 11:01, D. Hugh Redelmeier via talk
 wrote:
>
> | From: D. Joe via talk 
>
>
> | thanks for the link. typo in the URL, though. this seems to be it:
> |
> | https://opensrcsec.com/open_source_security_announces_rust_gcc_funding
>
> I noticed they type and tried that, but my firefox disliked that site:
>
> Secure Connection Failed
>
> An error occurred during a connection to opensrcsec.com. Unsupported
> hash algorithm used by TLS peer.
>
> Error code: SSL_ERROR_UNSUPPORTED_HASH_ALGORITHM
>
> The page you are trying to view cannot be shown because the
> authenticity of the received data could not be verified.
> Please contact the website owners to inform them of this problem.

There's always a certain amount of comedy in security-related sites
that get blocked based on security problems.  This site fails to load
in FF on Linux - but works fine in FF, Chrome and Safari on Mac.  It
also works in Chromium on Linux.

It doesn't appear to be a TLS version issue as I first guessed ... if
only I'd read the "error code" I would have known better, but TLS
version issues are commonest.  Sorry, I guess I don't have much to
add, but site encryption issues like this always interest me as I
spent a fair bit of time learning about site encryption ... and this
is still beyond me.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Anyone familiar with SMART?

[Giles Orr via talk]

Now that our list is back (Hallelujah, thank you Hugh and TLUG), I'm
resending this message I originally sent on the 23rd as I haven't yet
found an answer.

On Wed, 23 Dec 2020 at 15:18, Giles Orr  wrote:
>
> I just bought two new, identical, external USB hard drives.  While
> I've used SMART on and off over the years, I think this is the first
> time I used it immediately on connecting a drive.  I'm using
> 'gsmartcontrol', a GUI that I'm liking.  In the "Statistics" tab I'm
> seeing, under "Rotating Media Statistics (rev 1)", "Number of
> Mechanical Start Failures" = "1".  This is straight out of the box,
> brand new.  Where this gets really weird is that both of the drives
> show exactly the same error.  These are WD easystore drives.  So ...
> is this something I should be worried about?

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] "dd" in Windows?

[Giles Orr via talk]

On Mon, 14 Dec 2020 at 16:24, William Park via talk  wrote:
>
> How do you write "image" file to USB stick in Windows 10?
>
> In Linux, I would do 'dd', but it has to be Windows.  'dd' is available
> in Git for Windows, Windows Subsystem for Linux, BusyBox for Windows,
> and Ubuntu in Hyper-V VM.  But, I can't get them to access USB drive.

This might be useful - or it might not.  There was some mention of
image writers and Windows.

https://alternativeto.net/software/dd/

They mention Rufus, and so does this:

https://www.lifewire.com/how-to-burn-an-iso-file-to-a-usb-drive-2619270

lifewire is usually pretty reliable - although their explanation
looked pretty long (didn't read it).

I hope this helps.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] information storage ideas

[Giles Orr via talk]

On Tue, 8 Dec 2020 at 12:41, o1bigtenor via talk  wrote:
> I tend to work on quite a number of different things  if not at the
> same time then in
> quite short order. So far most projects will get some notes or phone
> call references
> or other information jotted down on paper. Over time this means that I
> all too often
> tend to redo things - - - sometimes to improvement but sometimes I
> don't know where
> the previous work is so I'm looking or I'm redoing.
>
> So I'm looking at collecting things like contact information (and
> their value/area etc etc),
> project ideas, info sources, project planning, project design
> parameters, project
> components all of which hopefully results in some in the end.
>
> Have been trying to use taskwarrior and its a decent reminder system but the
> storage of all the other 'stuff' isn't there. Been trying to just save
> things into a folder
> (that's not so useful when information is applicable to multiple projects).
>
> Has anyone found a 'reasonable' system that would effect this less than simple
> 'idea'?

If you're a Vim user, I highly recommend vimwiki:
https://github.com/vimwiki/vimwiki .  I live and die by NeoVim, and
spend every day with multiple sessions open across multiple machines.
I store my vimwiki in git, which allows me to sync it across all those
machines while managing potential data collisions.  I use a number of
Vim plugins, but vimwiki is probably the one I use the most ... unless
you count gitgutter.  :-)

If you're not a Vim user, I imagine there's an equivalent for Emacs.
If not Emacs, there are many, many personal wikis.  I've used and
liked the JavaScript-and-browser-based Tiddlywiki (although I haven't
touched it in years, so not sure of its current status -
https://tiddlywiki.com/ ).  It had the interesting property of all
being stored entirely in one file, but easily searchable and
displaying in bite-sized chunks as if it were many wiki pages.

Using a wiki and its associated mark-up language and commands takes
some time to adjust to, but it sounds like a good way to address the
problem you're outlining so it would probably be worth it?  And while
git is annoying, it's a great way to sync data across machines without
data collisions.

I hope this helps.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] SSL Certs for both web and email servers

[Giles Orr via talk]

On Thu, 3 Dec 2020 at 14:14, William Witteman via talk  wrote:
>
> Thanks for your help!
>
> I have not yet set the redirects from http -> https, but the result
> was achieved *much* more easily than I would have expected.
>
> I have an existing cert that I set up as a standalone, which dovecot
> has been using happily for a few years.  I did not know that I could
> expand what that cert covers, but the good people at EFF have made
> this very easy.
>
> I used this command and it Just Worked(TM):
>
> sudo certbot -d
> comma,separated,list,of,each,domain,and,subdomain,including,the,ones,already,in,place
> --expand
>
> And after a moment, all of my domains and subdomains are under the
> single umbrella that I already had.
>
> Note that the above list includes three different domains and a half
> dozen subdomains, all of which seem to just work now.
>
> Thanks again!

Heh - you top-posted, and I'm too lazy to fix the sequencing so now we
have mixed post order ... oh well!

If you're new to 'certbot' I recommend that you watch very closely as
you approach the three month mark.  Let's Encrypt's certs are only for
three months.  But if you've installed it on Debian, you should also
find that you have a twice-daily cron job that attempts to renew all
issued certificates automatically.  It will fail silently up until the
two month mark (don't quote me on this, I think it's two months), and
then just as quietly renew and replace the certs when it does.  It's a
lovely system, but you'll want to check maybe two weeks before they
need renewal.  If they haven't renewed automatically you'll need to
take a closer look at your system(s) to see what went wrong.

I wrote a shell script that takes a list of domain names as input,
then grabs the certificate for each domain and lists the names and
expiries for each.  It uses the 'openssl' command to extract the
expiry date, and a bit of date magic to determine if any are expiring
in less than a month and then highlights those.  I run it weekly
against my business's sites (that list is programmatically generated
too), and it's saved my ass a few times ...

> On Tue, 1 Dec 2020 at 06:37, ac via talk  wrote:
> >
> > On Tue, 1 Dec 2020 03:34:06 -0500
> > John Sellens via talk  wrote:
> > > On Tue, 2020/12/01 08:16:49AM +0200, ac via talk 
> > > wrote: | > I have three domains and a small but invariant number of
> > > subdomains | > that I want to encrypt - should I try to pull them all
> > > under one SSL | > cert, or do one for each domain, or one for every
> > > subdomain?  I don't | > need a wildcard, but I would like something
> > > relatively painless if | > possible.
> > > |
> > > | yes, in your case, and for painless and easy, just use the domain
> > > name | and one cert. so, instead of mail.example.com and
> > > www.example.com | - just use example.com.
> > >
> > > I think that might cause client complaints in some cases.
> > >
> > imho i do not think with three domains this will be an issue.
> >
> > what is the point of having mail.example.com if the IP number for
> > mail.example.com is the same as example.com ? the same can be asked
> > about imap.example.com and pop.example.com etc.
> >
> > This is just wasteful and increases the risk of issues, ads complexity
> > and does not serve any "real" technical, logical or functional purpose.
> >
> > The reason why mail.example.com used to be prevalent - pre container -
> > was because mail.example.com - was at a different IP number / different 
> > network
> > even...
> >
> > And, actually even if you had 100 domains on one server: reducing
> > complexity, reducing the amount of DNS lookups and reducing pebcac,
> > reducing comms, reducing traffic, reducing load and reducing wastage -
> > means:
> >
> > You are making it easier for clients
> >
> > And : You are even saving cycles, saving electricity, saving network
> > traffic and TOOOTEROOO:
> >
> > Saving the planet
> >
> > in case you did not know: In 2020 - 2030 - we will still get the vast
> > majority of our power from non sustainable fossil sources. so, we
> > should all try to be less wasteful, mind you, now with Alaska being
> > strip mined and auction sold, the planet has a lot more to waste.
> >
> > > I think letsencrypt now provides wildcard certifications, but you
> > > can use mutliple -d options when creating or updating a certificate
> > > e.g.
> > >
> > >   certbot certonly \
> > > --non-interactive \
> > > --expand \
> > > --webroot \
> > > -w /var/www/html/letsencrypt \
> > > --cert-name www.example.com \
> > > -d example.com \
> > > -d mail.example.com \
> > > -d blog.example.com
> > > And then the one certificate is valid for all those names.
> > >
> > a small number of invariant sub domains usually means
> > www.example.com, pop.example.com, mail.example.com,
> > imap.example.com and in this case - x3 domains
> >
> > but, one could also wildcard (*) just simply -d *.example.com and add
> > _acme-challenge TXT record to 

Re: [GTALUG] Network issues with github

[Giles Orr via talk]

On Sun, 29 Nov 2020 at 22:59, Alex Beamish  wrote:
>
> On Sat, Nov 28, 2020 at 11:19 PM Giles Orr via talk  wrote:
>>
>> Hi Alex.
>>
>> On Sat, 28 Nov 2020 at 16:50, Alex Beamish via talk  wrote:
>> >
>> > Hi All,
>> >
>> > This is probably a blindingly obvious question, but I'm a little stumped. 
>> > I've done a little work for local business, setting up a Linux server 
>> > (Ubuntu), developing some code and pushing it to github. It's all worked 
>> > wonderfully until a few weeks ago, when he had someone in to do something 
>> > to the network. Since then, Things Are Broken in ways that I don't 
>> > understand.
>> >
>> > When I try to do anything with github, I see the response
>> >
>> >   Received disconnect from 140.82.113.3 port 22:2: Connection blocked 
>> > because server only allows public key authentication. Please contact your 
>> > network administrator.
>> >
>> > Because I was worried I'd borked my account, this afternoon I tried again, 
>> > creating a brand-new account and ssh-ing in .. and still got the same 
>> > result.
>> >
>> > My github account works fine from my own machine, and also from my web 
>> > provider (pair.com), so I'm guessing there's something going on within my 
>> > client's network. Suggestions gratefully received.
>>
>> I apologize if this is something you've already looked at, but the #1
>> Google hit for "Connection blocked because server only allows public
>> key authentication" does look relevant:
>>
>> https://superuser.com/questions/1466177/connection-blocked-because-server-only-allows-public-key-authentication-putty-f
>
>
> Giles, Hugh,
>
> Thank you both for your responses. I am beginning to suspect that there is 
> some network thing that's breaking ssh.
>
> From my own machine, the result of ssh -vT g...@github.com looks like this: 
> it works fine.
>
> OpenSSH_8.2p1 Ubuntu-4ubuntu0.1, OpenSSL 1.1.1f  31 Mar 2020
> debug1: Reading configuration data /home/tab/.ssh/config
> debug1: /home/tab/.ssh/config line 22: Applying options for *
> debug1: /home/tab/.ssh/config line 338: Applying options for *
> debug1: /home/tab/.ssh/config line 339: Deprecated option "useroaming"
> debug1: Reading configuration data /etc/ssh/ssh_config
> debug1: /etc/ssh/ssh_config line 19: include /etc/ssh/ssh_config.d/*.conf 
> matched no files
> debug1: /etc/ssh/ssh_config line 21: Applying options for *
> debug1: Connecting to github.com [140.82.113.3] port 22.
> debug1: Connection established.
> debug1: identity file /home/tab/.ssh/id_rsa type -1
> debug1: identity file /home/tab/.ssh/id_rsa-cert type -1
> debug1: identity file /home/tab/.ssh/id_dsa type -1
> debug1: identity file /home/tab/.ssh/id_dsa-cert type -1
> debug1: identity file /home/tab/.ssh/id_ecdsa type -1
> debug1: identity file /home/tab/.ssh/id_ecdsa-cert type -1
> debug1: identity file /home/tab/.ssh/id_ecdsa_sk type -1
> debug1: identity file /home/tab/.ssh/id_ecdsa_sk-cert type -1
> debug1: identity file /home/tab/.ssh/id_ed25519 type -1
> debug1: identity file /home/tab/.ssh/id_ed25519-cert type -1
> debug1: identity file /home/tab/.ssh/id_ed25519_sk type -1
> debug1: identity file /home/tab/.ssh/id_ed25519_sk-cert type -1
> debug1: identity file /home/tab/.ssh/id_xmss type -1
> debug1: identity file /home/tab/.ssh/id_xmss-cert type -1
> debug1: Local version string SSH-2.0-OpenSSH_8.2p1 Ubuntu-4ubuntu0.1
> debug1: Remote protocol version 2.0, remote software version babeld-b85a2946
> debug1: no match: babeld-b85a2946
> debug1: Authenticating to github.com:22 as 'git'
> debug1: SSH2_MSG_KEXINIT sent
> debug1: SSH2_MSG_KEXINIT received
> debug1: kex: algorithm: curve25519-sha256
> debug1: kex: host key algorithm: rsa-sha2-512
> debug1: kex: server->client cipher: chacha20-poly1...@openssh.com MAC: 
>  compression: none
> debug1: kex: client->server cipher: chacha20-poly1...@openssh.com MAC: 
>  compression: none
> debug1: expecting SSH2_MSG_KEX_ECDH_REPLY
> debug1: Server host key: ssh-rsa 
> SHA256:nThbg6kXUpJWGl7E1IGOCspRomTxdCARLviKw6E5SY8
> debug1: Host 'github.com' is known and matches the RSA host key.
> debug1: Found key in /home/tab/.ssh/known_hosts:3
> debug1: rekey out after 134217728 blocks
> debug1: SSH2_MSG_NEWKEYS sent
> debug1: expecting SSH2_MSG_NEWKEYS
> debug1: SSH2_MSG_NEWKEYS received
> debug1: rekey in after 134217728 blocks
> debug1: Will attempt key: /home/tab/.ssh/music2012 RSA 
> SHA256:JzHBQSQHReaDXiXIEO4W3QtW/cqqoab6xuWt2V4eP30 agent
> debug1: Will attempt key: /home/tab/.ssh/id_rsa
> debug1: Will attemp

Re: [GTALUG] Network issues with github

[Giles Orr via talk]

Hi Alex.

On Sat, 28 Nov 2020 at 16:50, Alex Beamish via talk  wrote:
>
> Hi All,
>
> This is probably a blindingly obvious question, but I'm a little stumped. 
> I've done a little work for local business, setting up a Linux server 
> (Ubuntu), developing some code and pushing it to github. It's all worked 
> wonderfully until a few weeks ago, when he had someone in to do something to 
> the network. Since then, Things Are Broken in ways that I don't understand.
>
> When I try to do anything with github, I see the response
>
>   Received disconnect from 140.82.113.3 port 22:2: Connection blocked because 
> server only allows public key authentication. Please contact your network 
> administrator.
>
> Because I was worried I'd borked my account, this afternoon I tried again, 
> creating a brand-new account and ssh-ing in .. and still got the same result.
>
> My github account works fine from my own machine, and also from my web 
> provider (pair.com), so I'm guessing there's something going on within my 
> client's network. Suggestions gratefully received.

I apologize if this is something you've already looked at, but the #1
Google hit for "Connection blocked because server only allows public
key authentication" does look relevant:

https://superuser.com/questions/1466177/connection-blocked-because-server-only-allows-public-key-authentication-putty-f

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Fedora 33 uses RAM for swap?

[Giles Orr via talk]

On Wed, 25 Nov 2020 at 08:18, Giles Orr  wrote:
> A few days after upgrading the last of my machines from Fedora 32 to
> 33, I noticed my main machine has acquired a new disk:
>
> NAME   SIZE FSTYPE  LABELMOUNTPOINT
> zram04G  [SWAP]
>
> I didn't set that up, and I don't think it was there on F32.  So the
> OS has, without asking, co-opted 1/4 of my 16G of RAM to use as swap
> space.  This system has an SSD, so when I initially set it up (Fedora
> 27), I made a conscious decision to go without swap space.  I rarely
> push the limits of 16G.
>
> But now I'm in the situation that I have only 12G of RAM, so the
> system will become memory-starved earlier ... and what will it do?  It
> will go to swap.  Which is RAM anyway.  How does this help?  To me
> this seems like adding complexity without adding utility.
>
> Can someone please explain A) if I'm correct about this behaviour in
> the first place, and B) why it's useful?  Thanks.

Hugh, this machine was an upgrade from Fedora 32.  I guess the
decision was made to go with the new default rather than ask questions
during the upgrade.  Thanks Hugh, Mauro, Dave for the info - and my
apologies, I did zero research before turning to the list.  Not my
usual behaviour, I promise - I guess I was peeved because it was so
unexpected.

I assumed right out the gate that the statement "4G" meant it was
using 4G of RAM.  As has been pointed out, it's not allocated until
it's used, and it's compressed - so even when full it'll usually only
use 2G of RAM.  After having read about it some, it sounds like it's
mostly a win.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] First new home computer for decades - Raspberry Pi 400

[Giles Orr via talk]

On Wed, 11 Nov 2020 at 09:10, James Knott via talk  wrote:
>
> On 2020-11-10 9:56 p.m., Lennart Sorensen via talk wrote:
> > My second thought was: They removed the composite port.  I probably
> > would never personally use it, but I could imagine there are some people
> > that like attaching things to old CRTs for gaming that would miss it.
> > But perhaps most people like myself just don't care about analog video
> > anymore so they probably made the right call on that.
>
> A CRT monitor would have to be pretty old.  I bought my first LCD
> monitor almost 15 years ago and even then it was a refurb.  Then we get
> to wasted space and power.

And let's not forget weight: a 17" CRT was 20kg.  They're freaking
brutal to move.  All that lead shielding and a huge glass tube ...
The death of the CRT is mourned by very few indeed.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Scientific Libraries in Python for Drawing Physics Equations

[Giles Orr via talk]

On Mon, 2 Nov 2020 at 22:21, Nicholas Krause via talk  wrote:
> On 11/2/20 6:56 PM, Jamon Camisso via talk wrote:
> > On 2020-11-02 18:07, Nicholas Krause via talk wrote:
> >> Greetings,
> >> I'm wondering if anyone has used this before:
> >> http://qutip.org/tutorials.html. If someone has a recommendation or has
> >> used something
> >> similar for drawing out the graphs for Schrodinger wave functions or
> >> Quantum Field Theory that would be helpful. I'm drawing to draw them
> >> out on a computer with a library as drawing them by hand is also
> >> impossible for large scales of values and will be easier for field theory.
> >>
> >> This is for the future but it would be nice to hear if anyone has any
> >> experience with drawing out quantum physics equations from a programming
> >> library,
> >
> > I have zero experience or knowledge about it, but here's a reasonable
> > looking project that uses Seaborn (https://seaborn.pydata.org/) to
> > visualize wave functions:
> >
> > https://github.com/nnguyen19/helium_project
> >
> > The code is in here:
> > https://github.com/nnguyen19/helium_project/blob/master/Tung-Nhan%20Nguyen%20-%20Helium%20by%20wavefunction.py
> >
> > Seaborn is a very nice wrapper around matplotlib. From a cursory glance,
> > it looks like qutip is also a wrapper around pandas & matplotlib so the
> > two might be useful in combination with each other.
> >
> > Cheers, Jamon
>
> Jamon,
> Thanks I will look into that. The bigger problem isn't wave functions so
> much as quantum field theory. I'm not sure how many people about it.
> Basically its a merger of special relativity with quantum mechanics
> using fields. Not sure if its better to graph that out using a physics
> engine or just the above. If people need more information on that this
> may help:
> https://www.ribbonfarm.com/2015/08/20/qft/
> And for anyone who is asking General Relativity has issues with Bell's
> Theorem in quantum mechanics alongside other things I'm not aware
> of so its not part of the standard model. Off topic but the standard
> paper on that is this:
> https://cds.cern.ch/record/111654/files/vol1p195-200_001.pdf
> and it was proven in various labs in the 1960s if I recall my years
> correctly.
>
> Sorry for the weird question and this is my bigger concern if I
> was unclear before through thanks again James,

I have some (not much) experience with matplotlib, and have been
impressed by it.  I suspect that by itself matplotlib would be
inadequate to your task - I encountered Quantum Mechanics in
university and hated it, but that was ... decades ago, so I'm not
claiming to understand.  But it sounds like there are wrappers, and if
they're based on matplotlib I think they chose a good starting point.
Hope this helps, if only a little.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] YAD-Zenity-GTK on Pi - assistance needed

[Giles Orr via talk]

YAD ("Yet Another Dialog") is a fork of Zenity - essentially a way to
display GTK+ dialogues from shell scripts.  I've chosen YAD because it
supports a couple things Zenity doesn't have, namely form buttons and
tabs.  I've created an ugly but functional interface that's meant to
go on a 3.5" touch screen on a Raspberry Pi.  But here's the problem:
by the time it gets on a screen that small, the buttons are too small
to reliably poke with a finger.  So I want to increase the font size.
All that was needed to fix that on my Fedora desktop was:

# file: ~/.configs/gtk-3.0/settings.ini
[Settings]
gtk-font-name = Sans 32

And voila, 'yad' appears with a super-huge font.

However - this behaviour isn't replicated on the Pi.  It's an old(ish)
Debian stretch install on an old Pi B (not Pi 2 B, just "Pi B").  I'm
pretty sure that ~/.configs/gtk-3.0/settings.ini is the right file,
because when I mis-configured it, 'yad' complained loudly.  But it
ignores the font setting that works fine on another host.  I tried
'yad --font' which throws up what amounts to a GTK font selection
dialogue, and "Sans 32" is a valid font on the Pi.

Anyone got any ideas on this?

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Bash does-directory-exist question

[Giles Orr via talk]

On Sun, 12 Jul 2020 at 23:26, o1bigtenor  wrote:
> On Sun, Jul 12, 2020 at 8:48 PM Giles Orr via talk  wrote:
>> On Fri, 10 Jul 2020 at 17:51, John Sellens  wrote:
>> > On Fri, 2020/07/10 05:39:59PM -0400, Giles Orr  wrote:
>> > | I love this list!  I thought that '[ -w . ]' and '[ -w $PWD ]' were
>> > | practically equivalent.  "Practically" means, in this case, "almost."
>> > | But not quite - and the difference is the solution to the problem.
>> >
>> > It's a very important, though sometimes subtle, concept in unix-land
>> > that there are multiple names for just about anything.
>> >
>> > Here, obviously, $PWD is a variable substitution equivalent to /some/path,
>> > which likely existed at some point, but may or may not exist now.  The
>> > directory "." always (I think) exists, because a process always has a
>> > current directory open. (Hmmm, but opendir(".") might not work?)
>> >
>> > The other canonical example is "how do I remove a file that starts with 
>> > -?".
>> > The key to that of course is the multiple names thing "-file" (which looks
>> > like an option string) is the same as "./-file" (which doesn't).
>> >
>> > Once you understand that, the world opens up :-)
>> >
>> > Of course, most times "rm -- -file" works but I'm old enough (uh, I mean
>> > I've read about the history of unix) to know that -- didn't always exist.
>>
>> Here's a simple implementation of a Bash prompt using what we were 
>> discussing:
>>
>> PS1="\$(if ! [ -w "\${PWD}" ]; then echo -en '\[\033[41m\]' ; fi ;
>> echo '\w\[\033[0m\]\$ ')"
>>
> H - - - - I do believe that you suggested  " . . . simple . . .  ." - - - 
> yes?

For better or worse, that's about as simple as that kind of
implementation gets.  The if-then-fi should be fairly clear.  The '\w'
is a Bash prompt short form for "the working directory."  The super
ugly bits are colour codes.  Which you could do without, but without
them the directory name doesn't look different when the directory
becomes unwritable.  The '\033[41m' changes the background to red.
The '\033[0m' turns off all colour codes so everything after it is in
plain colours.  Both are surrounded in '\[' and '\]' to indicate to
Bash that they're non-printing characters, or more correctly that they
take up no physical space ... that's a long story I won't get into
here, suffice to say they're necessary in a prompt.

Without colour codes:

PS1="\w\$(if ! [ -w "\${PWD}" ]; then echo -en '(RO)' ; fi )\$ "

In this one - as in the last one - there's a lot of backslash-escaping
things so they don't get interpreted right away, instead being
interpreted each time the $PS1 is shown.  The '\w' and '\$' are both
special commands for the prompt, not backslash escapes.

That's the best I can do - I hope it helps.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Bash does-directory-exist question

[Giles Orr via talk]

On Fri, 10 Jul 2020 at 17:51, John Sellens  wrote:
>
> On Fri, 2020/07/10 05:39:59PM -0400, Giles Orr  wrote:
> | I love this list!  I thought that '[ -w . ]' and '[ -w $PWD ]' were
> | practically equivalent.  "Practically" means, in this case, "almost."
> | But not quite - and the difference is the solution to the problem.
>
> It's a very important, though sometimes subtle, concept in unix-land
> that there are multiple names for just about anything.
>
> Here, obviously, $PWD is a variable substitution equivalent to /some/path,
> which likely existed at some point, but may or may not exist now.  The
> directory "." always (I think) exists, because a process always has a
> current directory open. (Hmmm, but opendir(".") might not work?)
>
> The other canonical example is "how do I remove a file that starts with -?".
> The key to that of course is the multiple names thing "-file" (which looks
> like an option string) is the same as "./-file" (which doesn't).
>
> Once you understand that, the world opens up :-)
>
> Of course, most times "rm -- -file" works but I'm old enough (uh, I mean
> I've read about the history of unix) to know that -- didn't always exist.

Here's a simple implementation of a Bash prompt using what we were discussing:

PS1="\$(if ! [ -w "\${PWD}" ]; then echo -en '\[\033[41m\]' ; fi ;
echo '\w\[\033[0m\]\$ ')"

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Bash does-directory-exist question

[Giles Orr via talk]

On Fri, 10 Jul 2020 at 11:11, David Mason  wrote:
>
> (base) : ~/foo ; [ -w .. ] && echo true
> true
> (base) : ~/foo ; /bin/pwd
> pwd: .: No such file or directory
> (base) : ~/foo ; pwd
> /Users/dmason/foo
> (base) : ~/foo ; [ -w $PWD ] && echo true
> (base) : ~/foo ;
>
> So, /bin/pwd fails and [ -w $PWD ] also fails, as John hypothesized
>
> ../Dave
> On Jul 10, 2020, 11:01 AM -0400, John Sellens via talk , 
> wrote:
>
> On Fri, 2020/07/10 09:38:48AM -0400, Giles Orr via talk  
> wrote:
> | This gives immediate visual feedback on the write-status of the
> | current directory. But test's '-w' and '-d' both claim that you're
> | still in a valid directory under the above circumstances. Does anyone
> | know of a simple way to find out if the directory you're currently in
> | actually exists?
>
> The directory "." will still exist while you have it open (your current
> directory), but will be unreachable, as you observed with stat(1) and
> the number of links.
>
> Would checking for "test -d $PWD" work? I think $PWD is the full path
> and so if it's no longer reachable, the test should fail?
>
> Hope that helps

I love this list!  I thought that '[ -w . ]' and '[ -w $PWD ]' were
practically equivalent.  "Practically" means, in this case, "almost."
But not quite - and the difference is the solution to the problem.
Thanks everyone, particularly John and Dave.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Bash does-directory-exist question

[Giles Orr via talk]

I have a strange Bash question for you.  It's an edge case, but I've
run into it just often enough that I'd like to know how to deal with
it.

How do you determine if the directory you're in has been deleted?

I've done this to myself: 'cd' to a directory in one terminal, then
'rm -r' that same directory in another terminal.  It becomes very
confusing when you try to do something in the first terminal: 'touch
test.txt' responds with "touch: cannot touch 'test.txt': No such file
or directory".  Which is exceptionally hard to decipher because of
course it doesn't exist, I was trying to create it!

The worst case I've seen is 'git', and this is what's brought me back
to this puzzle. If you're in a directory and you 'git rm' the last
file in the directory, 'git' will helpfully delete the directory as
well.  Never mind that you're still in the directory, and are now in
the very confusing position of being in a non-existent directory.

One of my hobbies is tinkering with Bash Prompts.  It's fairly easy to
use 'test/[' to determine if a directory is writable with 'if [ -w .]
...' and then change the colour of the prompt based on the response.
This gives immediate visual feedback on the write-status of the
current directory.  But test's '-w' and '-d' both claim that you're
still in a valid directory under the above circumstances.  Does anyone
know of a simple way to find out if the directory you're currently in
actually exists?

I may have found my own answer: 'stat .' returns (in part) "Links: 0"
which would seem to indicate a non-existent node.  Is that a
definitive answer?  (I don't use 'stat' much.)  I'm also interested to
see if anyone else has a simpler or more "Bash" answer.

Thanks.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Java

[Giles Orr via talk]

On Fri, 5 Jun 2020 at 13:04, Slackrat via talk  wrote:
>
> Can anyone give me a pointer to what precisely I might do to correct the
> situation. The first three suggestions do not apply:
> #   Reduce memory load on the system
> #   Increase physical memory or swap space
> #   Check if swap backing store is full
>
>
> The situation requires a cold boot to correct and that ABSOLUTELY
> nothing be done in the user account except launching an X server
> prior to launching the application.
>
>
> [QUOTE]
>
> bash-4.3# ./HamSphere_4.010a
> Fatal error: out of memory
> Uncaught exception: Out of memory
>
> An earlier version of the application at least privides a logfile when
> it crashes
>
> Server:www.hamsphere.com:8010
> OpenJDK 64-Bit Server VM warning: INFO:
> os::commit_memory(0x7f942b028000, 262144, 0) failed; error='Cannot
> allocate memory' (errno=12)
>
> #
> # There is insufficient memory for the Java Runtime Environment to continue.
> # Native memory allocation (mmap) failed to map 262144 bytes for committing 
> reserved memory.
> # An error report file with more information is saved as:
> # /inconnu/hs_err_pid23337.log
> [/QUOTE]
>
> bash-4.3$ free
>   totalusedfree  shared  buff/cache   
> available
> Mem:5999704  217532 50332322760  748940 
> 5392888
> Swap:   7219376   0 7219376
> bash-4.3$
>
> LogFile - too long to be reasonably opened by emacs, but "cat
> /hs_err_pid23337.log |less" works fine but, at least to me, is not
> helpful.
>
>
> System is Slackware 14.2 on an
> "Ideapad 330 (17, Intel) | Durable, Easy-to-Use 17.3" (sic) ... Lenovo" 
> disaster.

If you'll forgive me fumbling about in an area I barely know anything
about ...  I'll point out that Java sometimes means very different
things by "out of memory" than we think it does, and in this case
we're only finding out about it after it's filtered through the
application logging.  I think this probably means "Java has run out of
the memory you said it could have."  You didn't know you told it about
how much memory it could have, did you?  That's because there are
default values, and it's a pretty common problem to bump up against
them.  (Again, not totally sure that's what's happening here.)

I would try:

export JAVA_TOOL_OPTIONS="-Xmx1024M"

(You should perhaps check if this is already set FIRST, but doing this
at the command line is specific to the current shell.)  Then try
running the application again.

And read this:

https://stackoverflow.com/questions/5374455/what-does-java-option-xmx-stand-for

for a passable explanation of a horrible subject.  Read not just the
accepted answer, but the other stuff as well.

I hope this is what you need.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Automatic .local and .lan hostnames

[Giles Orr via talk]

On Wed, 27 May 2020 at 11:31, James Knott via talk  wrote:
>
> On 2020-05-27 11:29 AM, Giles Orr via talk wrote:
> > So now I'm really interested in how these names are made available.  I
> > don't care which of the two names it provides, but I'd really, really
> > like it to do it consistently.  Can someone point me in the right
> > direction, and I'll go do my homework.  Thanks.
>
> Probablly mDNS.
> https://en.wikipedia.org/wiki/Multicast_DNS

Thanks, that's a perfect point of entry.  It sent me to Avahi, a name
I've been aware of for years but never previously investigated.  And
from there I discovered that 'systemctl status avahi.daemon' on the Pi
shows that the service died about a day and a half ago.  So I have my
work cut out for me trying to figure out why it died.  Is a restart
sufficient, or will it happen again?  Etc.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Automatic .local and .lan hostnames

[Giles Orr via talk]

I have a Raspberry Pi on my local network that acts as an MPD server.
It's hostname is "pib".  I have, for about a year and a half, been
referencing it from other machines on the local network as
"pib.local".  Yesterday, one of the machines stopped being able to
find "pib.local" but can now connect to "pib.lan" - which I don't
think used to work.  I've never set up my local machines to provide
.local or .lan addresses, it seems to be an
automatic service of both Fedora and Debian (or Raspbian).  And I've
always had problems with its indecisiveness about "is it .lan, or is
it .local?"  For some machines it's one, for some machines it's the
other.  But I think this is the first time it's switched.  And no, I
didn't make any configuration changes (or software upgrades) on the
machine that switched.

So now I'm really interested in how these names are made available.  I
don't care which of the two names it provides, but I'd really, really
like it to do it consistently.  Can someone point me in the right
direction, and I'll go do my homework.  Thanks.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] (off topic) What power bank should I avoid?

[Giles Orr via talk]

On Sun, 17 May 2020 at 21:28, William Park via talk  wrote:
>
> My LG Nexus 5X died, and I'm temporarily using an old Samsung Galaxy S3
> with poor battery.  I may need to use a power bank, until I get a new
> phone.
>
> Which power bank do you recommend (or not recommend)?

I've had two power banks over the years - I lost track of one, bought
a new one, found the old one, now I have two.  Different makers, one
is 10,000 mAh, the other is 11,400 mAh.  They're old enough that even
if I told you the make and model, they probably don't exist any more
... but they both still work well - partly because I've only used them
lightly.

So I have very limited experience with power banks.  But I'd suggest
that despite the added weight and cost, the larger ones (1 mAh and
up) are the way to go.  They'll always get you through a day, while
one of the puny ones (2500 mAh) will die after 3/4 charging your
phone, and won't even do that as the battery ages and weakens.
Something else to consider is that a charge indicator on the device is
very useful.

(Not an expert.  Just full of opinions.)

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Compressing an image of a microSD card

[Giles Orr via talk]

On Thu, 14 May 2020 at 18:04, Lennart Sorensen via talk
 wrote:>
> On Thu, May 14, 2020 at 05:20:18PM -0400, D. Hugh Redelmeier via talk wrote:
> > If your filesystem lives on some form of flash (SSD, SD card, USB
> > stick, ...) this can reduce the lifetime and performance of your
> > hardware.
> >
> > The wear-levelling firmware of the drive will think that every block of
> > the drive is "live" (contains valuable information).  This will increase
> > "write amplification".
> >
> > In any case, if you do do this, be sure to use fstrim afterwards.  (I'm
> > not sure that SD cards and USB sticks support trim.
>
> I think the idea was to do it on the image loop back mounted, not on
> the original device.

What all this discussion has made me realize is that I have to either
A) modify the original SD card or B) loop mount and modify the backup.
Option A was what I was initially proposing, but forcing a write
(zeroing out most of the contents) across the whole card isn't a great
idea.  Not terrible, but not great.  Option B involves modifying the
backup, and I think this is a worse idea because 1) a backup should be
an accurate recreation of the source, and 2) modifying the backup
effectively means you'd be restoring something different, and do I
really feel comfortable counting on that?

The more I think about it, the less I want to modify either of them.
Thanks for clarifying the process: it solidified my opinion on what
should be done (although not in the direction I expected).  ie. 10G
really ain't that big, I'll just keep the backup as is.  Thanks
everyone.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk