Stuart Longland (VK4MSL)

https://mastodon.longlandclan.id.au/@stuartl

Software developer, electronics enthusiast, Amateur radio operator, lover of classic rock.

Blog: https://vk4msl.com
Github: https://github.com/sjlongland

Public Key

npub12n5weher8thl0qlj398tvw2rmwpr33z8j443lpg28xyuzkef9dksnd9dex

NIP-05 Address

[email protected]

Profile Code

nprofile1qqs9f68vmu3n4mlhs0egjn4k89pahq3cc3re26cls59rnzwptv5jkmgpqqc5yuvw

Publishing to

Author Public Key

npub12n5weher8thl0qlj398tvw2rmwpr33z8j443lpg28xyuzkef9dksnd9dex

Show more details

Last Notes

2026-08-22 19:59:09 UTC

- reply

It looks more visually appealing.
Might be an issue if you have a disability that precludes bending down to plug in a socket. Also harder to keep rain & dirt out.
Perhaps somewhere between the two, the socket could be installed in a bollard. That could be made not much wider than the socket itself. Still fairly discrete without the maintenance and accessibility headaches.

2026-08-22 03:13:24 UTC

Thank-you `pip` and `venv` for trying to smuggle a `x86-64` binary into ` .deb` intended to run on a AM3358… this technique works really well for platform-independent builds!

2026-08-16 09:38:23 UTC

- reply

Yikes, not good news.

2026-08-15 10:08:02 UTC

- reply

Linux Foundation support of "AI" techbros is one of the reasons why I'm now starting to distance myself from them and related groups.
Not sure what the connection Linux Australia has with this group, but I've quietly distanced myself from them too just in case. linux.conf.au 2012 is the last of their conferences I'll be attending.

2026-08-15 06:37:33 UTC

- reply

Indeed, motors handle it by spinning backwards.

2026-08-14 11:31:25 UTC

- reply

Means the installer image (which has a kernel and *some* of the modules) is now older than the modules packages up on the mirror.
Sadly, the ABI between kernel versions is not stable, so the only recovery method is to swap out the kernel image … that requires a download and a reboot.

2026-08-09 21:43:53 UTC

- reply

That's Grafana reading from InfluxDB.

2026-08-09 03:52:33 UTC

- reply

The plot thickens… so I went to make a cup of tea… turn the kettle on. Nothing.
No life out of the kettle. Glance over at the microwave oven, it's dead too.
Unplug kettle, move it over to the other side of the kitchen, plug it in… turn it on… off it goes. Oookay, all of GP1 is dead. Check the breaker.
The breaker of course tripped when the fridge went blam. Still plugged in, I try resetting the breaker, it stays closed, so a transient fault… I go upstairs… LEDs on the fridge controller are illuminated, but no fridge lights.
I'd say the thing that went bang is the PSU for running the fridge lighting, the fridge compressor controller still works. We can probably live with that… crisis averted.
Not sure what the "noise" is on GP1's signature, but it looks like the compressor is doing something, the fridge seems to be retaining the cold.
https://mastodon.longlandclan.id.au/system/media_attachments/files/117/063/519/548/491/490/original/3f3312399f6b0b83.png

2026-08-08 21:14:17 UTC

Looks like we'll be fridge-shopping today. I got up, with one loudly meowing cat indicating his hunger… I get Sam's mince out and give him a spoon-full.
Open the fridge to put the tray of mince back in… *BANG*… a flash erupted from the top-left at the back of the fridge. Fridge lights go out.
I wondered if maybe this was just a PSU circuit for the lighting… shut the door, I figured I'd either see the signature power load on circuit GP1, or I wouldn't.
Had my shower, then checked the power meter… sure enough… that fridge has ceased to be.
https://mastodon.longlandclan.id.au/system/media_attachments/files/117/061/953/528/124/136/original/f76beb8de1b4320e.png

2026-08-07 05:26:56 UTC

- reply

Bugger… but as you say, you got them cheap.
The interesting thing is with Chicago, is they had a number of lead singers… many bands build their identity around the lead singer… here the band's identity was the brass section.
The ones really to look for would be these ones:
- 25 or 6 to 4 (Peter Cetera)
- Alive Again (Peter Cetera)
- Does Anybody Really Know What Time It Is (Robert Lamm)
- Feelin' Stronger Every Day (Peter Cetera)
- Hard To Say I'm Sorry (Peter Cetera)
- If You Leave Me Now (Peter Cetera)
- Just You 'N' Me (Peter Cetera)
- Make Me Smile (Terry Kath)
- Saturday In The Park (Robert Lamm & Peter Cetera)

2026-08-06 11:27:02 UTC

- reply

Rather than roll the dice on yet another Chinese vendor… I'm taking a different tack.
I've acquired a roll of 0.3mm PVC.
https://www.bunnings.com.au/nuplastex-137cm-x-3-5m-0-30mm-crystal-clear-pvc-film-tablecloth-blinds-awnings_p0495783
It's a 3½m roll of the stuff, 1370mm wide. Should be enough.
There are five possible ways I can join pieces of it:
- sew it, then use seam sealer (as suggested by @npub1ucv…5a8g)
- glue it / chemical welding (what I was thinking of initially)
- thermal welding
- ultrasonic welding
- RF welding
Whilst I have devices that can put 100W of RF anywhere between 1.8MHz and 29MHz… I don't think my radio finals will take too kindly to this sort of abuse… and a RF welding rig is likely well outside my price range.
Ultrasonic rigs are up around the AU$700+ mark… pricey for a one-off.
Thermal welding rigs need to get to 300°C… soldering iron territory. I have an old Jaycar one that has temperature control and a somewhat buggered tip. Might be an option here.
I'll try a small sample of material and see what works.

2026-08-05 02:06:45 UTC

- reply

With `setxkbmap -option comp:ralt` you just tap right-ALT then press `-` three times to get `—`. (Not sure how you do an en-dash.)
LLMs can bugger off.

2026-07-31 21:23:56 UTC

- reply

Everything I do is "godless".
- Sleeping
- Eating
- Drinking
- Having a shower
- Doing laundry
- Feeding the cat
- Changing the cat's litter
- Going to work
- Doing the grocery shopping
I'd literally die (of starvation and thirst) if I had to stop doing those things.

2026-07-29 00:24:59 UTC

- reply

Peoples' greed is why we can't have nice things.

2026-07-28 03:59:11 UTC

- reply

I'm waiting of the day they bury this LLM quakery 6 down and 3 across.

2026-07-26 03:02:26 UTC

- reply

Whilst researching this, I had a look at some of the libraries out there to see how they do it on different operating systems.
Microsoft Windows and Apple MacOS X provide system APIs for doing this.
On Linux, some libraries go poking around in `/dev/serial/*`, others go nosing through `/sys/class/tty/*`. Some just look for specific names in `/dev` and hope for the best. `pyserial` also checks `/proc/tty/drivers` on Linux, but everywhere else in POSIX land it uses `glob`.
Most do not consider BSD at all. You're on your own to learn what serial ports your computer has. It's a mess.
#NetBSD has a `/kern` virtual filesystem which is similar to `/sys` in Linux… so there *is* scope there to maybe put some sort of virtual text file that dumps a mapping of devices and what device nodes they're at. My mind also thinks about OpenFirmware's DeviceTree.
BSD in general, has the same problem Linux had years ago: it assumed the hardware never changed. In the 80s and 90s, that was mostly true, your devices were hard-wired into "busses". You mostly ignored that there was a ISA bus, you just "knew" you had a device at address 0x300 and "talked" to it. It was there at power on, and there at shutdown.
SCSI was a little more complicated, you spoke to your disk drive *through* this HBA card.
Then came along Firewire, USB and Bluetooth. Devices could come and go as they pleased. Moreover, USB is a tree structure and Bluetooth is a mesh radio network. A flat structure doesn't quite describe the situation. I'm seeing a lot of wisdom behind the idea of Open Firmware's DeviceTree… and it seems to me, the kernel of a modern OS needs to track a live changing representation of this. A big part of keeping track will be providing a mechanism for the userspace to learn of changes in this tree, and to find where things are.
While I don't expect them to wholesale copy Linux, I think there is merit in the kernel calling a userspace application to "set up" devices when they are detected, and "tear down" those devices when they are removed.
This is `udev` in Linux and it is very handy even if it can be a pain in the bum to debug.
I've also seen the virtual `/dev` filesystem route used. Linux used to have `devfs`, and SyllableOS also used such a system. Not saying that's the best approach, but it is another way of exposing this information to the system.
Not sure which is the "right" approach, but I don't think a static "nothing ever changes" approach is viable here.

2026-07-26 01:11:57 UTC

Playing around with #NetBSD again… so I have a Intel NUC running it. One of the thing I frequently have to deal with, being an embedded software developer *and* an amateur radio operator, is #SerialPorts.
More often than not, USB serial ports. Now, I'm used to the Linux way of doing things, where everything is accessed through a /dev/tty* device.
We used to do it the BSD way with a device like /dev/cua0 (16550 UART on ISA 0x3f8 IRQ 4; COM1 for you DOS people) being your serial port ignoring carrier detect, and /dev/ttyS0 was the same serial port but required CD to be up. This got deprecated years ago, so /dev/cua0 disappeared, and we were left with /dev/ttyS0.
I grabbed a test rig I made for work, this is Zephyr RTOS on a Raspberry Pi Pico. It enumerates two CDC-ACM serial ports, one with the Zephyr shell, and the other is a USB serial port. You use the shell port to tell it how to route the RXD/TXD lines on the other serial port, with the board supporting RS-485, RS-422 and TTL serial. On Linux, it shows up as `/dev/ttyACM${n}` and `/dev/ttyACM$(( ${n} + 1 )`. `udev` also creates these symbolic links:
```
stuartl@rikishi ~ $ ls -l /dev/serial/by-id
total 0
lrwxrwxrwx 1 root root 13 Jul 26 10:58 usb-WideSky.Cloud_Pty_Ltd_WideSky_Hub_Test_Rig_E663AC91D3665F30-if00 -> ../../ttyACM0
lrwxrwxrwx 1 root root 13 Jul 26 10:58 usb-WideSky.Cloud_Pty_Ltd_WideSky_Hub_Test_Rig_E663AC91D3665F30-if02 -> ../../ttyACM1
stuartl@rikishi ~ $ ls -l /dev/serial/by-path/
total 0
lrwxrwxrwx 1 root root 13 Jul 26 10:58 pci-0000:00:14.0-usb-0:2:1.0 -> ../../ttyACM0
lrwxrwxrwx 1 root root 13 Jul 26 10:58 pci-0000:00:14.0-usb-0:2:1.2 -> ../../ttyACM1
lrwxrwxrwx 1 root root 13 Jul 26 10:58 pci-0000:00:14.0-usbv2-0:2:1.0 -> ../../ttyACM0
lrwxrwxrwx 1 root root 13 Jul 26 10:58 pci-0000:00:14.0-usbv2-0:2:1.2 -> ../../ttyACM1
```
This is handy because if I have two of these, it doesn't matter which gets detected first, I know which device is which. The links ending in `-if02` or `.2` will be the data port, and the `-if00` / `.0` ones will be the shell control port.
In NetBSD, I plug this same MCU in and I get 4 devices:
- `/dev/dtyU0`
- `/dev/dtyU1`
- `/dev/ttyU0`
- `/dev/ttyU1`
The files actually always exist, but they get mapped to this physical device when the driver detects it and enumerates it on USB.
If I plug in another dongle, I get another pair, a `/dev/dtyU2` and a `/dev/ttyU2`. The choice of `/dev/dtyU[012]` is dependent on which order the devices is detected… which on USB is indeterminate as they can be hotplugged.
The only clue I get is the `usbdevs -v`; it shows a device number and serial number information.
```
vk4msl-nbsd# usbdevs -v
Controller /dev/usb0:
addr 0: super speed, self powered, config 1, xHCI root hub(0x0000), NetBSD(0x0000), rev 1.00(0x0100)
port 1 disabled
port 2 disabled
port 3 disabled
port 4 disabled
Controller /dev/usb1:
addr 0: high speed, self powered, config 1, xHCI root hub(0x0000), NetBSD(0x0000), rev 1.00(0x0100)
port 1 powered
port 2 addr 17: full speed, power 100 mA, config 1, USB-Serial Controller D(0x2303), Prolific Technology Inc.(0x067b), rev 4.00(0x0400)
port 3 addr 1: low speed, power 100 mA, config 1, CASUE USB KB(0x6a21), vendor 2a7a(0x2a7a), rev 0.01(0x0001)
port 4 addr 18: full speed, self powered, config 1, WideSky Hub Test Rig(0x5678), WideSky.Cloud Pty Ltd(0x1234), rev 4.03(0x0403), serial E663AC91D3665F30
port 5 powered
port 6 powered
port 7 addr 3: full speed, self powered, config 1, product 0a2a(0x0a2a), Intel(0x8087), rev 0.01(0x0001)
port 8 powered
port 9 powered
port 10 powered
port 11 powered
…
```
lacking is the knowledge that there are *two* devices allocated to device 18 (the test rig).
Given a `/dev/ttyU${n}` or `/dev/dtyU${n}` device node, how do I find out about the device attached to it? Seems assuming a device node is fraught with danger.

2026-07-25 09:09:14 UTC

Trying to find things online in 2026… bloody tedium.
I do a lot of radio comms events where I'm out on a checkpoint. In all weather… bright sunshine… pouring rain… freezing cold. Everything.
Running a checkpoint involves paperwork, which you must keep dry, but still use out in the elements. I also have a tablet computer, which *really* does not like even the slightest spray on the screen. I thought about this, and bought one of these:
https://www.ebay.com.au/itm/306249867707 — that specific item is no longer available, but you can see others like it such as https://www.ebay.com.au/itm/188473117167
Good, except one problem… despite getting the biggest one they offer, the hood is tight. Barely fits over my head, and if I've got a headset on for a radio, it's a real struggle. But other than that, it keeps a clipboard or tablet dry just fine. The clear fabric is ideal for this as it means any high-vis clothing underneath is visible.
Okay, it wasn't expensive and nearly solves the issue, but it proved a concept. Let's go look for a better one.
"clear enclosed cape" is a query one must use with great caution. The search engine is very quick to throw you down the BDSM rabbit hole very quickly if you're not careful. Either that, or you get sellers that don't understand the difference between a jacket (has sleeves), a cape (like a sleeveless jacket) or a poncho (rectangular fabric).
Image search eventually narrows down something I'm after… from a vendor that's better known for BDSM stuff, but whatever. Based in Germany. €58 + €30 shipping… okay, let's go. I put the order in.
A few hours later, an email:
> Comments: due to very long delivery times and difficult import regulations for PVC clothing, we unfortunately do not currently ship to Australia. We ask for your understanding.
>
> New status: Canceled
Well, news to me. I've brought stuff in from elsewhere without issues. So I'm back to keeping a search engine from going off the deep-end. Ugh!
#rant

2026-07-23 00:59:21 UTC

- reply

Very hard to stay "plugged in" at a radio comms checkpoint.
I'm usually on my feet taking numbers using a Java application I wrote for that purpose.
I also use the machine on the bus… nowhere to plug in.

2026-07-23 00:54:05 UTC

- reply

Okay, recent personal experience: my tablet.
Panasonic Toughpad FZ-G1 Mk1.
Intel Core i5 3437U @ 2.9GHz
4GB RAM. 512GB SSD.
Debian 13. Recently replaced battery.
Under KDE Plasma on X11 (Wayland's support for onscreen keyboards is broken)… I was getting maybe 4 hours runtime.
Running XFCE4, the battery life jumped by 2 hours.
It may not have in practice been much slower, but the slowness overall added up… and the battery stats speak for themselves. The machine was spending a lot of time paging to swap, and that's what made it feel slow.
I guess it depends on what your definition of "really old" is. Most of mine are in their teenage years. The Pentium II I mentioned earlier saw the millenium tick over: I used to run KDE 3.0 all the time on that, but KDE 4 was too heavy for it, I switched to FVWM on there and never looked back.

2026-07-23 00:02:34 UTC

- reply

Well, "sluggish" and "bloated" is a relative term.
Compared to a lean window manager like FVWM or a lightweight desktop like LXQT, it is.
Compared to Microsoft Windows (Microsoft and KDE have been riffing off each other UI-wise for years), it looks like an underfed supermodel.
At the end of the day, the desktop has to work the way _you_ need it to… and do the things _you_ need. Some are able to make BSD work very well for what they do, and that's great and should be encouraged.
Not familiar with GhostBSD but if you've got what you need working well in it, that's fantastic. Wikipedia tells me the underlying kernel is FreeBSD. Like Linux, the modern versions of this (and sadly OpenBSD) have some slop in them, but aside from that they are fine operating systems.
I recently gave NetBSD a spin, and while there is some promise, I'm struggling with getting my Yubikey going in OpenPGP mode (which I _do_ need) and it looks like the Bluetooth audio support lags far behind what's in Linux (the docs seemed to suggest it solely supports the 8kHz mono CVSD audio CODEC… forget HiFi sound… but maybe the docs are out of date, not sure).
That said, I definitely should look at running more of my servers on NetBSD… starting with migrating the OpenBSD ones initially.

2026-07-22 23:23:51 UTC

- reply

I must admit I was a staunch KDE user up until 2009 when KDE 4 was released. KDE 3 ran okay on late 90s hardware and could do everything one needed.
After that, it seemed to lose its way. It lost its ability to define key sequences as keyboard shortcuts or the ability to just use the logo key on its own as a keystroke in a shortcut (even today, it must be used as a modifier with another key). What didn't help is I was running a Pentium II 300MHz and 160MB RAM which really crawled with KDE4.
The KDE applications were fine, it was just the Plasma desktop environment that was particularly sluggish. So I adapted to using FVWM2 and just used the KDE applications in that.
The laptop in question used a joystick pointing device which was uncomfortable to use and only had 1 USB port, so I made the configuration keyboard-oriented. It works a bit like `vi`: I press the logo key, it pops up a menu… that's like `vi`'s "command mode".
From there, I press different keys to invoke actions… the menu has a helpful label and underlined keyboard mnemonic guides me as to what I'm after.
I wish I could do the same with XFCE actually. I have XFCE on my tablet (keyboard navigation is a moot point when you have no keyboard), and its a nice desktop environment to use… but I'm still yet to see something as configurable as the classic window managers like FVWM & co.
I'm now looking at dropping its applications because of slop.

2026-07-20 23:46:49 UTC

- reply

…and if you're lucky, the comments match the code.

2026-07-20 06:20:34 UTC

- reply

Well, I don't have a login on that Wiki so no… (I don't think it existed back in my day, Gentoo/MIPS handbook was in GuideXML.)
It might be needed on the Pi4/CM4 and earlier… I'm not sure. But it breaks things on the CM5.
Firefox scrolling is a _lot_ better having changed that setting.

2026-07-20 05:59:24 UTC

- reply

One gripe with the Argon One Up is WiFi performance seems to be weak. Reading the documentation, it appears the CM5 defaults to using a PCB-trace antenna which lives behind a massive heatsink inside the base of the laptop.
Not a recipe for success on microwave bands.
https://www.jeffgeerling.com/blog/2022/enable-external-antenna-connector-on-raspberry-pi-compute-module-45/ describes how to switch it to the "external" antenna, which is presumably better placed.
Doing this, it seems to receive a lot better.
#ArgonOneUp #CM4 #CM5

2026-07-19 23:40:03 UTC

Been updating the OS on my old faithful (Panasonic CF-53), as well as setting up Gentoo on the Argon One Up CM5.
Package builds are demonstrably faster on the CM5 than the CF-53. As in big builds like `gtkwebkit` are building a lot faster on the CM5 than they are on the CF-53.
The CF-53 is a Core i5-3320M; hyperthreaded dual-core x86-64 running 2.6GHz, there's 8GB RAM.
The CM5 is a Broadcom BCM2712; quad-core ARM Cortex A76 running at 2.4GHz, and the module I have sports 16GB RAM.
The CM5 is leaving the old machine in the dust in all areas except video acceleration… haven't quite got that working in X11 yet (I'll move to Wayland when it's ready… it isn't now for my use case).
I guess I'll know more once I get more of my standard desktop applications up and running.
#CM5 #RaspberryPi #Gentoo

2026-07-17 20:49:07 UTC

- reply

No, there'll be a fuse in the connection from the grid to the premise… usually on the pole or in a pillar box.
Inside the switchboard of the house you'll find circuit breakers, but at the grid end, it's a fuse.

2026-07-16 23:30:52 UTC

- reply

They require helmets here too… imagine my surprise a week or two back, I see one fellow riding a motorcycle along Milton Road near Lang Park Stadium… wearing only a backward-turned baseball cap on his head.
(Don't get me started on off-angle baseball caps either… but that was beside the point.)

2026-07-16 05:30:51 UTC

- reply

It can also happen that you're not on the same phase as your neighbours, one might be on phase-A, you're on phase-B and the other is on phase-C; A and C might be working but B might not.
I've also had it where the fuse on the street pillar blew. Very weird having no power in our house while all the neighbours have power like nothing had happened.

2026-07-15 02:43:39 UTC

Hmm, so gfortran (part of the gcc suite) now has slop in it.
https://slopscan.ava.pet/repo/https%3A%2F%2Fgcc%2Egnu%2Eorg%2Fgit%2Fgcc
Of course it's got a long way to catch up to LLVM…
https://slopscan.ava.pet/repo/https%3A%2F%2Fgithub%2Ecom%2Fllvm%2Fllvm%2Dproject
Given Zig uses LLVM, I wonder where that puts them?
#NoLLM #gcc #fortran #llvm #zig

2026-07-11 06:36:48 UTC

- reply

I've built a custom kernel and updated my `cmdline.txt` to point `root=` at the Gentoo partition. Damn this thing flies on Gentoo.
I had Debian momentarily on NVMe as I re-arranged things on the eMMC, and didn't notice much of a speed-up, notably opening a terminal window took ~30 seconds or more, whether running from eMMC or NVMe.
On Gentoo, it's practically instant (booting from NVMe).
I have some graphics acceleration issues to sort out in X11, but things are a lot more responsive.
Maybe it's all that `systemd` junk. It certainly isn't my CFLAGS because they're as plain as they come: `-O2 -pipe`.
#CM5 #Gentoo

2026-07-11 01:24:43 UTC

- reply

Well… tried a few things. GRUB boots, great… but it appears GRUB knows nothing about `devicetree`, and so the boot process hangs the moment control is handed to the kernel.
Ubuntu 26.04 LTS ISO (arm64) and the standard Raspbian kernel both fail to boot. Guess I'll look elsewhere for dual-booting.

2026-07-10 23:36:56 UTC

- reply

To get this far, you need to do these things:
1. partition NVMe with a GPT; first partition formatted FAT32 and set up as an EFI system partition in the partition table
2. unzip the release zipfile to that first partition
3. grab the device tree overlays out of https://github.com/raspberrypi/firmware/tree/master/boot and place them alongside (I just put all the `.dtb` and `overlay` directory in the root of this partition)
4. merge any special settings into `config.txt` (the default one is pretty bare-bones and the Argon One Up needs some adjustments)
Doing that, it boots into UEFI firmware. Allegedly I can run Windows 11 on this (technically, not legally), but who'd want to do that? What I plan to do is run GRUB so I can dual-boot between Debian 13 and Gentoo.
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/898/307/466/036/352/original/2a1a2572ba2ac948.jpg
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/898/307/466/999/760/original/41e5b04ef4178921.jpg

2026-07-10 23:19:10 UTC

Trying https://github.com/ansleythompson/blueberry on the Argon One Up laptop with a 16GB RAM / 64GB eMMC CM5, booting from NVMe.
The internal keyboard doesn't work which is annoying, but now I'm looking at a UEFI shell and wondering how you load a Linux kernel in this thing. I think I might need to get GRUB going now.

2026-07-09 23:41:30 UTC

```
[394845.189522] sr 1:0:0:0: [sr0] tag#25 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=1s
[394845.189538] sr 1:0:0:0: [sr0] tag#25 Sense Key : Medium Error [current]
[394845.189544] sr 1:0:0:0: [sr0] tag#25 Add. Sense: Unrecovered read error
[394845.189549] sr 1:0:0:0: [sr0] tag#25 CDB: Read(10) 28 00 00 2c 5d a2 00 00 02 00
[394845.189553] I/O error, dev sr0, sector 11630216 op 0x0:(READ) flags 0x800000 phys_seg 1 prio class 2
[394847.012792] sr 1:0:0:0: [sr0] tag#26 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=1s
[394847.012808] sr 1:0:0:0: [sr0] tag#26 Sense Key : Medium Error [current]
[394847.012813] sr 1:0:0:0: [sr0] tag#26 Add. Sense: Unrecovered read error
[394847.012819] sr 1:0:0:0: [sr0] tag#26 CDB: Read(10) 28 00 00 2c 5d a2 00 00 02 00
[394847.012822] I/O error, dev sr0, sector 11630216 op 0x0:(READ) flags 0x800000 phys_seg 1 prio class 2
```
… that's a BD-R that I burned in January, and here in July, it's unreadable.
I have a disc here that I've just tried on 3 separate drives (including the one that burned it), totally unreadable.
How the hell are we supposed to archive stuff? I've got about 300GB of data I want to archive off to hand to my soon-to-be-ex-workplace. BD-Rs with ~22GB/disc seemed the most logical choice