Re: [GTALUG] A dos port of the Linux program neofetch.

[D. Hugh Redelmeier via talk]

Interesting.

The blurb about MySysInf is a little confusing.
I haven't seen a clearer blurb.  Maybe there is one in the ZIP file.

The first confusing thing is that the blurb seems to call it MySysInf one 
place and Fetch4FD another.  And the URL calls it Fetch4FD (apparently the 
old name).

Googling for MSD gets a lot of unrelated hits.  Here's a useful 
description 

MySysInf is a port of a BASH script called Neofetch 

The program lists a bunch of interesting things about your System.

There seem to be a lot of variants of this program!
I've added an entry for fetch4fd and MySysInf to the Wikipedia page.

I would think that the code for discovering low level resources would be 
very different for Linux and DOS.  Porting this code must have been hard.

| From: Karen Lewellen via talk 
| To: talk@gtalug.org
| Cc: Karen Lewellen 
| Date: Sat, 27 Apr 2024 13:37:24 -0400 (EDT)
| Subject: [GTALUG] A dos port of the Linux program neofetch.
| 
| Hi folks,
| I may only end up using Linux via dos ssh into Linux shells, but boy am I
| happy when someone ports a Linux tool for DOS.
| 
| https://sourceforge.net/p/freedos/news/2024/03/fetch4fd-system-info-program/
| 
| My goal was  something comparative to MSD,  and Poof!
| Cheers,
| Karen
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] ot: backlab?

[D. Hugh Redelmeier via talk]

Do you mean "Hacklab" as in ?
I would have assumed you did, but you used "backlab" twice.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] monitor characteristics: EDID

[D. Hugh Redelmeier via talk]

I'm trying to figure out the characteristics of the display on my Asus 
Vivobook S notebook.

One way to figure this out is to look at the monitor's EDID
(Extended Display Identification Data).

I installed a program called edid-decode since otherwise the EDID is just 
binary bytes.

Then I used this command:

  edid-decode /sys/class/drm/card1-eDP-1/edid | less

The path is to one of the edid's exposed by the Linux kernel.  I picked 
eDP because that should be the built-in display.

I've included to result at the end of this message.

Things I noted

- manufacturer is SDC (Samsung Display Company)

- manufactured in the 25th week of 2021 (older than I thought)

- bits / color channel: 10 (very nice)

- only color format: RGB 4:4:4
  That means no chroma subsampling but RGB doesn't support chroma
  subsampling anyway

 EDID 
edid-decode (hex):

00 ff ff ff ff ff ff 00 4c 83 6d 41 00 00 00 00
19 1f 01 04 b5 1f 14 78 02 0c f1 ae 52 3c b9 23
0c 50 54 00 00 00 01 01 01 01 01 01 01 01 01 01
01 01 01 01 01 01 39 ff 40 5c b0 08 20 70 14 20
08 04 38 c3 10 00 00 1b 39 ff 40 5c b0 08 48 77
14 20 08 04 38 c3 10 00 00 1b 00 00 00 0f 00 ff
0a 78 ff 0a 3c 28 80 02 00 00 00 00 00 00 00 fe
00 41 54 4e 41 34 35 41 46 30 31 2d 30 20 01 72

02 03 0f 00 e3 05 80 00 e6 06 05 01 74 60 07 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 b7



Block 0, Base EDID:
  EDID Structure Version & Revision: 1.4
  Vendor & Product Identification:
Manufacturer: SDC
Model: 16749
Made in: week 25 of 2021
  Basic Display Parameters & Features:
Digital display
Bits per primary color channel: 10
DisplayPort interface
Maximum image size: 31 cm x 20 cm
Gamma: 2.20
Supported color formats: RGB 4:4:4
First detailed timing includes the native pixel format and preferred 
refresh rate
  Color Characteristics:
Red  : 0.6796, 0.3203
Green: 0.2373, 0.7226
Blue : 0.1396, 0.0498
White: 0.3125, 0.3291
  Established Timings I & II: none
  Standard Timings: none
  Detailed Timing Descriptors:
DTD 1:  2880x1800  120.001014 Hz  16:10   219.842 kHz653.37 MHz 
(312 mm x 195 mm)
 Hfront   20 Hsync  32 Hback   40 Hpol P
 Vfront   16 Vsync   8 Vback8 Vpol N
DTD 2:  2880x1800   60.000507 Hz  16:10   219.842 kHz653.37 MHz 
(312 mm x 195 mm)
 Hfront   20 Hsync  32 Hback   40 Hpol P
 Vfront   16 Vsync   8 Vback 1840 Vpol N
Manufacturer-Specified Display Descriptor (0x0f): 00 0f 00 ff 0a 78 ff 0a 
3c 28 80 02 00 00 00 00 '.x..<(..'
Alphanumeric Data String: 'ATNA45AF01-0 '
  Extension blocks: 1
Checksum: 0x72



Block 1, CTA-861 Extension Block:
  Revision: 3
  Native detailed modes: 0
  Colorimetry Data Block:
BT2020RGB
  HDR Static Metadata Data Block:
Electro optical transfer functions:
  Traditional gamma - SDR luminance range
  SMPTE ST2084
Supported static metadata descriptors:
  Static metadata type 1
Desired content max luminance: 116 (616.884 cd/m^2)
Desired content max frame-average luminance: 96 (400.000 cd/m^2)
Desired content min luminance: 7 (0.005 cd/m^2)
Checksum: 0xb7  Unused space in Extension Block: 112 bytes
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] GPU advice needed

[D. Hugh Redelmeier via talk]

GPU-for-AI advice from the cheap seats (I've never done this).

AI stacks seem very fragile.  Any time you make a substitution the
whole thing might misbehave.

You can either copy EXACTLY a working configuration from someone else or 
you can sign up for adventure.  Most likely you will try some of both.

NVidia cards come in at least three flavours: those for AI, those for
pros (running things like autocad), and those for gamers.
They try hard to segment the market so they can charge as much as
possible for the segment.

Serious AI cards cost up to US$20,000.  Maybe more.

How do they keep gamer cards from being great AI cards?  I don't know
all the techniques, but they include:

- crippling the speed of 64-bit floating point

- limiting the amount of on-board memory (pretty important)

- (I think) limiting the ability to partition the card into virtual
  cards for multiple processes to run in parallel safely.

- limiting how GPUs can scale up (interconnect)

This is on top of the crippling they did to hobble cryptocurrency
mining.  I don't know what that amounted to technically.

Advice for buying a card:

- how much do you want to spend?  Not enough!

- you probably want as much on-board RAM as possible because swapping
  stuff between the computer's RAM and the GPU's RAM is apparently a
  serious bottleneck.  I don't have a cost/benefit curve.  I'm pretty
  sure that it is like real RAM: performance falls off a cliff when
  you don't have enough.

- fans probably matter because (1) good cooling should be quieter than bad
  cooling (not in data-centre cards: nobody cares about noise there),
  and (2) without good cooling, throttling will happen.
  Read good review sites to get opinions about card cooling issues.

- Guess: the version of PCIe used might matter.  We're in a period of
  transition.  (Remember: the motherboard and the GPU have to both
  support the PCIe version you target.)

- if I were at all adventurous, I'd look at Intel graphics cards.
  They are probably a bargain on a per teraflop basis.
  Intel tries hard to push AI on Linux.  They are pretty good
  open-source players and seem more competent than AMD.

- my impression is that AMD for AI on Linux smells a bit like a lost
  cause: they care about their commercial GPU-compute customers but
  not us.  ROCm only really seems to work on their industrial GPU
  cards.

- the latest gen of NVidia is apparently not much a step up from the
  previous one.  Check the performance, not just the model number.

Fun fact: Mac folks never cease to brag about how great Unified Memory
is on the M1 etc.  That seems silly when you realize that integrated
GPUs have always had this.  But there is apparently a case where the
performance benefit is great:

- tonnes of memory on the Mac (very expensive), more than you can get
  on an NVidia gamer GPU

- working with a model that doesn't fit in the NVidia GPU
  but does fit in the Mac's RAM.

The Mac's RAM has very fast access from both the GPU and the CPU.
Much faster than the PCIe bus.  In fact, much faster than an X86 can
access bulk RAM.

Phoronix.com seems to focus on Linux and GPUs.

https://ca.pcpartpicker.com/ might be helpful finding deals.

Where to buy?

Check out
Amazon (easy returns; sometimes good prices)

Canada Computers (sometimes bad customer service but I've not really
encountered this)

Best Buy (sometimes)

NewEgg (sometimes)

Memory Express
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Securely wiping SSDs

[D. Hugh Redelmeier via talk]

| From: David Thornton via talk 

| I grilled my local hardware security friend who said you should never trust
| the secure delete feature in ssd . In general it has been found insecure.

It seems hard to accidentally screw up Secure Delete:

- everything written to the flash layer is encode
- everything read from the flash layer is decoded
- secure erase simply changes the key

It is fairly easy to intentionally screw up Secure Delete.
Trusting Secure Delete involves trusting the manufacturer:
there is no sure way to test for this failing.

Does your guy have examples or even indications of Secure Delete failing?
In particular, what does "In general it has been found insecure." mean?  
What has been found?  How has it been found?
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Securely wiping SSDs

[D. Hugh Redelmeier via talk]

| From: D. Hugh Redelmeier via talk 

| Some disks have a feature where they have a key that encrypts every block.  
| The key persists in the drive.  But if you want to wipe the disk, you 
| change this key.  Then every block is nonsense until it is rewritten.
| I haven't got time to look up the name for this feature.  It is optional.  
| I think that most enterprise drives have it.

The feature is called "Self-Encrypting Drive" (SED).  This seems like
the only way to safely allow a drive to be re-used by a someone you
don't trust.

<https://www.techspot.com/guides/869-self-encrypting-drives/>
Note: that's an old article.

The simplest way of using it makes the encryption / decryption
invisible to the computer.  You just have to change the key to do a
"secure erase".

There is a next level in which the drive requires a password when
powered on.  This isn't needed for Secure Erase.

Apparently a lot of drives have Secure Erase but the feature is rarely
advertised.

Sad fact: you may only realize that you want this feature long after you 
bought it.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Securely wiping SSDs

[D. Hugh Redelmeier via talk]

It depends on your paranoia level.

In the good old days, there were no levels hiding the disk sectors from 
the computer.

Then came automatic sector mapping for bad blocks.  Really convenient.  
But how are you going to wipe those mapped-out blocks?  As far as I know, 
there is no way to do so.  But there are probably only a few such blocks 
and they are probably somewhat defective.  We'll ostrich the problem away.

Then came flash memory with wear-levelling firmware between the computer 
and the bulk flash.  And the devices are over-provisioned: there is more 
flash than "simulate" disk space.  I don't think you can be sure to have 
erased it.  Maybe I'm wrong.

Some disks have a feature where they have a key that encrypts every block.  
The key persists in the drive.  But if you want to wipe the disk, you 
change this key.  Then every block is nonsense until it is rewitten.
I haven't got time to look up the name for this feature.  It is optional.  
I think that most enterprise drives have it.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] DOS > bash

[D. Hugh Redelmeier via talk]

| From: Ron / BCLUG via talk 

| How does `rename` (a Perl program, I believe?) handle globbing and pathname
| expansion?
| 
| 
| i.e. in a folder with 3 files: one.sh, two.sh, file.tar, typing:
| 
| rename *.sh *
| 
| means rename gets passed these parameters:
| 
| one.sh two.sh one.sh two.sh file.tar

The DOS ren command is very simple.
So too is the Unix Utilities one included in Fedora by default.

rename [options] expression replacement file...

The expression and replacement are simple strings, not patterns.
There are options like --all and --last that affect where the expression 
can be matched.

What you want is
rename .sh '' *.sh
which means
for each filename matched by *.sh
change every occurrence of ".sh" in its name
to '' (empty)

I think that regular expressions in at least "expression" would be nice.
I'd often use ^ or $, to prevent surprises.
So the perl-based version might be more useful.

But I so rarely use rename that I have to read the man page each time.
Once or twice a year.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] DOS > bash

[D. Hugh Redelmeier via talk]

| From: William Park via talk 

| Ubuntu/Debian distros
| will include 'rename' where you can use regex.

So does Fedora.

It is much more powerful that DOS's ren command.

Even so, I so rarely use rename that I have to read the man page each 
time.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Debian Live Linux -- Change Overlay Filesystem -- From Tempfs Ramdisk To Hard Drive ?? [was] Re: Debian Live Linux -- Overlay Filesystem -- Where Allocated ??

[D. Hugh Redelmeier via talk]

| From: D. Hugh Redelmeier via talk 

| 2.  (Less useful to you) Various distros are experimenting with an 
| immutable core.  So you could have the same problem on your disk as you 
| now have on your stick. <https://fedoramagazine.org/what-is-silverblue/>

There is a group of immutable desktop Fedora spins called Atomic
Desktops:
  <https://fedoramagazine.org/introducing-fedora-atomic-desktops/>

The writeup says they aren't exactly immutable.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Debian Live Linux -- Change Overlay Filesystem -- From Tempfs Ramdisk To Hard Drive ?? [was] Re: Debian Live Linux -- Overlay Filesystem -- Where Allocated ??

[D. Hugh Redelmeier via talk]

1.  It is easy to install a Linux system onto a USB stick: just tell the 
Linux installer to use the (second) USB stick instead of the HDD.  The 
result has ordinary filesystems that can be updated, just like the one on 
your hard drive.

2.  (Less useful to you) Various distros are experimenting with an 
immutable core.  So you could have the same problem on your disk as you 
now have on your stick. 

I bought a USB SSD and installed Fedora on it (see 1 above).  I boot from 
it for a bunch of maintenance tasks.  For example: resizing partitions 
(gparted).
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Pi Day; Raspberry Pi virtual Meet-Up tonight

[D. Hugh Redelmeier via talk]

Pi Day (3.14) is no longer just a silly math undergrad joke.  Even piza 
companies observe it.



The Toronto Raspberry Pi meetup group is restarting tonight.


A new Raspberry Pi OS release is out.

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] amazing but expensive Windows notebook (will run Linux)

[D. Hugh Redelmeier via talk]

Two left at the moment 


The notebook: ThinkPad P16 with most nice features.
The price ~$1400 + tax

Too rich for me.  But I can drool.

Although this is sold through ebay, the vendor is Lenovo itself.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] uses and dangers of AI in science

[D. Hugh Redelmeier via talk]

I found this article quite interesting.


---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Monthly meeting on Tuesday March 12

[D. Hugh Redelmeier via talk]

Scraped from :

The Future of LUGs and the Open Source Community. It’s AI, Baby! with 
Marcel Gagné

12 March, 2024 at 07:30 PM

Join GTALUG and our guest this month, Marcel Gagné, to discuss the current 
state of Linux User Groups and AI.

Time: March 12th, 2024 19:30 AM Eastern Time

Join us on Big Blue Button: https://blue.lpi.org/b/eva-zjc-gjy-kgl.
Location---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Grrr: CRA vs evince

[D. Hugh Redelmeier via talk]

Background / Public Service Announcement:

This year, the CRA has started requiring taxpayers to file returns for
joint accounts (I think joint accounts for spouses are not included).
These are called "bare trusts" and one needs to

- register with the CRA to get a trust number (kind of like a SIN)

- file a T3 tax return (but there will be no tax)

- file a T3 Schedule 15 form for each beneficial owner.

The deadline for filing for 2023 is the end of this month.


Linux-related:

The fillable form for Schedule 15 is in an Adobe-proprietary variant PDF.
evince doesn't understand it but Firefox does.


Evince displays it as
==
Please wait...
If this message is not eventually replaced by the proper contents of 
the 
document, your PDF
viewer may not be able to display this type of document.
You can upgrade to the latest version of Adobe Reader for Windows®, 
Mac, 
or Linux® by
visiting http://www.adobe.com/go/reader_download.
For more assistance with Adobe Reader visit 
http://www.adobe.com/go/acrreader.
Windows is either a registered trademark or a trademark of Microsoft 
Corporation in the United States and/or other countries. Mac is a 
trademark
of Apple Inc., registered in the United States and other countries. 
Linux 
is the registered trademark of Linus Torvalds in the U.S. and other
==

On the other hand, the T3 tax return fillable PDF works fine with evince.


I think that the CRA should ensure that its forms are not in a
proprietary format.

Today I learned that Adobe Reader again supports Linux.  I'm pretty
sure it will be horrible: large, not really native to my distro,
closed source.---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] .local question

[D. Hugh Redelmeier via talk]

| From: D. Hugh Redelmeier via talk 

| Tonight I'm creating an experimental NextCloud server.  Let's say it's 
| hostname is nc.
| 
| - ping nc.local works
| - ssh nc.local works
| - host nc.local works
| - Firefox and even links cannot see nc.local.

This report turns out to be very incomplete.

These tests were on nc.
On any other host, nc.local doesn't resolve at all.
Odd.

avahi is running.
$ avahi-browse --all
shows many things, but none of them being my PCs.

/etc/nsswitch.conf has a line:
hosts:  files myhostname mdns4_minimal [NOTFOUND=return] resolve 
[!UNAVAIL=return] dns

So I'm guessing that the nc.local is being resolved by "myhostname",
not mdns4_minimal.

On my desktop:
$ sudo netstat -l -p -4 | grep avahi | cat
udp0  0 0.0.0.0:mdns0.0.0.0:*   
890/avahi-daemon: r 
udp0  0 0.0.0.0:48151   0.0.0.0:*   
890/avahi-daemon: r 
$ grep mdns /etc/services | cat
mdns5353/tcp# Multicast DNS
mdns5353/udp# Multicast DNS
mdnsresponder   5354/tcpnoclog  # Multicast DNS Responder IPC
mdnsresponder   5354/udpnoclog  # Multicast DNS Responder IPC

The firewall allows mdns

BTW, I'm tunning fedora 39 on all the boxes involved.  I was going by
this tutorial; it's bit old:
<https://fedoramagazine.org/find-systems-easily-lan-mdns/>

Pausing for dinner.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] .local question (Off Topic: how did you acquire a /24?)

[D. Hugh Redelmeier via talk]

| From: Ron / BCLUG via talk 

| D. Hugh Redelmeier via talk wrote on 2024-02-26 22:43:
| 
| > (I have /24 globally routable IP addresses.)
| 
| I'm curious about the story behind how you acquired what's become such a rare
| item?

If you asked 30+ years ago you got a "Class C" assignment.
This isn't delegated.

More recently, you get subnets delegated by your ISP.  If you change
ISPs, you change IPs.

| Also - how does one use them behind a residential ISP?

I have an arrangement with a small ISP.  I've known one of the guys
there for a long time.

Unfortunately they can only resell DSL so the speed is low.

Most of my traffic goes through NAT and over Bell Fibre (with a Bell
dynamic address).

If it mattered, I could rent a box at a colo and get fast speed for my
/24.  I could tunnel the traffic to my home through Bell.
I have no idea whether colos are used to Bring Your Own IPs.

| Do you have your own AS number?

I haven't needed one.  I've always assumed that I could get one if I
wanted to.

I've switched ISPs a few times over the years.  I switch as the
current one dies or is consolidated.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] .local question

[D. Hugh Redelmeier via talk]

I used to give all my machines permanent IP addresses and added entries 
in my domain for them.  (I have /24 globally routable IP addresses.)

I then got lazy and let most be assigned dynamically.  But not servers.

Now I'm even more lazy.  I'm starting to use .local.  So machines declare 
their hostname and mDNS / bonjour gets to resolve name.local.  Neat.

Tonight I'm creating an experimental NextCloud server.  Let's say it's 
hostname is nc.

- ping nc.local works
- ssh nc.local works
- host nc.local works
- Firefox and even links cannot see nc.local.

Why is that?
Firefox is set to use the default DNS, not Firefox's own.

If I give Firefox the actual IP address, Apache balks because it cares 
about the DNS name used.  But that's further than nc.local got.

I guess Firefox doesn't do mDNS.  Why would that be?  Or am I making a 
wrong guess?

Clearly I could give to computer a real DNS name, but I'm kind of 
stubborn and want to understand what's going on.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] War Story: adding RAM to a laptop can increase the speed

[D. Hugh Redelmeier via talk]

| From: mwilson--- via talk 

| My main desktop is a Raspberry Pi 4B+ with 8GiB.

Neat!

| `top` says that swap is full (100MiB), but there's 1.2 GiB of memory
| available.

I'd be tempted to add swap.  But then again, if that is on an SD card,
maybe not.

I don't know the kernel policies.  Or rather, I know a lot of
different kernels' policies.  So this will be unreliable.

1. Paging policies tend to migrate pages that haven't been used
   recently, but must be preserved, to swap.  Code pages don't need to
   be written to swap because they are immutable -- they can be
   reloaded from the original file.

2. If there is no memory pressure (i.e. more RAM than things to occupy
   it), probably nothing goes to swap.

3. It used to be that RAM was used for processes: code and data.  For
   a few decades, files have been allowed to fill up available RAM but
   could be easily evicted to their place in the filesystem when
   memory pressure got high.  Eviction often requires no I/O: if the
   file was only being read.  Consequence: RAM may all be used but
   since things can be evicted there is room for more stuff: you
   haven't run out.

4. In the last couple of years a new level of swap has been created:
   compressed swap stored in a RAM-disk.  This sounds crazy but makes
   sense if the stuff being swapped is significantly compressable.
   And if writing to swap is slow (eg. SD card)

So:

Your swap being full may mean that some stuff is stranded in RAM
even though it isn't used much.

Your RAM having 1.2 GiB available does not actually indicate that RAM
is tight.  Much of the occupied space might be evictable files.

If you were truly short of RAM + swap, I think you would see OOM
messages in the log.

| GIMP that came with Raspberry Pi OS is not wonderfully stable -- might
| crash once a day on average.  Doesn't seem directly due to memory shortage
| though.  Console messages from one time that I ran it from the command
| line hinted that there might be mutual exclusion problems with GtK, or
| something.

Are you running it from GDB?  In theory that might help you figure out
what's going on.  But GDB would like to to install the symbols for
executables that you care about and that may overwhelm your SD card.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] War Story: adding RAM to a laptop can increase the speed

[D. Hugh Redelmeier via talk]

| From: Howard Gibson via talk 

| On Mon, 19 Feb 2024 12:31:08 -0500 (EST)
| "D. Hugh Redelmeier via talk"  wrote:

| > What tasks to you do that challenge computers with only 8 GiB?

| I probably spend too much time on YouTube.  I use GIMP to mess with 
| photographs.  I use LaTeX a lot to prepare documents, but if I am 
| sharing info, I use LibreOffice, and the Microsoft office document 
| formats.

Thanks for the reply.

We do some of those things on 8 GiB.

Which of those tasks actually pinch when you run them on an 8 GiB system?
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] War Story: adding RAM to a laptop can increase the speed

[D. Hugh Redelmeier via talk]

| From: Howard Gibson via talk 
| 
|I have long ago given up trying to run computers on 8GB RAM.  For 
|GNU/Linux, 16GB minimum.

What tasks to you do that challenge computers with only 8 GiB?

For me, it is Firefox with a lot of tabs.  I'm sure that some web pages 
are way more RAM-intensive than others (see about:performance) but I 
rarely pick pages based on weight.  I admit that I keep way too many tabs 
open but mostly on my desktop which has 24 GiB of RAM.

I regularly run Fedora / GNOME desktop on 4 G machines.  They work but are 
not fast, for multiple reasons.  They have Atom-microarchitecture 
processors and eMMC "disks", both of which slow them down.  On the plus 
side, they are delightfully inexpensive, small, energy efficient, and 
silent.

I have several notebooks, of varying ages, with 8 GiB of RAM.  They all 
work fine.

RAM is cheap enough to overbuy, except when you are forced to buy it with 
the computer.  Manufacturers really pad the price of computers with more 
RAM.  Most notebooks now come with soldered RAM, not socketed.

Some notebooks come with soldered RAM plus an empty or occupied socket.  
That's not as good as two sockets. Best performance, all other things 
being equal, is to have the same RAM size in the socket and soldered.  
Any variation comes at a cost.

One reason I bought this particular cheap notebook is that it came with a 
perfect memory configuration for upgrading to 16 GiB.

A certain amount of RAM is tied up in fixed overhead assignments: the 
kernel, the memory used by the iGPU, the memory used by NVMe drives.  So 
doubling RAM makes more than twice as much RAM available for applications.

In this particular case, adding RAM added RAM bandwidth, improving the 
speed of computation.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] War Story: adding RAM to a laptop can increase the speed

[D. Hugh Redelmeier via talk]

I bought a very inexpensive laptop from costco.ca.
Dead deal: 


Unlike most laptops these days, it has two sockets for RAM.
Only one was occupied.  It had an 8 GiB stick of RAM.

- two stick should perform better: allow dual-channel memory access

- many consider 8 GiB too little RAM these days.  I'm not sure that's true
  but more cannot hurt

I bought another 8 GiB stick with the same specs from Amazon for $25.99 
and installed it.


I ran passmark benchmarking software before and after the upgrade



The Passmark Rating (whatever that is) went up about 20%.
You can see that the laptop still has modest performance.

I'm glad the HP included one 8 GiB stick instead of two 4 GiB sticks.  
That made it cheap for me to upgrade (add one stick instead of replacing 
two) but it made the off-the-shelf performance worse.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Air Canada claims its chatbot is liable, not AC!

[D. Hugh Redelmeier via talk]

There's surely more to this story.

A customer sued AC for the consequences of bad advice AC's chatbot
gave.

In an argument that appeared to flabbergast a small claims
adjudicator in British Columbia, the airline attempted to
distance itself from its own chatbot's bad advice by claiming
the online tool was "a separate legal entity that is
responsible for its own actions."
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] meeting idea: AI Explorers' Reports

[D. Hugh Redelmeier via talk]

| From: Evan Leibovitch via talk 
| 
| I have confirmed that Marcel Gagné will lead our March meeting on "Open
| Source AI", as was discussed at both the recent Friday and Tuesday
| discussions.

Great!

| It is an open question whether he should give a presentation, or simply
| lead a round-table with a brief introduction and a list of relevant topics
| that should be discussed.
| Based on our discussions, the latter appears the more popular choice.
| 
| Thoughts?

If Marcel can and would fill a meeting himself, that would be great.

We could have a separate meeting for explorers' tales.  But I admit that 
it might be hard to guess how many would come forward.---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] I asked Google's Gemini to compare iGPUs

[D. Hugh Redelmeier via talk]

Sorry for the previous partial response.  Trying again.

| From: Evan Leibovitch via talk 

| My original Gemini search said nothing about DirectX or OpenGL.

You mean your prompt, right?
Which was the same as my prompt, right?

Things that my run got wrong and your run didn't:

- did not mention OpenGL, but it didn't get it wrong.

- got the EU count right.

- got DirectX == DX right (I think;  I don't really know what "Ultimate" 
  means)

Extract from your run:

*i3-1005G1 (Ice Lake generation, 2019):*

   - *Features:* DX12 compatible,

*i3-1215U (Alder Lake generation, 2022):*
   - *Features:* DX12 Ultimate compatible, 

*Here's a summary:*
Featurei3-1005G1i3-1215U
Generation Ice Lake (2019) Alder Lake (2022)
Execution Units 32 64
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] I asked Google's Gemini to compare iGPUs

[D. Hugh Redelmeier via talk]

| From: Evan Leibovitch via talk 

| My original Gemini search said nothing about DirectX or OpenGL.

You mean your prompt, right?
Which was the same as my prompt, right?

>From you previous message, the answer was

---
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

Re: [GTALUG] I asked ChatGPT to compare iGPUs.

[D. Hugh Redelmeier via talk]

| From: Evan Leibovitch via talk 
| 
| For what it's worth, I fed Hugh's question into Google Gemini and got this
| (including the formatting, this is just cut/pasted):

Interesting.

As described in my second posting, I too tried Gemini.

- I got a different table from Evan (boxes and columns)

- several details in the answers it gave me were wrong.
  Those mistakes did not appear in Evan's run.

I wonder why it did a better job for Evan.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] I asked Google's Gemini to compare iGPUs

[D. Hugh Redelmeier via talk]

In the last message, I described how ChatGPT 3.5 confidently gave a wrong 
answer to my prompt "compare igpu in i3-1005G1 and i3-1215U".

Now I'm trying free Google Gemini.  New this month!

It's answer is much better.  But that makes its serious mistakes more 
difficult to catch.

First of all, I blame Intel.  It really doesn't describe the differences 
in iGPUs when showing the difference between the processors:


Gemini incorrectly says the i3-1005g1 supports DirectX 12.1 whereas Intel 
says 12.
Gemini says the i3-1005g1 supports OpenGL 4.5 whereas Intel says 4.6.
Gemini says that the i3-1215u's iGPU has 80 Execution Units whereas Intel 
says 64.

I'll stop here.  Gemini has wasted my time.  If I were less careful, it 
would have mislead me.

This is the comparison that I trust the most (but it could be wrong too)

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] I asked ChatGPT to compare iGPUs.

[D. Hugh Redelmeier via talk]

I have two models of dirt-cheap notebook.  One was $300 for an "open box" 
not quite two years ago and the other was $300 ten days ago.

It's easy to google on the web to compare the CPUs but not so easy to find 
out the difference between the integrated GPUs.  So I asked ChatGPT 3.5 
(the free one).

"compare igpu in i3-1005G1 and i3-1215U"

The answer looked OK but I'm quite sure that it is very wrong.  It thought 
that the i3-1005G1 was newer and more advanced than the i3-1215U.  The 
first clue is that it thinks that the i3-1215u is part of Intel's 8th gen 
Core processors whereas it is part of the 12th gen.

I then asked it "why do you thing i3-1215u is 8th gen?".  It appologized 
and said that that processor was part of the 10th gen "Comet Lake".  It is 
not.

I wonder if this is due to its training date being before the i3-1215u's 
launch (first quarter of 2022).

It sure would be nice if it knew what it didn't know.

Note: I'm not cherry-picking this example.  This is the first prompt I've 
given in weeks.

 ChatGPT output:
When comparing the integrated graphics processing units (iGPUs) of the 
Intel Core i3-1005G1 and the i3-1215U, there are a few key differences to 
consider:

Architecture:
The i3-1005G1 is part of the 10th generation Intel Core 
processors, codenamed Ice Lake. It features the Gen11 graphics 
architecture.
The i3-1215U, on the other hand, belongs to the 8th generation 
Intel Core processors, codenamed Kaby Lake Refresh. It features Intel HD 
Graphics 615.

Performance:
The i3-1005G1 generally offers better graphical performance 
compared to the i3-1215U due to the newer architecture and improved 
integrated graphics in the Ice Lake processors.
The Gen11 graphics in the i3-1005G1 typically provide better 
support for modern gaming and multimedia applications compared to the 
Intel HD Graphics 615 in the i3-1215U.

Features:
The i3-1005G1 supports newer technologies such as Intel's Adaptive 
Sync, which helps reduce screen tearing during gaming or other 
graphic-intensive tasks.
The i3-1215U may lack some of these newer features and 
optimizations found in the 10th generation processors.

Power Efficiency:
The i3-1005G1, being a part of the Ice Lake architecture, benefits 
from Intel's advancements in power efficiency, resulting in potentially 
better battery life and lower power consumption compared to the i3-1215U.

In summary, while both iGPUs serve their purpose for light gaming and 
everyday tasks, the i3-1005G1 generally offers better performance and more 
advanced features due to its newer architecture.

 second prompt

User
why do you thing i3-1215u is 8th gen?
ChatGPT

I apologize for the confusion. It seems there might have been a mistake in 
my response. The Intel Core i3-1215U is actually part of the 10th 
generation Intel Core processors, codenamed Comet Lake. Let me correct the 
comparison between the iGPUs of the i3-1005G1 (Ice Lake) and the i3-1215U 
(Comet Lake):

Architecture:
The i3-1005G1 features the Gen11 graphics architecture as part of 
the Ice Lake processors.
The i3-1215U, being a part of the Comet Lake processors, features 
Intel UHD Graphics.

Performance:
The i3-1005G1 with Gen11 graphics generally offers better 
graphical performance compared to the i3-1215U with Intel UHD Graphics due 
to the architectural improvements in the Ice Lake processors.
The Gen11 graphics typically provide better support for modern 
gaming and multimedia applications compared to Intel UHD Graphics.

Features:
The i3-1005G1 may support newer technologies and features compared 
to the i3-1215U, although both are part of the 10th generation Intel Core 
processors.

Power Efficiency:
The i3-1005G1 might have an advantage in power efficiency due to 
the architectural improvements in Ice Lake, potentially leading to better 
battery life compared to the i3-1215U.

So, to clarify, both the i3-1005G1 and the i3-1215U belong to the 10th 
generation Intel Core processors, but they differ in their graphics 
architecture and associated features.

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] meeting idea: AI Explorers' Reports

[D. Hugh Redelmeier via talk]

AI is an exploding field.  Lots of stuff is new to us.  Many of us are 
trying things.

I think that it would be interesting if those of us that have experimented 
would chat in a meeting about their experiences.  For example:

- what were you trying to do?

- how is it working out?

- what approaches are you thinking might be interesting or productive?

- what approaches have you tried and how did they fail?

- did you have fun?

I would love to hear explorers tales.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] This month's meeting: Friday (Tomorrow) AND Tuesday

[D. Hugh Redelmeier via talk]

Our web site has the announcement and the list doesn't.  So I'm
copying the page to this message.

NOTE THE UNIQUE TIMING.

The Future of Linux User Groups and Open Source Advocacy
with Stefano Maffulli, Executive Directorof the Open Source Initiative

09 February, 2024 at 11:00 AM

Join GTALUG and our guest this month, Stefano Maffulli, to discuss the
current state of Linux User Groups and Open Source advocacy, and how
our community can be most effective in the coming years.

Stafano is Executive Director of the Open Source Initiative (OSI,
opensource.org), maintainers of the Open Source Definition, and
involved in ongoing advocacy and community development worldwide.

PLEASE NOTE THE MEETING DATE/TIME. To address timezone and
availability issues, Stefano's live presentation will take place
Friday February 9 at 11:00am Toronto time (16:00 UTC).

A recording of the presentation as well as followup Q will be held
during the usual GTALUG meeting time Tuesday February 13 at 19:30
Toronto time (Feb 14 at 00:30 UTC).

We're going to use Big BLue Button for this meeting:
https://blue.lpi.org/b/eva-zjc-gjy-kgl

Time: February 09th, 2024 11:00 AM Eastern Time

Join us on Big Blue Button: https://blue.lpi.org/b/eva-zjc-gjy-kgl

Location
Big Blue Button - https://blue.lpi.org/b/eva-zjc-gjy-kgl

Schedule
7:30 pm Meeting and presentation.

Code of Conduct

We want a productive happy community that can welcome new ideas,
improve every process every year, and foster collaboration between
individuals with differing needs, interests and skills.

We gain strength from diversity, and actively seek participation from
those who enhance it. This code of conduct exists to ensure that
diverse groups collaborate to mutual advantage and enjoyment. We will
challenge prejudice that could jeopardize the participation of any
person in the community.

The Code of Conduct governs how we behave in public or in private
whenever the Linux community will be judged by our actions. We expect
it to be honored by everyone who represents the community officially
or informally, claims affiliation, or participates directly. It
applies to activities online or offline.

We invite anybody to participate. Our community is open.

We encourage you to read the complete Code of Conduct before attending
the meeting.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] RISC-V News

[D. Hugh Redelmeier via talk]

| From: Evan Leibovitch via talk 

| On Fri, Feb 2, 2024 at 11:38 AM D. Hugh Redelmeier via talk  
wrote:
|  
|   For serious applications, the openness RISC-V helps but doesn't make 
everything you
|   need open and free.  Or even available.  You actually need chip designs 
-- what SiFive
|   sells.
| 
| 
| If the RISC-V design is open source, what is SiFive selling? Something easily 
copyable? Support
| and documentation?

Scott has answered this, but I will expand on it.

All important processor architectures are well documented: so we know what 
the processor is supposed to do.

To implement it (in a chip, for example) and sell it, you need to have the 
right to do so.  For commercially important architectures, these rights 
cost real money or are unavailable.  ARM is the only one that sells rights 
at a price that is worth paying.  RISC-V is becoming commercially 
important and those rights are free.

Once you have the right to implement it, you then have the hard work of 
actually designing and producing a system (including a processor) that 
will actually run programs for that architecture.

You've seen photomicrographs of processor chips.  All those blobs are not 
random: they are the product of a lot of work.  Not unlike writing a 
program.  And to make such a design that meets performance goals 
multiplies the work and experience necessary.  There are not many 
teams with that experience.

All that work needs eventually to be paid for by a sufficiently promising 
market, one that doesn't exist now.  Chicken-and-egg.  So a successful 
process involves a step-by-step development of capabilities and markets.  
We depend on something like venture capitalists or governments to prime 
that pump.

|   ARM has a vibrant ecosystem with all these things available for 
licensing.  And ARM
|   doesn't seem to be too greedy.  Even so, it has taken a long time to 
get ARM
|   processors that match x86 at the high end.
| 
| 
| Assuming that ARM's time to implement was not a matter of laziness, doesn't 
its experience suggest
| that RISC-V will take similarly long -- or longer -- to evolve from low-level 
SBCs to high-end
| computing?
| 
|   So: if you want a short time to delivery, ARM is way ahead.
| 
| 
| Plus, ARM has clients such as Qualcomm and Samsung and Apple that have 
decades of experience in
| implementation of the architecture and lots of high-quality fabs.

Not really.

High-end processor fabrication is now a specialized business.  Only 
Samsung, TSMC, and Intel have the capability as far as I know.  There are 
stories out of China suggesting that they are trying really hard to get 
there (SMIC, for example).

Most ARM processor designs are done by ARM in-house and are licensed at 
rates that seem to be good enough for the market.  Interestingly, ARM 
charges considerably more for you to use their architecture without using 
one of their designs.  The few companies with that second license include 
Apple (who got their license by being a fo-founder of ARM as we know it) 
and Qualcomm (I think).

This second license requires conviction that your processor design team 
can do better than an off-the-shelf design for your application.
A few companies have tried and failed (eg. AMD).  Of the shelf ARM designs 
have not been good enough to break into the datacentre but a couple of 
non-ARM designs are doing OK (Amazon's Graviton, for example).

Note: users of the second license could just use RISC-V instead, for free.

In my opinion, the software support for RISC-V is mature.  Linux supports 
it as a first class architecture and that's all I need.  Android supports 
it or soon will (I don't remember which).  Microsoft does not support it, 
as far as I know.

| So the definition of "short" here could be an understatement.

Yes.  ARM has taken a lot longer than I expected.

The NetWinder was a credible machine for desktop or server Linux something 
like 25 years ago.  It failed for several reasons but one was that there 
was no ARM implementation that got near x86 performance.

Remember: that same fact killed off all the RISC desktops (HP, MIPS, 
SPARC, Power) and 68k, NS32032, etc.

|   If you think more strategically, RISC-V has some advantages.
| 
| 
| Not sure I gather this conclusion from the rationale.
|  
|   The US used a foot-gun on Huawei by banning ARM from dealing with 
Huawei. The largest
|   damage is to ARM: China can no longer think of ARM as a reliable 
partner.  So China
|   will switch to RISC-V (there really isn't
|   a better choice).
| 
| 
| Sure, this means that Chinese R will focus on RISC-V. As will that of its 
military clients and
| other lesser-aligned countries such as Pakistan, Russia and Brazil. OTOH, US 
allies won't be
| sinking much into RISC-V for fear of running afoul of the same embargoes and 
sanctions that have
| hit AMD, TSMC etc. As a result you have Western R entrenching around ARM, 
such as the rec

Re: [GTALUG] free old tower computers

[D. Hugh Redelmeier via talk]

| From: D. Hugh Redelmeier via talk 

| - Dell Inspiron 530

Includes an ATI video card with DVI out.

| - Dell OptiPlex 780

Includes an ATI video card with DVI and HDMI out.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] free old tower computers

[D. Hugh Redelmeier via talk]

(This is an experiment.  I don't know if list members area interested in 
this kind of thing or find it a waste of list attention.  Let me know.)

- Dell Inspiron 530
  No HDD or SSD
  DVD burner
  Core 2 Duo CPU E6550 @ 2.33GHz
  3 GiB of DDR2 RAM (2 x 1G, 2 x 512M)
  tested: boots CentOS on an HDD that I will remove.

- Dell OptiPlex 780
  No HDD or SSD
  DVD burner
  Core 2 Duo CPU E7500  @ 2.93GHz
  4 GiB of DDR3 in one slot (3 empty)
  Tested: boots Fedora 39 from USB stick

Each has some kind of Windows license sticker.

Available for pickup near Yonge and York Mills.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] war story: preliminary adventures with RDP / Remmina / Flatpak

[D. Hugh Redelmeier via talk]

| From: Michael Hill via talk 

Thanks for this information.

| On Tue, Jan 30, 2024 at 6:47 PM D. Hugh Redelmeier via talk
|  wrote:
| 
| > Running Window in a VM would be great but I haven't figured out how to do
| > that.  Licensing is part of the issue.

| About seven years ago I needed to start using it full time, so I
| bought a Windows 10 Pro license and an SSD.

Yeah.  It is a matter of pride/stubbornness that I don't want to buy a 
second Windows license for the machine.  I already involuntarily paid for 
one.

| My understanding is your ThinkPad should store the Windows license key
| that came with it in a folder in Linux. I've seen mine but I haven't
| tried using it.

At one point, licenses were stored (partly) in the firmware.

These days, once you've initialized Windows, Microsoft records your 
"entitlement" in its cloud.

If you change enough things, Windows won't consider it the same computer 
and you have to buy a new license or argue with humans at Microsoft.

The VM doesn't look at all like the real machine to Windows.  Starting 
from the fact that the firmware is quite different.---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] RISC-V News

[D. Hugh Redelmeier via talk]

(Top posting because unmangling Evan's message is hard.)

For serious applications, the openness RISC-V helps but doesn't make 
everything you need open and free.  Or even available.  You actually need 
chip designs -- what SiFive sells.  You also need a lot of other modules 
for things like USB, PCIe, Power Management, ...

ARM has a vibrant ecosystem with all these things available for licensing.  
And ARM doesn't seem to be too greedy.  Even so, it has taken a long time 
to get ARM processors that match x86 at the high end.

So: if you want a short time to delivery, ARM is way ahead.

If you think more strategically, RISC-V has some advantages.

The US used a foot-gun on Huawei by banning ARM from dealing with Huawei.
The largest damage is to ARM: China can no longer think of ARM as a 
reliable partner.  So China will switch to RISC-V (there really isn't 
a better choice).  RISC V International is in Switzerland to try to evade 
US games.

Space stuff has long term horizons.  That's another area that has shown 
RISC-V interest.  But that's not a business that uses a large number of 
processors.  Space designs rarely feed back into the mainstream.

China is wary of buying from a US company like SiFive.  Chinese companies 
are developing their own expertise and products.  So SiFive is surely 
suffering from the above-mentioned foot-gun blast.

As a software guy, I don't actually have a horse in this race.  Linux runs 
on all these platforms.  I like "open" and RISC but my desktop is going to 
be x86 for some time.

There are RISC-V Single Board Computers in the Raspberry Pi space but they 
are inferior to the the Raspberry Pi line and other ARM-based SBCs.  
Mostly based on the open Alibaba processor designs


Seagate disks have RISC V processors but the consumer would never know.

ESP32-Cx chips/modules/boards have RISC-V processors.  Not powerful enough 
for Linux.

| From: Evan Leibovitch via talk 
| 
| Interesting.
| 
| Of course it is always useful to read beyond the cheery predictions.
| Buried under all the positive upward chart lines is the news (from the same
| publication) that a major RISC-V "pioneer" has just undergone layoffs (20%
| of engineering) and restructuring
| 

| .
| 
| - Evan
| 
| On Thu, Feb 1, 2024 at 11:23 AM Ivan Avery Frey via talk 
| wrote:
| 
| > https://www.eetimes.eu/navigating-the-risc-v-revolution-in-europe/---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] An anomaly with the `date` command

[D. Hugh Redelmeier via talk]

| From: Michael Galea via talk 

| Looks like its m=$((`date +%m` + 1)) for you!

Try that in December.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] An anomaly with the `date` command

[D. Hugh Redelmeier via talk]

| From: mwilson--- via talk 

| Discovered when I ran my script to run pcal and refresh my next-month
| calendar, and got March.
| 
| 
| mwilson@ningabel:~$ date
| Tue 30 Jan 2024 04:23:27 PM EST
| mwilson@ningabel:~$ date -d'this month' +%m
| 01
| mwilson@ningabel:~$ date -d'next month' +%m
| 03
| mwilson@ningabel:~$ which date
| /usr/bin/date

I never knew you could do this with date(1).  The man page punts to the
info pages (yuck).

I think that "next month" adds the length of the current month to the
time.  This is a 31-day month, we're quite late in it, so this hops
over all of February.

$ date
Tue 30 Jan 2024 06:46:05 PM EST
$ date -d'this month'
Tue 30 Jan 2024 06:46:10 PM EST
$ date -d'next month'
Fri 01 Mar 2024 06:46:17 PM EST

What you want to do is go back to the first of the month (all modern months
have a first) and then leap ahead a month.

$ date --date="$(date +%Y-%m-01) next month"
Thu 01 Feb 2024 12:00:00 AM EST
$ date --date="$(date +%Y-%m-01) next month" +%m
02

The phony natural language that date accepts is a hazard.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] war story: preliminary adventures with RDP / Remmina / Flatpak

[D. Hugh Redelmeier via talk]

We use Windows for tax programs.  Linux doesn't have these programs.  So 
we need to run Windows once in a while.

Dual-booting works fine but it isn't the most convenient setup.

Running Window in a VM would be great but I haven't figured out how to do 
that.  Licencing is part of the issue.

So: Remote Desktop (RDP) would seem to be a solution.

The RDP server (i.e. the computer running Windows, backwards from X) has 
to be running Win 11 Pro.  We have a computer that can do that.

There are several Linux RDP clients.  Remmina gets the best press.  So 
that's what I chose.

Remmina comes pre-installed on Fedora 39 (at least I don't remember 
installing it)/

My adventures so far are documented here:
 
 


Summary:

- Remmina didn't work

- Gnome Connections did work (but people recommend against it as a client)

- Remmina was installed as a flatpak

- when I replaced it with normal packages, it worked

$ flatpack remove org.remmina.Remmina
$ sudo dnf install remmina

If I were on Team Flatpak I'd debug why the flatpack version didn't work.
My hunch is that it may be because I'm not using the systemd-resolver.

More to learn.  Like what these diagnostics mean:

[11:08:55:918] [51240:52086] [ERROR][com.freerdp.common.settings] - 
[freerdp_settings_get_bool] Invalid key index 131
[11:08:55:918] [51240:52086] [ERROR][com.freerdp.common.settings] - 
[freerdp_settings_get_bool] Invalid key index 0

[11:08:55:114] [51240:52086] [WARN][com.freerdp.crypto] - Certificate 
verification failure 'self-signed certificate (18)' at stack position 0

[11:09:36:154] [51240:52086] [WARN][com.freerdp.core.rdp] - pduType 
PDU_TYPE_DATA not properly parsed, 562 bytes remaining unhandled. Skipping.

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Ongoing war story (currently issues of user trust)

[D. Hugh Redelmeier via talk]

[I seem to have forgotten to send this.]

| From: Peter King via talk 

| Can't say that I disagree with any of this.  I protested when the UofT decided
| to amalgamate all its services on Microsoft Server (to no avail), and even
| more so when they made it all but impossible not to use Outlook (after using
| mutt happily for years and years) - on the grounds that Outlook somehow had
| "more modern" security, which turned out to be doublespeak for "proprietary
| closed-source protocols" for accessing the mailserver that they now
| controlled.  Rewriting links and pushing their brand is the completely
| predicable result.

Hear hear!

I have (rarely used) courtesy accounts at U of T.  I just tested.
Mail to me at cs.toronto.edu isn't touched by Microsoft.  Postfix and
Exim only.

| Just recently I was told that the University would not allow me to ssh in to
| my office computer "because ssh had to be protected from the internet" (!),

Yuck.  I can still ssh into the one U of T CS system I tried.  I don't 
know the current nodes that general CS users are supposed to use (it's 
been that long).

| and instead I was supposed to use some binary blob to create a VPN into the
| UofT network

Open Standards!  They should use IPSec and not require binary blobs of
unknown safety.

SSH is good as long as passwords are not used.---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Ongoing war story (currently losing the battle)

[D. Hugh Redelmeier via talk]

| From: Peter King via talk 

| 
| The motherboard on the failing system is a non-UEFI Asus P6T.  The CPU is an
| Intel i7 950.  I have 32GB of Crucial DDR3 RAM in it.  The whole thing dates
| from 2009/2010 or so.  I'm pretty sure I replaced the motherboard at least
| once already.  There are four or five spinning disks of various sizes and
| ages.

Wow, that's an old processor.  Introduced in second quarter of 2009.
Intel says that the max RAM is 24 G.  The MB has 6 sockets -- an
oddity of that generation of Core processors.
  

  

This processor uses a lot of power: 130 W TDP.  Its fans probably make a 
significant noise.


| A few minor updates.
| 
| First, the problem remains the same: I never get through the POST, much less
| to the BIOS.  No beep codes (or beeps at all), no display, no nothing; it just
| remains silent as the fans spin.
| 
| Second, there seems to be power to the computer.  The internal MB power
| indicator lights up, the fans spin up, the hard drives seem to all spin up,
| and the graphics card at least lights up.

I wonder if one of the power supply "rails" is bust.  The power supply 
supplies several different voltages, all on different rails.  Sometimes a 
heavily-used voltage is supplied on more than one rail.

One place that old systems deteriorate is in the connectors.  Unplugging 
and plugging things in a few times can clean the contacts and sometime get 
them working again.

I have a power supply tester.  Who knows it it is useful?  Something like 
this (not a recommendation for this listing):
  
They are fairly inexpensive.
You could borrow mine if you were willing to pick it up and drop it off.
Oh: and I have to check if I can find it.  I'm near Yonge and York mills.
Alternatively you could check with a voltmeter (not as easy as you'd 
think).

| (Where "giving up" means looking for another computer I can migrate
| all/most/some of the existing hardware to -- maybe an old tower that has room
| for lots of spinning drives.)

How many spinning drives?  The more you have, the fewer systems are 
available.  Other than that, there are a million surplus systems that are 
cheap and faster than yours.  You might find too few SATA ports on a new 
MB.

The current trend is to put lots of drives in a NAS.  Some you can run as 
a JBOD (Just a Bunch of Disks).  That might be a good long-term strategy.  
Too bad they are so expensive.

We have a USB dock for bare SATA drives.  That's OK (not great) for 
accessing the half dozen or so hard drives with TV programs recorded by 
MythTV (recording stopped about 6 years ago).---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] anyone have this bell item?

[D. Hugh Redelmeier via talk]

I thought digital telephone signals were a big problem for you so I 
wondered if digital TV would be a similar problem.

| From: Karen Lewellen via talk 
|
| Not sure I follow while digital would be a problem for me question though?
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] anyone have this bell item?

[D. Hugh Redelmeier via talk]

I'm not convinced that it will work.  (I don't know that it won't.)

One clue is that Bell won't offer it to you.

Another is that the web site is for Bell's Manitoba Telephone System 
subsidiary.  There's a good chance they use different technology since 
MTS started as an independant company.

If it works, it has to connect "upstream" to another box with WiFi.  Will 
your house's Bell box inter-operate with this thing?

The TV signal will be digital all the way to the VIP2502. Is that a 
problem for you?


| From: Karen Lewellen via talk 
| To: talk@gtalug.org
| Cc: Karen Lewellen 
| Date: Mon, 22 Jan 2024 17:17:04 -0500 (EST)
| Subject: [GTALUG] anyone have this bell item?
| 
| hi all,
| Layered times, leaving me reading more than commenting.
| My landlord wishes to provide, but cannot seem to successfully secure from
| bell, the box below.
| 
| it's a basic bell cable box vip2502
| 
https://www.bellmts.ca/support/tv/fibe-tv/installation-and-setup/arris-vip2502-reference-guide
| Anyone have this sitting in a closet?
| Might be wise to confirm the idea will work.  This entire property, even
| outside, has bell fibe for home coverage including internet..But my on going
| issues with them have me a bit scared.
| Thanks,
| Kare
| Who has not given up on Over The Air, just gauging offered options.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Ongoing war story (currently issues of user trust)

[D. Hugh Redelmeier via talk]

| From: Peter King via talk 

| Well, more exactly, Microsoft Exchange Server insists on authentication being
| through XOAUTH2.

I don't know what XOAuth2 is.  Googling gets me things that would take too 
long to read.

If XOauth2 is the same as OAuth2 (I'm sure it isn't), this might be a good 
resource (less than 2 years old):


|  I far prefer the use of a console
| client (mutt) and to keep my email off the server and on my main computer.

I don't do this, so this is just guess work and suggested research
directions.  I run my own mail server and mail goes in and out of my
house using SMTP.  I then read and write mail with alpine(1), another
console client.

One could use fetchmail to fetch from imap and deliver to a local MUA.

- you'd need to figure out how to get fetchmail to use XOAuth2.
  Google results suggest that this has been done, perhaps with a
  patch to fetchmail.

- you'd have to set up local mail (Mail Delivery Agent) in some way.

- reading (local) mail with your choice of MUA (mutt) is easy.

- I don't know how you would handle outbound mail: that's not
  fetchmail's job.

  + can you still send mail with your mutt setup?  If so, I guess this
subproblem is solved.

  + you'd like a copy to be local

  + your MUA probably has a config for an SMTP server.  Can exchange
accept that?  Surely with some unknown-to-me authentication.

| I
| suppose I should see whether there is now some ready solution for mutt, and
| then figure out how to migrate all the email since switching out of
| Thunderbird and into my mutt archives

There are a few ways of storing mail in a user's files.  Mutt can
probably use a couple.  maildir and mbox are common.


That second claims Thunderbird uses the mboxrd variant of mbox.

| -- but too many battles at the moment
| getting hardware to work to take time out for reconfiguring software.

Yeah.

| If anyone knows a straightforward way to do either of these things (get mutt
| to work with Exchange and migrate from Thunderbird to mutt) I would be very
| pleased to hear about it!

I hope these notes help.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] war story: Brother printer won't turn on

[D. Hugh Redelmeier via talk]

| From: Evan Leibovitch via talk 

| I've had a CDP-L2520DW for at least a decade.
| Touch wood, no problems like this.

I've found Brother laser printers to be fairly reliable, especially 
considering how complex and mechanical they are.  I don't buy other brands 
so I cannot compare them.

I've bought several refurb Brothers from their web site.  Since the 
pandemic, the prices seem to be way higher.

Brother seems to allow clone toners. That's good.

Brother's say that the toner cartridge is empty a bit prematurely.  
Luckily there are sequences of key presses that override this.

| FWIW, I've found Brother tech support to be quite helpful

Good to know.  I've rarely found any tech support useful so I try them 
last.  Google is much better.  I wonder if chatGPT would be better still.

|  While I have no
| idea where she was located, she did know her stuff.

I always ask CSRs "what City are you in" as a kind of stamp collecting.  I 
do it at the end because it isn't important and I don't want to throw the 
conversation off.

| For those of us not intrepid enough, or lacking confidence in keeping a
| soldering gun steady -- I wonder if Brother might make available a
| replacement circuit board.
| More expensive than buying a single capacitor but less than replacing what
| has otherwise been a pretty good unit (from a company that has supported
| Linux very well).

I'd like to say that soldering is a basic life-skill, but it isn't. One 
could try Dropping in on the Hacklab open house (Tuesdays starting at 
18:00)? 

Ohh: I thought that the Repair Cafe had died but I'm wrong: 


The repair is pretty easy if you've done any soldering.  I solder about 
twice a decade, so I'm not very good at it.

I don't know what soldering guns are good for.  Probably not electronics.  
I use a cheap soldering iron I bought three to five decades ago.  
Probably from Radio Shack.  For electronics work, you want a low power and 
small iron.

The latest cool cheap soldering iron is the PINECIL V2.  It has a RISC-V 
processor and open source software.  Why, I don't know.  I'd have bought 
one but the shipping doubles the price and irks me.

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Ongoing war story (currently losing the battle)

[D. Hugh Redelmeier via talk]

| From: Peter King via talk 

I don't remember you telling us what the older (now also failing)
machine model is.  AMIbios from 2009 suggests to me that it predates
UEFI.  If so, it probably isn't worth investing any time to fix it.
Publish or perish (unless you have tenure)!

(This system's problem could be as simple as a broken power switch.)

A notebook or a miniPC might be a good stopgap.  You get to tune that
suggestion to match your requirements.  Shopping too hard will waste
your time too.

A mini PC ought to just plug into your exiting desk setup.  When you
are no longer using it, it takes little space in you cupboard,
bookshelf or the back of your desk.  It is easy to stick in a bag to
carry to another site.

These BeeLink Mini PCs look OK to me but are not as deeply discounted
as they have been.  You might well prefer Minisforum.  Support isn't a
strong suit (but then your experience with Lenovo has been
disappointing).

All these are from amazon.ca.  The listings have coupons that come and
go so the prices wander.  And there are so many listings of the same
devices from the same vendor!  Some have an additional
10%-of-list-price coupon that I cannot see because I've used mine up.


$367 Ryzen 5 5560u 16G RAM / 500G NVMe SSD
This is a notebook CPU.  Low power in both sense.
Easy to update RAM and SSD but you cannot add, only replace.


$329 after coupon; same as above.


$570 after coupon Ryzen 7 5800h 32G RAM / 1T NVMe SSD
This ups each of the specs in the previous box.


$749.99 after coupon Ryzen 7 7840HS 32G RAM / 1T NVMe
This adds a better processor with a better iGPU and faster RAM.

Fun alternative: Best Buy has the Asus Ally gaming console, open box,
for $549.99.  It has essentially the same guts as the 7840HS system
but only 16G RAM and 512G SSD.  It has only one port (USB C / USB 4 /
power) but a docking station fixes that.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] war story: Brother printer won't turn on

[D. Hugh Redelmeier via talk]

My LASER printer / scanner, a Brother MFC-L2729DW, would not power on.
Normally something like this is due to a power supply failure, but I
googled and found this:



(The video is incomplete; there is another one that is linked from the
first, but all the useful info is in the first.)

Following his instructions, I found that I could reset the machine to
work.  But he said that it would fail again on a power failure or even
just turning it off.  Furthermore, doing this too many times breaks the 
printer.

It turns out that settings are kept in a RAM that is powered by a
supercapacitor (instead of a battery).  Furthermore, that
supercapacitor wears out.  All you need to do to fix the printer is
to replace the supercapacitor.

The video claims all recent Brother printers have these
supercapacitors.

The video shows you how the replace the supercapacitor.  You need a
new supercapacitor, which he will sell you.

Armed with this information, I

- bought a suitable supercapacitor from digikey.ca (cheaper and faster
  than ordering from Australia).  Actually I bought a slightly smaller 
  one, but that seems to be OK.  The original was 0.33F, 5.5V.
  I paid $3.48 for two, plus $8.00 shipping, plus HST.

- pulled apart the printer, as per video

- hacked out the old supercapacitor and soldered (badly) the new
  supercapacitor.  (Doing the job right would have required removing
  the circuit board.  Instead, I left the old supercapacitor's legs
  and soldered the new one's legs to the old ones.)

- struggled to get the plastic panel back in place.

It works.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Doctorow: "Sympathy for the spammer"

[D. Hugh Redelmeier via talk]

| From: mwilson--- via talk 
| Subject: Re: [GTALUG] landline power [was Re: "AI" on getting correct
| technical answers]

| Cory Doctorow posted a relevant article today:
| 
https://pluralistic.net/2024/01/15/passive-income-brainworms/#four-hour-work-week
| 
| Money quote:
| "while we're nowhere near a place where bots can steal your job, we're
| certainly at the point where your boss can be suckered into firing you and
| replacing you with a bot that fails at doing your job"

Thanks for the pointer.  A great article!

I recommend that everyone read it.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] landline power [was Re: "AI" on getting correct technical answers]

[D. Hugh Redelmeier via talk]

| From: Alvin Starr via talk 

| Bell and Rogers are now both offering VOIP based home phone services.
| I assume that they have batteries to keep things running in the event of a
| power outage but It would be interesting to have someone on list confirm that.
| I remember many years ago working with an ISDN ATA device from Bell that had
| NiCad batteries that did not last all that long and had real degradation
| problems.

They try to hide the fact that their "home phone" service is over
VOIP.  For one thing, I think that they have dedicated bandwidth so
saturating your internet service won't break your phone service.

Rogers has or had UPS built into the home phone box.  And they have
batteries on their neighbourhood boxes.  (During a long power failure,
they even brought a generator for the one near us.)

Bell had a UPS built into their home internet+phone boxes.  But not
the latest ones.

(The old Bell system had large lead-acid batteries in the COs.  Old
handsets were actually powered by the CO.  Modern ones have their own
power for many functions.  So we used to expect the phone to work
during "hydro" failures.)
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] mailing list convention

[D. Hugh Redelmeier via talk]

There is a long standing convention for this list.

1. when replying to a post, only quote the parts you are responding to.
   Why?  Because this makes it easier for the reader and there are
   more readers than writers.

2. your response should immediately follow the part of the quote you are 
   responding to.
   Why?  That makes a linear reading work.

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] "AI" on getting correct technical answers

[D. Hugh Redelmeier via talk]

| From: o1bigtenor via talk 

| > I have never used a cellphone to access https://chat.openai.com -- this is 
where I signed up for free and later where I upgraded to Plus. The process to 
register is quite easy, I don't quite understand the issues you and Kevin are 
having.

| One doesn't use a cell phone (well I wouldn't) but a cellphone number
| is demanded to
| verify one's identity. Without a working cellphone connection one is
| today - - - a NON-person.

This is "two-factor authentication" (often called 2FA).

Lots and lots of things require it.  And then assume that that number
is a good way to contact you.  Grrr.

SMS is not a great second factor.  Especially in the cases where there
is no first factor.

A related issue: places that ask my phone number and then assume it
works for SMS.  My phone number dates back more than forty years and
it is a landline (no SMS).

The thing is that the world has changed faster than some of us have
adjusted.  I feel that way about many things; at least Flash is gone.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] AI - Llama 2 an open source AI that can run on a Raspberry PI

[D. Hugh Redelmeier via talk]

| From: Colin McGregor via talk 

Interesting report!

Some nit picking of the answers.  Just from my memory -- there are
surely other bugs.  Summary: don't trust these models!  They don't do very 
basic fact-checking.

| > Who was Canada's best Prime Minister?

Why are all the answer Liberal prime ministers (except for John A.)?

| 1. Sir John A. Macdonald (1867-1873, 1878-1880):

|  and he also negotiated the
| Northwest Passage Agreement with the United States, securing Canada's
| western border.

Google finds no "Northwest Passage Agreement".

| 2. William Lyon Mackenzie King (1921-1930, 1935-1948):

Not exactly.  He lost power in 1926 June and regained it in 1926
September.  Remember the King-Byng constitutional crisis?

| King was a
| skilled diplomat and leader who guided Canada through two major
| crises: World War I

Really?  He wasn't even a member of parliament doring the Great War.

| 5. Jean Chrétien (1993-2003): Chrétien is credited with helping to
| stabilize Canada's economy and promote international trade. He also
| played a key role in the Belfast Agreement, which brought an end to
| the Troubles in Northern Ireland. Additionally, he implemented
| policies that helped to promote diversity and inclusion within
| Canadian society.

I don't remember Cretien being part of Belfast Agreement (better known
as the Good Friday Agreement).  Retired General John de Chastelain was
one of the three commissioners of the Independed International
Commission on Decommissioning.

Off the top of my head, there are a couple of high points in Cretien's years:
avoiding the Iraq War and imposing election financing limits (a poison pill
for Martin, but still a Very Good Thing; contrast with US!).

Cretien and Martin are jointly credited with drastically and painfully
improving Canada's finances in the 1990's

| -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
| 13b
| -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
| > Who was Canada's best Prime Minister?

| 5. Jean Chrétien (1993-2003): Chrétien's tenure saw significant
| economic growth, the repayment of Canada's national debt, and the
| negotiation of several major international agreements, including the
| Free Trade Agreement with the United States.

No: the FTA was Brian Mulroney.---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] lazy jail server admin forced to act

[D. Hugh Redelmeier via talk]

| From: Alvin Starr via talk 

| SPF should insure that only your email servers can be used to send mail from
| your domain.

Related:

My Postfix server validates HELO / EHLO records.  These records declare 
the name of the server contacting my server.  Postfix makes sure that the 
declared name resolves to the IP address from which the server was 
contacted.  At least that's what I think it does.

From /etc/postfix/main.cf:

smtpd_helo_restrictions =
permit_mynetworks,
reject_invalid_helo_hostname,
reject_non_fqdn_helo_hostname,
reject_unknown_helo_hostname

This catches a lot of SPAM.  But it also blocks a few badly configured 
servers that I want to get messages from.  The ones that I've noticed:

- some parts of Sunnybrook hospital

- rakuten

I have been unable to contact the technical people at these sites to get 
them to fix their problem.


| There was once a guy who insisted on running an open relay and believed that
| anybody blocking open relays were infringing on his rights.
| He was VERY loud about his convictions that open relays were not the problem,
| spammers were.

John Gilmore?  He's generally soft-spoken but has strong convictions.


Hugh Daniel?  Sadly gone.  But he definitely was loud.


BTW, both are/were strong forces for good.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] lazy jail server admin forced to act

[D. Hugh Redelmeier via talk]

| From: Kevin Cozens via talk 

| You probably need to add a google authentication record to your domain name to
| reduce the chance of your email messages being rejected by gmail.

I don't know what that is.

Googling responses suggest you might be talking about a Google
Worspace thing.  I don't have a Google Workspace.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] video: The Dark Side of Open Source

[D. Hugh Redelmeier via talk]

This video was recommended to me:

Chris Titus Tech: The Dark Side of Open Source


Apparently Titus recommended Thorium, a mod of the Chromium browser.
Now he feel burned because of a couple of non-mainstream Easter eggs.

It seems mostly overwrought silliness to me.  But you can decide for 
yourself.

The story isn't really about open source.  It is about trust and 
verification of software.  The bigger / more complex the object, the 
harder it is to trust.  A very very deep problem.

How does open source relate to this?

- (we think that) it is harder to sue an open source project than a 
  commercial software producer.

- the infrastructure for open source (GitHub, for example) lets you build 
  and distribute new mixes things without a lot of effort.  So one oddball 
  can create and distribute a useful system

- a larger team, needed in the past, would probably have an average
  weirdness that is less than some random single creator.

- open source software can be examined.  This is likely how the
  "problems" with Thorium were discovered.

I don't even know why Thorium was interesting.  It is a hacked version
of Chromium.  Are the hacks interesting?  Apparently its main
advantage is that it is compiled with higher optimization.  If they
judged it worth doing, the Chrome project could do this itself.  As
could the distros that package Chrome or Chromium.

The only browsers that I (reluctantly) trust enough to use are
FireFox, Chrome, Chromium.  Links or Lynx when desperate.
Browser-of-the-month isn't a club for me since the browser is my main
exposure to security threats.

There is a very interesting question here: how can software earn trust? 
Any software, including open source software.

A recent enthusiasm has been to implement procedures to prevent "supply 
chain attacks". Things like "software bills of materials" (provenance of 
components).  The (deserved) whipping boy has been NPM, the repo for open 
source JavaScript.  Equally scary things exist for Python, Perl, and Rust, 
for example.

The Thorium browser problem could be classified as a supply chain problem.

Reliable software is hard.  We have to work on it any way that is 
effective.

PS: I'm looking at Titus' video recommending Thorium in the first place.

- He gushes about how much faster it is than Chromium and Chrome.
- He suggests that the author has added accelerators not in chromium.
- A few nice little things.
- He mentions "multi-threading improvements" which seems unlikely.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] lazy jail server admin forced to act

[D. Hugh Redelmeier via talk]

| From: D. Hugh Redelmeier via talk 

| How to read this:

|   mx:
|   email sent by mimosa.com should only come from its servers
|   declared in MX DNS records

More accurately, this means that mail that says it is from mimosa.com is 
OK if it comes from one of the SMTP servers listend in MX records.

In other words, this test only marks good things.  Then the "~all" says 
anything that isn't good is bad.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] lazy jail server admin forced to act

[D. Hugh Redelmeier via talk]

| From: John Sellens via talk 

| You may find that you need DKIM and DMARC as well. If you’re using 
| postfix it’s fairly easy to hook opendkim in.

Yeah.

| My biggest hurdle was trying to find clear concise guides. 

Links?---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] lazy jail server admin forced to act

[D. Hugh Redelmeier via talk]

Mails from my domain have started to be rejected by gmail.
To placate gmail, I've added an SPF record to my doman:

mimosa.com. IN  TXT "v=spf1 mx ~all"

How to read this:

v=spf1:
this TXT record is an SPF version 1 record

mx:
email sent by mimosa.com should only come from its servers
declared in MX DNS records

~all:
any other mail from mimosa.com is bogus.
(~ and - are slightly different.)

That seemed pretty painless.  I guess I should have done this a decade 
ago.

I hope this works.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] meeting Tuesday?

[D. Hugh Redelmeier via talk]

We are supposed to have a meeting Tuesday but I have not seen an 
announcement.

The venue might be changed from our Big Blue Button room.  Our host, LPI, 
was set to switch from BBB to some NextCloud facility.

If we don't get further directions, try 
  
at 19:30 on Tuesday.
There's plenty of stuff to chat about.

PS: this is partly a probe of the mailing list.  It went offline without 
notice last month.  I was able to restart it when I found out about the 
problem.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] short window on a great deal for a mini PC

[D. Hugh Redelmeier via talk]

For me, it worked out to be $293.39 + tax for a Ryzen 7 5800h system with 
16G RAM and 500G SSD.

This is a "Lightning deal" so it will be over soon (but I can no longer 
see when).

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Exim vs Postfix [was: Local-only mail server]

[D. Hugh Redelmeier via talk]

| From: Giles Orr via talk 

| "exim4" is usually (not always) installed on Debian systems.

| From: Anthony de Boer via talk 

| Exim4 would be the right solution on whichever host you designate your
| mailserver.

| From: Michael Galea via talk 
| 
| On my mail server (which runs an exim4 smarthost) my /etc/aliases resembles:

I used to use Sendmail but switched to Postfix some time this century.  
Why do you choose Exim instead?  Is it just that Exim is the default on 
debian?

Postfix tries to be a Sendmail replacement, but better.  That made the 
transition easier.

Security-by-design has been a focus of Postfix.  I seem to remember at 
least one disastrous security problem with Exim.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] queue unjamming [was Re: Meeting tonight?]

[D. Hugh Redelmeier via talk]

| From: Lennart Sorensen via talk 
| 
| On Sun, Dec 17, 2023 at 02:36:05PM -0500, D. Hugh Redelmeier via talk wrote:
| > There was a gathering at the site of what would have been a meeting.
| > No meeting was organized.  This lack was not intentional.
| 
| An in person gathering.  Haven't been to one of those in a while.

By "site" I meant <https://blue.lpi.org/b/eva-zjc-gjy-kgl>.
Same as always since the start of the pandemic.
Beware: this is scheduled to change before the next meeting.

We could go back to TMU but that would require a doorman to let 
stragglers into the building and the Board thinks that's too onerous.
(I think that it would be manageable but I'm not yet going to in-person 
events.  BTW waste-water numbers for the COVID virus are quite high.)

| Maybe once I finish moving I will try to make it down again.  18 days
| until closing.

Good luck!

| So next time it is stuck, email you rather than ask on discord.
| Got it. :)
| 
| I even considered doing that.

Yes.

Or do both in case I've been replaced.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] queue unjamming [was Re: Meeting tonight?]

[D. Hugh Redelmeier via talk]

| From: Anthony de Boer via talk 
| 
| > On Dec 17, 2023, at 12:44, Lennart Sorensen via talk  
wrote:
| > 
| > Is there a meeting tonight given I think it is the second tuesday of
| > the month?

There was a gathering at the site of what would have been a meeting.
No meeting was organized.  This lack was not intentional.

| Certainly seems to be Sunday today…
| 
| > Wasn't sure if an annoucement was stuck in a queue somewhere waiting to
| > be permitted through.

No there was no announcement.

| And it does appear that a great queue-unjamming has occurred.

I hadn't seen any message newer than December 8 so I posted one.  When I 
didn't see it come out on the list, I rebooted the machine 
penguin.gtalug.org. That unstuck the system.

I should have tried to figure out how it got stuck but we intend to 
replace the system soon so the knowledge wouldn't be useful for 
long.---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] tiny GNAME change that I don't like

[D. Hugh Redelmeier via talk]

The latest version of GNOME is included in the latest Fedora and Ubuntu.

(It's hard to talk about this GUI stuff because I (and probably you) don't 
know the names of visual components.)

On the top bar of the screen, "Activities" is on the far left.
Next to it used to be the name of the application that had focus.
If you clicked on that name, you would get a list of instances of that 
application, plus a couple of other potentially useful choices.

When I have a lot of windows for a particular application, I sometimes use 
this to choose the instance I want to have focus.

This feature is gone.  You can still get a similar effect using 
Windows-Tab and Windows-DownArrow, but it is kind of awkward having to 
hold the Windows key.  Furthermore, I find the description less useful.

Am I the only one who cares about the loss of this feature?

The word "Activities" has been replaced with a horizontal oval and a dot.  
I don't know what that change signifies.  I don't even know how to 
interrogate GNOME about this.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] ot: sort of, is it really impossible to get real cable anymore?

[D. Hugh Redelmeier via talk]

| From: Evan Leibovitch via talk 

| Synopsis: I think you've just re-complicated something that took a while in
| this thread to simplify.
| And buying a new TV is not considered an option, the roughly $100 solution
| being proposed is expensive enough.

Yeah, if the problem is JUST to get OTA TV, almost all of what I said is 
just noise.

Sorry, everyone, and especially Karen.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] ot: sort of, is it really impossible to get real cable anymore?

[D. Hugh Redelmeier via talk]

I have some concerns, but am somewhat uninformed.

I've run out of time to edit this.  I apologize that it so is long and 
disorganized.

General concern: the complex chain of devices has many points of 
failure.  Getting it to work might be difficult.  Keeping it working might 
be difficult.

General concern: many components these days expect navigation of on-screen 
menus.

General concern: some HDMI signals (eg. from Rogers or Bell) are 
"protected" by HDCP.  The general idea is that you are presumed to be a 
pirate and only approved devices are trusted by Rogers or Bell (or Blu-Ray 
players, ...).  Interesting point: if HDCP authentication fails, I think 
that a low resolution signal is still allowed -- maybe good enough for 
Karen.

What are all the video sources that you want to connect to your TV?

- we've been talking about OTA (Over The Air) TV signals

- we've mentioned Rogers Cable TV signals.  Or, perhaps Bell TV signals.
  Those are a problem now since both companies want to give you an HDMI 
  signal, protected by HDCP.  Both want the "smarts" in a Set Top Box that 
  they provide you.  The STB must be manipulated through a complicated 
  on-screen menu system and a remote control.  Gone are the days when a 
  raw cable signal can be "tuned" by a third party device (eg. TV set).

- an additional "modern" way of using Rogers or Bell has been extended to 
  using an App on an Android or iOS box (or Rakuten or Fire OS).  These 
  seem to require on-device visual menus but they MAY have accessibility 
  features.

  In this case, the signal is "streamed" to the box through the internet 
  (IP).

  Those boxes may have their own display (think smart phone, tablet, or 
  smart TV) or output to a TV (eg. Apple TV, Fire TV Stick, ...) via HDMI.

- Is your computer expected to be an input?  I assume that it produces 
  VGA.  That can be converted to HDMI with a simple dongle, if that turns 
  out to be useful.

- Are your various assistive devices expected to be inputs?  What are 
  their output signals?

About converter boxes being discussed:

- The Mediasonic HOMEWORX HW130STB seems to have on-screen menus I take it 
  that this will be very difficult for Karen There is a chance that only 
  setup requires on-sheen menus

- An ATSC => NTSC converter will only work for over-the-air signals.

- Alternately, it may be possible to use an HDMI => NTSC Channel 3
  converter (yuck) or HDMI => composite converter.

  This does NOT include a tuner.  You'd then need to get a tuner with the 
  right output (HDMI)

  If you have multiple HDMI sources you would need an HDMI switch feeding 
  into the HDMI => NTSC converter
  
- a DVD or Blu-Ray player?

All these problems go away if you can find a new monitor or TV set that 
could work for you.  Of course there is a new set of problems:

- a new TV might cause you sensory problems (that might not be the right 
  term but I think you know what I mean)

- new TVs are mostly "smart" in a way that is going to be a problem 
  (complex on-screen menus and many states).
  There are a few "dumb" TVs still available but even they might not be 
  dumb enough.

- Your best bet might to be to get a computer monitor: they usually have 
  fewer settings.  You might want one with speakers or audio-out of some 
  kind.  Most modern monitors have HDMI-in; some only have DisplayPort, 
  but that's almost trivial to convert at FullHD resolution (1920x1080 or 
  perhaps 1920x1200) or less.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] article about Window Snyder [was Re: Protonmail WAS: Re: Linux Unicorns]

[D. Hugh Redelmeier via talk]

| From: Giles Orr via talk 

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

Interesting post.
Interesting article.

The article linked to another that I found very interesting.  I'd never 
heard of Window Snyder before (Window is her first name and has nothing to 
do with MS Windows.):

?
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Linux Unicorns

[D. Hugh Redelmeier via talk]

| From: Evan Leibovitch via talk 

| On Fri, Nov 24, 2023 at 12:19 PM D. Hugh Redelmeier via talk <
| talk@gtalug.org> wrote:
| 
| It seems to me that Linux has won and been commoditized on the server side.
| > No high-level discussion needs to address this.  Several distros are good
| > enough and almost interchangeable in abilities.
| >
| 
| Agreed.
| 
| The whole CentOS/Rocky/Alma/Oracle/SUSE/RHEL dustup has brought some
| instability to the realm (and for those of us old enough, bears remarkable
| resemblance of the old SysV/OSF1 wars). Ditto some of the jerk moves
| Canonical seems to be doing to leverage the Ubuntu base. But I think that
| these instabilities will settle down, even if they only evolve into
| stalemates.
| 
| On the PC side, if you don't wish to demand a dedicated machine, you
| > have to fit into what the market has.
| > 1. Windows of various vintages
| > 2. MacOS.
| > 99. Linux.
| >
| 
| People want virtual dedicated desktops, for something other than UI
| testing? Uh, OK.
| Where do Chromebooks fit into this?

I was talking about an application publisher.  If they want to run on the 
user's existing PC...

| Old Unicorns came from different times.  25 years ago, it wasn't obvious
| > that Linux would be where it is (and isn't).
| >
| 
| Speak for yourself. 
| 
| I was kinda certain that Linux would overtake Unix early in its life; that
| feeling was confirmed when the first Beowulf clusters came online and DEC
| imploded in the late 90s, and fully cemented when Oracle bought Sun less
| than a decade later. The increasing commoditization of PC components
| combined with the neverending balkanization of Unix hardware increasingly
| made the progression to dominance inevitable, the main issue was how long
| it would take.

25 years ago, it was clear Unix was not much of a contender.  And Windows 
NT looked like a winner.  You could choose to build a Unicorn on Windows 
in the cloud.  I think you'd be nuts to use Windows now.

| Netflix uses NetBSD for at least some stuff.  Making a decision like that
| > today would take serious conviction.
| >
| 
| BSD is still a good choice for some, so long as they're OK with
| self-support. It failed to get Linux's mindshare for two unrelated reasons:

So you are agreeing, right?

|- They got caught up in their own little petty Unix wars in the 90s,
|just when the world was looking for some platform stability (that Microsoft
|and Apple were providing on desktops). Going with one Linux distro over
|another, especially on the server side, was far less of a critical decision
|than choosing Open- or Net- or FreeBSD.
| 
|- Also, the GPL enabled big vendors to contribute code without a
|competitor being able to take it and re-lock it proprietary, as Apple
|famously did to BSD-licensed code in Project Darwin. The GPL enabled the
|old Ray Noorda philosophy of "co-opetition" which BSD could never match
| 
| 
| 
| > | Does anyone on the list actually know (as in personal experience)
| > | whether/which (or even if) most tech unicorns are Linux based or with a
| > | Linux origin story?
| >
| > That may be interesting history but I doubt that it is worth consideration
| > for a new business.
| >
| 
| Agreed. Linux is infrastructure, not a factor in whether an effort gets
| funded or expands.

Or even work.  In other words, it isn't an innovation any longer.

| More interesting is to figure out how to use the cloud without being tied
| > to a particular vendor.
| >
| 
| 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/>
| .

Yeah, I was grasping at straws.

There is room for innovation in infrastructure but it is hard to
become a unicorn there since the area is mostly occupied by giants:

- Amazon

- Microsoft

- Google

- Oracle...

Being cloud platform agnostic is smart.  How you do it isn't something
I know much about.

I vaguely remember that Netflix can switch platforms.  (The few tech
people that I know of there are fairly impressive.)  That's probably
very useful when negotiating with their cloud providers.---
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

[GTALUG] cheap cheap Virtual Private Servers

[D. Hugh Redelmeier via talk]

This is the time of year (black Friday / Cyber Monday) when you can get 
very cheap VPSes.  Not necessarily good, but cheap.

I have a couple that cost about US$10 or so a year.  Sometimes useful for 
pinging.  You can do actual useful things with them, but the resources are 
fairly constrained.


---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] GNOME's "Attach Modal Dialogs"?

[D. Hugh Redelmeier via talk]

By default, when a GNOME application pops up a dialog, it gets glued to 
(one of?) the application's window: the dialog sits in front of the 
Window, glued in place.  If you move the dialog, the window moves with it.

I'm not sure why this is considered useful behaviour.  Perhaps because if 
you misplace the dialog, you might not know that the application is 
waiting for a response (I've had that happen in Windows).  But it means 
that the dialog might hide something you need to see on the main window to 
fill in the dialog.

At least on Fedora, you can change this behaviour using "GNOME Tweeks".
Under Windows, set Attach Modal Dialogs to "off".

Does anyone else like this default?


---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] mbox vs Maildir [was Re: Linux friendly email providers?]

[D. Hugh Redelmeier via talk]

| From: Howard Gibson via talk 

|I figured that separate email files would be using up a lot of
| sectors, but my bandwidth is limited by my Blu-ray discs.  My actual
| backup is a gzipped tar file.  Sectors should not be a problem, should
| they?

Uncompressed TAR format: 


tar uses blocks.
The default block-size is 512 bytes (that dates back to 7th edition UNIX 
in the late 1970s) (for many applications you want to use a larger block 
size but not for maximal compression).

Each file uses one header block plus the number of blocks to hold the 
contents.  So maildir is probably significantly worse than mbox if no 
compression is used.

I expect any compression technique to be fairly good at compressing a run 
of 0 bytes.  Such bytes are expected for filling out each header block and 
the last block of each file.  I cannot quantify "fairly good".

mbox has none of that per-message overhead, compressed or not.

Without testing, I cannot say how much worse compressed tar files of 
maildir would be, compared with compressed tar files of mbox.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Linux Unicorns

[D. Hugh Redelmeier via talk]

| From: ac via talk 

| I have been wondering about tech unicorns and Linux.

It seems to me that Linux has won and been commoditized on the server
side.  No high-level discussion needs to address this.  Several distros 
are good enough and almost interchangeable in abilities.

On the PC side, if you don't wish to demand a dedicated machine, you
have to fit into what the market has.
1. Windows of various vintages
2. MacOS.
99. Linux.

On the mobile phone and tablet side
1. iOS (not the most users but the most money)
2. Android.
3. Web App of some kind

On the dedicate device side: an engineering choice.  Embedded Linux if
the device is beefy enough.

All these decisions are easy.

On the device side, it is just an engineering decision.

Old Unicorns came from different times.  25 years ago, it wasn't
obvious that Linux would be where it is (and isn't).

Netflix uses NetBSD for at least some stuff.  Making a decision like
that today would take serious conviction.

| Does anyone on the list actually know (as in personal experience)
| whether/which (or even if) most tech unicorns are Linux based or with a
| Linux origin story? 

That may be interesting history but I doubt that it is worth
consideration for a new business.

More interesting is to figure out how to use the cloud without being
tied to a particular vendor.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] mbox vs Maildir [was Re: Linux friendly email providers?]

[D. Hugh Redelmeier via talk]

| From: Ron / BCLUG via talk 

| I've seen mbox files get corrupted (all mailbox
| messages in one file, and a line like "From: " is the message delimiter.
| Terrible!)

I don't remember seeing that corruption in the last few decades of using 
mbox. The horrors of in-band signalling are well known -- maybe the 
software I use reflects that knowledge.

| I've recently switched to using Maildir format (server *and* Thunderbird).
| One message per file.

You mentioned that you were running out of space on your system.  If a lot 
of that space is mail messages, I would bet that Maildir is costing you a 
lot of it.  Each message is taking a multiple of the allocation unit size 
(1KB? 4KB?) and a large part of that is likely unused (the tail of the 
last unit).

My intuition would be that since mail messages are usually short, and the 
distribution of sizes isn't uniform, you are probably using at least 25% 
more disk space with Maildir.

But intuition is surprisingly bad for computer things.  With Thunderbird 
conversions, you could easily measure this for a real-world example

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] issues with "fractional scaling" for displays

[D. Hugh Redelmeier via talk]

In GNOME (and surely other systems) you can specify that you want the
contents of the display scaled.  For example, 200% scaling doubles the
size of everything on the display in both dimensions.

The control for selecting scaling is in GNOME's "Display Settings", under
"scale".  Easiest way to get there: right-click on empty part of the
desktop and select "Display Settings"

This is useful if you have a HiDpi (high resolution) display and low 
resolution eyes.  Note: this is different from changing the display
itself to a different resolution.

Scaling by integral amounts has been available for a long time.  But it 
isn't that useful.  It would be better if the system could handle scaling 
by non-integral amounts like 125%.

For example, I have a 13.3" with 2560x1600 resolution.  100% scaling
is very fine.  Sometimes 150% scaling is easier to read.

Wayland allows this non-integral scaling.  Yay!
XWayland copes poorly with this: everything gets a bit fuzzy, even at 
integral scaling points (but perhaps not at 100%).
XWayland is used by programs that use the X Window System APIs.

(I don't even know which programs that I use are using XWayland so I
don't know how badly this would affect me.)

I was looking forward to fractional scaling in Fedora 39 / GNOME 45.
But it seems to be turned off, again:
See 

It appears that the Fedora folk think that:
1. fuzziness at integral scaling in XWayland is unacceptable
2. making a clear user choice "fuzziness is OK" isn't reasonable

Fractional scaling has been available before this with a magic
incantation.  I haven't tested it with Fedora 39.

 $ dconf write /org/gnome/mutter/experimental-features 
"['scale-monitor-framebuffer']"

This only takes effect after the next login.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] [GTALUG-Announce] Membership Dues

[D. Hugh Redelmeier via talk]

This message from a year ago is still/again relevant.

Summary:

1. Please join GTALUG.  I quoted Warren's message explaining how to
join and pay for membership.

2. Consider volunteering to help run GTALUG.

From: D. Hugh Redelmeier via talk 
To: Toronto Linux Users Group 
Date: Sun, 8 Oct 2023 15:11:30 -0400 (EDT)
Subject: [GTALUG] GTALUG meeting 19:30 Tuesday, Oct 10

(This is not the official announcement.)

We will meet virutally, via Big Blue Button: 
<https://blue.lpi.org/b/eva-zjc-gjy-kgl>

The first few minutes will be our Annual General Meeting.  This is where 
the board presents the financial statement for the past year and where we 
the members of GTALUG elect the board of directors.  There might be more 
business.

Please volunteer to stand for election: we can use your energy and ideas.
No experience is necessary.

To be a member, you have to have paid membership dues for this year.
I have appended a copy of Warren's helpful email from last year.

After the short AGM, we'll have a regular meeting.
As I understand it, Mike Hoye has agreed to give our October talk.
Mike is a great speaker and I look forward to this.
I have not heard what his topic will be.

| From: Warren McPherson via talk 
| To: annou...@gtalug.org
| Date: Thu, 24 Nov 2022 23:43:11 -0500
| Subject: [GTALUG] [GTALUG-Announce] Membership Dues
| 
| With the agm coming up, now is a great time to renew membership and make a
| small twenty dollar contribution to the organization.
| For those looking forward to the chance to wind their way through the
| Interac E-Transfer process again here is my summary of how it works.
| (Actually I forget each time so the novelty never wears off -- luckily it
| is easy.)
| 
| Log in to your bank and look for "Interac e-Transfers" you may need to go
| to the "Transfers" page and then you will see this option.
| 
| Select "email method" and use the email address:
| members...@gtalug.org
| 
| Putting your name in the message can be helpful as the notification I get
| can sometimes be a touch obtuse.
| 
| Thank you to those who have already contributed!
| (I know who you are.  Ha ha)
| 
| Thank you,
| Warren
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] growing an EFI System Partition (ESP)

[D. Hugh Redelmeier via talk]

Fedora 30 was released today.  In the release notes, it talks about EFI 
sizes:


Make the EFI System Partition at least 500MiB in size

The minimum size of the EFI System Partition (ESP) created by Anaconda
has changed from 200 MiB to 500 MiB. The maximum size, which is used
in most cases, remains at 600 MiB.

The reasons for this change include:

 * This partition is used to deploy firmware updates. These updates
   need free space of twice the SPI flash size, which will grow from
   64 to 128 MiB in near future and make the current partition size
   too small.

 * The new minimum is identical with what Microsoft mandates OEMs
   allocate for the partition.

See also:

https://fedoraproject.org/wiki/Changes/BiggerESP

https://github.com/rhinstaller/anaconda/pull/4711

https://github.com/rhinstaller/anaconda/pull/5081

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] growing an EFI System Partition (ESP)

[D. Hugh Redelmeier via talk]

(This is a recipe.  If you find errors or think of improvements, please 
let me know.)

== context ==

The ESP is the partition to hold anything the UEFI firmware can 
run.  Mostly bootloaders but also diagnostics and firmware updates.
It is a FAT filesystem in a partition with a special type.
In a running Linux, it is usually mounted on /boot/efi.

This stuff is a little bit tricky and not well-documented.  I tried to be 
very detailed to make sure you don't get lost.

== motivation ==

The ESP doesn't have to be very big, but sometimes it isn't big enough.

I often buy computers pre-loaded with Windows -- typically cheaper than 
computers without Windows.  Sometimes the ESP is 100MB.  This has pinched.  
So I want to increase the size.  I have a feeling that 200MB would be a 
good size, but I will probably settle on 400MB.

== naive method and a failure ==

Tools:

- a live USB Linux system.  [You don't want to be doing partition surgery 
  on the disk from which you are running.]

- some place to temporarily hold the contents of the ESP.
  It could be on another partition of the same disk or on a USB devicee.

- gparted for partition manipulation.  You could substitute with something 
  like parted(8).

- gnome "disks" (gnome-disk-utility).  I find it convenient for mounting, 
  unmounting, and looking at the properties of disks and partition.  You 
  can choose to do these things with other tools.

1. boot the live Linux stick.

2. if your live system doesn't include gparted, install it.
   Fedora: "sudo dnf install gparted"
   debian or Ubuntu: "sudo apt install gparted"

3. run gparted and point it at the disk with the too-small ESP.
   Make sure that none of that disk's partitions is mounted.

3. shift and shrink partitions to make sufficient unallocated space 
   adjacent to the ESP.

   If you are going to install Linux on the disk, this would be a good 
   time to make space for it too.

   Make sure to tell gparted to "apply" these changes.

4. If you changed the size of any NTFS partitions, gparted might have 
   damaged some redundant metadata.  Fix it now.

   - boot into Windows

   - ask it to scan and repair the volume(s)

   - reboot into your live Linux system and get back into gparted

5. ask gparted to resize the ESP.  It should succeed in resizing the 
   partition but may fail in resizing the FAT filesystem.  The 
   failure message will be rather obscure:

No Implementation: GNU Parted cannot resize this partition to this 
size.  We're working on it!

   What it means is that libparted-fs-resize doesn't have code to deal 
   with some common forms of FAT.  The message is wrong: the partition has 
   been resized but the filesystem it contains has not.  Oh, and nobody 
   has worked on this in a decade.

   If you had no failure, STOP HERE, you are done.

Continuing past FAT filesystem resize failure:

1. exit gparted

2. mount the old ESP. GNOME "disks" has a button for that.

3. copy the contents of the ESP somewhere safe.  It could be to a USB 
   device or even a non-ESP partition on the system disk.
"sudo cp -a ESP SAVED" 
   ESP is the ESP mount point.
   SAVED is to be a new directory in the place where you have enough free 
   space

4. unmount the old ESP

5. Make a new, larger FAT filesystem, filling the partition:
"sudo mkfs.vfat X"
   where X is the device name of the ESP partition.
   On my system, it was /dev/nvme0n1p1 but yours might well be different
   This will wipe the old FAT filesystem and make an empty new one that 
   fills the partition.

6. mount the new ESP

7. copy the old contents of the ESP into the new ESP.
"sudo cp -a SAVED/* ESP
   Note: the /* bit is to avoid creating an extra directory layer in the 
   ESP.  * isn't great but FAT filesystems don't have files starting with 
   ".".

8. shut down Linux; unplug the live stick

FINI
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] "RISC-V technology emerges as battleground in US-China tech war"

[D. Hugh Redelmeier via talk]

| From: Evan Leibovitch via talk 

| I think there's something very different at play here. While I'm sure
| existing chipmakers are whispering in Congress' ear, they don't have any
| selling to do. Since the US has already put export controls on advanced
| chipmaking technology and equipment in China, RISC-V can be trivially
| advanced as a path to circumvent such controls. It's no secret that the HQ
| of the RISC-V consortium was moved from the US to Switzerland explicitly to
| inhibit any one country (or bloc of countries) from inhibiting its progress.

Right.  There are a bunch of hard things about making a competitive 
processor chip or SoC.  Designing an Instruction Set Architecture is not 
one of them (I've done it myself).

Launching a new architecture and getting enough infrastructure to be 
useful seems to be the hardest.  RISC-V is almost there.  Its design is 
complete enough to manufacture.  There are a bunch of SoCs that you can 
buy now.  Linux runs on them.  That genie is out of the bottle.

The remaining barriers have nothing to do with RISC-V International.

Designing fast implementation designs ("micorarchitecture") is hard.  
There are open implementations but they lack credibility.  That may 
change.

ARM licences a bunch of microarchitecture and most ARM chips use one or 
two of these.  (The major exception is Apple -- they use their own 
microarchitecture.)

The ISA also is missing ancillary functions:

- PMIC
- USB
- GPU
- PCIe / NVMe
- TPU
- SATA
- DRAM controller

Fabricating chips from these designs is hard too.  That's what the US has 
cracked down on: ASML in the Netherlands has a lock on making masks.  
This ban hurts a tiny company like the Netherlands a lot.

TSMC, headquartered in Taiwan is the leading fab in recent years. They are 
barred from selling some stuff to PRC.  This hurts them a lot.  It might 
be sufficient to cause PRC to invade.


A US ban on RISC-V International contributions from the USA will have a 
few effects:

- development will be slowed

- China will pull further ahead in control and contribution to RISC-V

- US looks like an unreliable partner.  (Look at NVidia's stock this 
  week.)

In effect China will be ahead from this move.

The only way to keep technical leadership is to advance more quickly than 
you competitors.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] "RISC-V technology emerges as battleground in US-China tech war"

[D. Hugh Redelmeier via talk]

Grrr.

Some US lawmakers want to restrict US companies from working on RISC-V.
They are worried it will benefit China (of course it will, and the rest of 
us too).

It is true that ARM's ownership has allowed the US government to make 
things very difficult for Huawei.  No wonder China likes RISC-V.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] (very off topic) where to buy old AM/FM receiver?

[D. Hugh Redelmeier via talk]

| From: Michael Galea via talk 
| 
| I priced out various pre-amps online, got over the shock of their price
| and built one from some online plans.

Good on you!

I don't know if this phono pre-amp is any good, but its from an OK brand.  
It currently costs $29.99.  There is a newer model with two vacuum 
tubes(!).

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] (very off topic) where to buy old AM/FM receiver?

[D. Hugh Redelmeier via talk]

| From: Evan Leibovitch via talk 

Good advice about estate auctions and downsizing.

| The best estate-auction site of which I'm aware is Maxsold,

Thanks.  I did not know about that site.

| TMI time: In another message Hugh wrote:
| 
| New receivers often don't support RIAA equalization needed for record
| > players.  You can get an external phonograph pre-amp to do that or you can
| > get an old receiver with a phono input. (Since phonographs are cool again
| > MAYBE phono in has been revived on some   models.)
| 
| 
| In my experience (and my first job was at a stereo store, TAT near
| Yonge-Steeles), a pre-amp might be necessary, but it has nothing to do with
| equalization. Rather, there used to be two stylus-cartridge types used for
| turntables -- ceramic and magnetic. Ceramic cartridges were cheaper and had
| lower-quality sound, but produced high sound levels similar to other
| sources like cassette or CD players. Magnetic cartridges were more
| expensive and accurate, but far quieter and needed a pre-amp. Most
| receivers made from the 70s to the 2000s. and many but not all made today,
| have a separate "phono" input that indicates a pre-amp is built in. Given
| that these days vinyl records are bought for perceived better fidelity,
| ceramic-based cartridges aren't even made anymore; the difference now is
| moving-magnet or moving-coil. And I just shocked myself seeing what can be
| spent on just a cartridge these days, in support of a medium I thought was
| obsolete.
| 
| (Whether the "record player" William has is ceramic or magnetic depends on
| the stylus cartridge installed, not the turntable itself. Most likely
| magnetic.)

These things are arcane.  But there are two different things going on
here.

LPs are recorded in a way to minimize the average width of a track.
Without that, low notes would have large excursions.  This processing
uses the RIAA Equalization Curve.

Signals from other devices (tuners, tape decks, computers, ...) are
not processed this way.

"Moving Coil" cartridges were essentially audiophile things when I paid
attention.  They generated a lower signal level so they needed a
special preamplifier.  For signal integrity, it was felt that this
should be as close to the cartridge as possible and was sometimes
called a "head preamp".  Eventually, the industry came up with a
cheaper/simpler/inferior solution: a (passive) step-up transformer.


So:

- before other processing, a signal from a moving coil cartridge
  needs to be specially boosted.

- any record player signal needs to be fed to an input that reverses
  the RIAA equalization ("phono").  Or processed by a device to
  reverse RIAA before going to a normal audio input.  (It may be that
  new beginner turntables do this themselves since phono inputs are
  almost extinct and explaining this stuff to a customer is hard.)

- other signals must fed to an input that doesn't handle RIAA
  equalization.  Those inputs are labelled almost anything but "phono"
  or synonyms.  For example "audio in", "line in", etc.

In computer networking terms:

- use of a moving coil cartridge is a local decision so the MC pre-amp
  or transformer is a local decision.

- use of RIAA equalization is an interoperation standard.  Essentially
  all LPs require this.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] (very off topic) where to buy old AM/FM receiver?

[D. Hugh Redelmeier via talk]

| From: William Park via talk 

| While cleaning my house, I found box speakers and LP record player. Brand name
| "Sanyo".  What's missing is AM/FM receiver that came as a set.  Do you know
| where I can buy an old stereo AM/FM receiver?

Adding to what others have said:

- new receivers often don't support RIAA equalization needed for
  record players.  You can get an external phonograph pre-amp to do
  that or you can get an old receiver with a phono input.  (Since
  phonographs are cool again MAYBE phono in has been revived on some
  models.)

- AM and FM are fading out too.  AM goes first -- it really sounds 
  terrible, at least in North America.

- you might want some new features like Bluetooth so you can send
  sound from your phone, tablet, or computer without wires.

- Back in the day, the nominal speakers were 8 Ohms impedance.  But
  bundled systems like yours were often 4 Ohms.  If you are
  calculating Watts, this makes a difference.  I never worried about
  this because I never pushed the edge.  I guess "impedance matching"
  is a thing too, but I didn't think about it.

The latest Hi Fi gizmo we bought was something like this:

- it doesn't have a tuner
- it doesn't have a phono equalizer
- it doesn't have a switch for multiple inputs
but it is small and efficient.
We use it for ancient iPod => good 1980 speakers.
Ours has no Bluetooth.

AliExpress has these things for laughably low prices but they often
come without full cases and power supplies.  I think that they are
meant for car sound systems (12v power in).

Here's one at Amazon, with a full case, and multiple inputs, FM tuner
(no AM), but they don't mention that you need to provide a 12V / 3A
power supply.  No phonograph in.  Quality is unknown to me.

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] why I like shared libraries -- no longer a popular position

[D. Hugh Redelmeier via talk]

| From: D. Hugh Redelmeier via talk 

| It turns out that the bug is in libwebp.  "WebP codec is a library to 
| encode and decode images in WebP format."
| 
| libwebp is used in a lot of programs.  On my Fedora 38 system, it is a 
| shared library so it can be fixed in one update.  Except where the library 
| is copied (for example, statically linked, or used in a container of some 
| sort).

Still more reverberation: copies exist in several Python projects.  
This paper appears to report on rather hard work to discover problems with 
projects in PyPI.

<https://sethmlarson.dev/security-developer-in-residence-weekly-report-16>

Perhaps analogous work is needed in many repositories.  Of course libwebp 
might not be the only dangerous library.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] No keyboard at boot?

[D. Hugh Redelmeier via talk]

| From: Peter King via talk 

| Apparently the problems/errors/failures I reported are a known problem that
| Lenovo has refused to address:
| 
http://reddit.com/r/techsupport/comments/ztvro4/lenovo_legion_5_t5_are_seemingly_bricked_with_no/

Wow!  That's truly terrible.  And very hard to understand how Lenovo
could let this bleed on.

Reading that, it makes me wonder if you could do a firmware update
every month to ward off the bug.  But I cannot think of a rational
mechanism that would make that work.

Does the T5 victim community have a preferred replacement motherboard?
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] No keyboard at boot?

[D. Hugh Redelmeier via talk]

| From: Peter King via talk 

| Booted from a USB with Arch Linux on it; chroot into the now-working computer;
| ran a system update (massive: 500+ packages).  Something in there -- likely
| the new kernel and firmware -- fixed the keyboard problem.  So I didn't have
| to reinstall, just get far enough in to do an upgrade.

Wow.  Heroic measures.

| I now have a functional computer to replace the Lenovo T5.  For those of you
| keeping score, I *still* do not have the replacement MB I ordered from Canada
| Computers on 27 September.

I'm a broken record: I'd cancel the MB order and get Lenovo Warranty
service.

I have to admit my experience with warranty service hasn't been
stellar.  Problems still being resolved.  Here are the ones I've dealt
with today:

- an Asus notebook died (lights on but not POST).  Sent it in.  They
  now say that their are going to transfer the patient to the US.  In
  return for the inconvenience, they are extending the warranty 90 days.
  I hope that isn't an indication of how long they will have my
  machine.

- HP Aero with flaky WiFi.  I know how to make it work in Linux
  (suppressing sleep modes on the card) but I think that that impairs
  battery life.  In Windows, it is unreliable.  Talked for hours with
  support, getting nowhere.  Fun fact: asked ChatGPT (through Edge or
  Bing on Windows).  It came up with an answer that was congruent
  with my beliefs -- impressive.  Support Chat fails now.  Grrr.

- got a Dyson refurbished ed vacuum in July.  One broken part and one
  missing part.  They sent a replacement for the broken part promptly.
  But the missing part was out of stock.  I had to phone them today to
  remind them to send it to me.  CSR didn't know what a "wand clip"
  was (neither do I -- I've never had one).  After perhaps half an
  hour of chat, they will send one.---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] sigh, hardware resources?

[D. Hugh Redelmeier via talk]

Don: thanks for offering a computer to Karen.

It has an AGP slot (empty) and 3 PCI slots (empty).  Not quite old
enough for ISA.

Someone may know how to break the WinXP password from a Linux system.
If Win XP mattered, I assume that it could be re-installed.  But I
don't think Karen cares about Windows.

If I remember correctly, Win XP doesn't have include a free-standing
DOS (it is based on Windows NT).  The last Microsoft OS which included
DOS was probably Windows ME.

I wonder if any drivers are an issue for MSDOS-7.x (what Karen seems
to run).

Video drivers for DOS itself are surely not needed.  But for programs
that try to do graphicy things, a driver might be needed.  I don't
know if Karen uses a program with such needs.

Sound drivers are more of an issue.  Karen seems to want sound.  There
wasn't really a standardized way of doing sound in DOS.

Disk drives should just work, I think.

Is the drive CD and DVD?  Can it burn?  The one at newegg only reads
and only CDs.

How much RAM?  The newegg model has 512MiB -- plenty for DOS, not so
much for Linux.

How much HDD?  The newegg model has 40GB of disk -- plenty for DOS.

Karen: does this look like a computer you can use?
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] CVE-2023-43641: libcue bug allows Remote Code Execution in (at least) GNOME

[D. Hugh Redelmeier via talk]

TL;DR: update libcue to a version released after October 10.

I read this last night:



Summary: 

- libcue has a bug that allows an attacker to execute arbitrary machine 
  code

- libcue handles CUE sheet files

- GNOME (at least) uses libcue for file indexing when it finds a suitable 
  file

- a bad actor can give you a bad CUE file without you knowing it (eg. via 
  browsing a web site)

- GNOME's tracker will automatically run libcue on the CUE file -- kaboom.

You can update your system to get a fixed libcue. The fix is in version 
2.3.0.

Fedora 37 and 38 have fixes in version 2.2.1-13.  I checked this by using 
"rpm -q --changelog libcue | less".
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] WikiConference North America is in Toronto next month

[D. Hugh Redelmeier via talk]

Where: Toronto Reference Library (Yonge, North of Bloor)
When: November 9-12
Cost: US$25 for most of us.  Free if you find that a burden.

You must register by Oct 29.

I really value Wikipedia.  I do a bit of editing.
Sounds like it would be interesting but I avoid crowds these days.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] sigh, hardware resources?

[D. Hugh Redelmeier via talk]

| From: Karen Lewellen via talk 

| In fact, the P3 or greater aspect is far more important than the ISO slot, if
| that makes sense.

Why do you need the power of a P III?  Generally speaking, DOS software 
was built for much weaker processors (eg. 486 or older).

| Although not installing freedos, it can manage more current hardware, so can
| the DOS I run to some extent.
| www.freedos.org

FreeDOS looks like a good choice: it is still maintained.
What DOS do you use?

Here are some technical requirements that may make a recent machine 
unsuitable:

DOS requires a BIOS as opposed to UEFI firmware.
For about a decade, all machines came with UEFI firmware.
But some UEFI firmware can emulate BIOS.

FreeDOS requires that the disk be formated using MBR, not GPT.


DOS requires disks to have 512-byte sector.  That's less common these days 
-- most current drives 4096-byte sectors.  Sometimes they can emulate 
512-byte sectors.

MBR disks are limited to 2TB or less:


FAT32 partitions are limited to 2TB as well:

FreeDOS has supported FAT32 since 2002 (recent, in DOS terms).
Does your DOS support FAT32?
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] sigh, hardware resources?

[D. Hugh Redelmeier via talk]

| From: Karen Lewellen via talk 

| A dvdr or combination cd and dvdr would be appreciated, or at least room for
| both.

Do you need a drive with the ability to "burn" or is reading enough?

As far as I know, all drives that can handle DVDs also handle CDs.

There are a whole lot of different kinds of drives:

- CD, read-only
- CD, read / write
- CD & DVD, read-only
- CD read / write, DVD read-only
- CD read / write, DVD read-write

There was a war about the way writeable DVDs should work.  The two
standards were DVD-R and DVD+R.

I think DVD+R won.  Usually you don't have to worry about this.

Then there are some drives that can handle BluRay discs.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] sigh, hardware resources?

[D. Hugh Redelmeier via talk]

Distilling your toughest requirements:
- You need DOS
- perhaps you would get no value from any additional OS
- you would like an ISA slot.

That means the the computer has to be more than at least a decade old.  
Perhaps two decades (I don't remember all transitions).

I'm pretty sure we can come up with hardware that matches those 
requirements.  But it will only become harder as time goes on.

It would be great if you could figure out how to live with current 
hardware.  I understand that isn't easy for a variety of good reasons.

If you get such hardware, are you able to set it up?

| From: Karen Lewellen via talk 

| First at least a Pentium 3,  I can be flexible on processor speed as I have
| utilities working around some of the DOS barriers.
| As much memory as possible, my previous machine had almost 800 meg of ram,
| again managers made it work, even had onboard graphics memory although that is
| less important.

Those DOS memory managers are quite arcane.  I imagine that configuring 
them is mostly a lost art.  DOS kind of deals with 640K of RAM, with these 
memory managers fudging access up to perhaps 4M of RAM.

I don't see how 800M of RAM can fit into a DOS world.

Perhaps you meant 800M of Hard Drive (but you said 20G of HDD later).

All current machines have gigabytes of RAM.  Quite a different scale.
---
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

[D. Hugh Redelmeier via talk]

| From: Giles Orr via talk 

| 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.

That doesn't match my (possibly unreliable) model of The Way Things
are Supposed to Be.  I think:

- the syslog facility has been replaced by the journald facility

- all the APIs for syslog now go to journald.

- That may require rebuilding old packages.  (Not changing, just
  rebuilding, so that the newer libraries get linked.)

- I would expect that the new stable debian had rebuilt all packages

Any of these beliefs could be wrong.

|  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.

So modifying my ealier question, can you see a commonality in the syslog
entries just before each crash?

| 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.

Yeah.  And you won't know when it becomes safe to run CUPS again.

The best way of gaining confidence would be if a bug fix were
released.  But you don't even know the bug.  It doesn't even sound
clear enough to report.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] editing and publishing our meeting recordings?

[D. Hugh Redelmeier via talk]

This message is unauthorized by the board, but I thought I'd float these 
ideas anyway.

Alex used to do a wonderful job of filming, editing, and publishing our 
meetings.  Lots of work.  Alex has stepped back.

Since we've had virtual meetings, we have made recordings.  But none has 
been edited and uploaded to YouTube.  (Did Alex upload them to other 
places too?) https://www.youtube.com/@GtalugOrg

We used to get a modest number of views -- usually more than the number of 
people attending the meeting.  One of my segments seems to have 20k views!

Is anyone interested in trying to do editing and uploading?  This might be 
something that could be shared.

Some reasons why this would be useful

- for our members who miss meetings

- outreach: to let more folks know that we (still) exist and what we do

- to share more broadly

PS: in my opinion, the number of views does not correlate with the merits 
of the video's content.  That hints to me that better marketing should be 
able to get more people watching.
---
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

[D. Hugh Redelmeier via talk]

| 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!
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] journalctl / systemctl [was: Re: Debian has suddenly become unstable]

[D. Hugh Redelmeier via talk]

| From: Bob Jonkman via talk 

| > One thought might be to disable cups (`journalctl disable --now cups`)
| 
| Did you mean 'systemctl disable --now cups'  ???

I often make that mistake.  Although the names are logical, there is
some kind of cognitive trap here.

From a human factors standpoint, it would be better if they didn't
look alike.  Of course that is less logical and elegant.

I'd also prefer "journalctl" to be shorter to type.  Also, ctl seems
to be redundant: what other command can you give the journal?  On my
system, "jou" TAB completes to "journalctl".



The --now flag is a feature that I've wanted but never noticed.
Thanks!  I wonder if I'll remember when I need it.

The naming is not too good since the requested action was always
immediate.  --now asks for a different but related action.

I've previously daydreamed about a syntax for systemctl which would
allow a sequence of COMMANDS in a single invocation just as a list of
UNITs is allowed.

systemctl [OPTIONS...] COMMAND [UNIT...]

For example: sytemctl enable,start,status dovecot
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] US urged to ban RISC-V

[D. Hugh Redelmeier via talk]

Land of the Free?

"I fear that our export-control laws are not equipped to deal with the 
challenge of open-source software - whether in advanced semiconductor 
designs like RISC-V or in the area of AI - and a dramatic paradigm shift 
is needed," Warner said in a statement to Reuters.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] No keyboard at boot?

[D. Hugh Redelmeier via talk]

| From: Peter King via talk 

| I haven't tried to ssh in, because at the moment it tries to boot
| into a static ethernet address which isn't available, and so network
| connectivity fails.

When I have my ethernet ports configured in an inconvenient way, I
have plugged in a USB ethernet dongle and SSHed in through that.
Usually the dongle will have a default configuration that is
functional in an ordinarly LAN.

Where is the inconveniet network configuration?  Do you use
NetworkManager (config seems to be in /etc/NetworkManager) or the old
scripts (/etc/sysconfig/network-scripts on some system).  If you can
hide that obsolete configuration, you might be ahead.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] GTALUG meeting 19:30 Tuesday, Oct 10

[D. Hugh Redelmeier via talk]

(This is not the official announcement.)

We will meet virutally, via Big Blue Button: 


The first few minutes will be our Annual General Meeting.  This is where 
the board presents the financial statement for the past year and where we 
the members of GTALUG elect the board of directors.  There might be more 
business.

Please volunteer to stand for election: we can use your energy and ideas.
No experience is necessary.

To be a member, you have to have paid membership dues for this year.
I have appended a copy of Warren's helpful email from last year.

After the short AGM, we'll have a regular meeting.
As I understand it, Mike Hoye has agreed to give our October talk.
Mike is a great speaker and I look forward to this.
I have not heard what his topic will be.

| From: Warren McPherson via talk 
| To: annou...@gtalug.org
| Date: Thu, 24 Nov 2022 23:43:11 -0500
| Subject: [GTALUG] [GTALUG-Announce] Membership Dues
| 
| With the agm coming up, now is a great time to renew membership and make a
| small twenty dollar contribution to the organization.
| For those looking forward to the chance to wind their way through the
| Interac E-Transfer process again here is my summary of how it works.
| (Actually I forget each time so the novelty never wears off -- luckily it
| is easy.)
| 
| Log in to your bank and look for "Interac e-Transfers" you may need to go
| to the "Transfers" page and then you will see this option.
| 
| Select "email method" and use the email address:
| members...@gtalug.org
| 
| Putting your name in the message can be helpful as the notification I get
| can sometimes be a touch obtuse.
| 
| Thank you to those who have already contributed!
| (I know who you are.  Ha ha)
| 
| Thank you,
| Warren
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] Raspberry Pi 5 announced (shipping starting near the end of October)

[D. Hugh Redelmeier via talk]

Faster.  Some new capabilities.  US$5 increase in base price over original 
Raspberry Pi.
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk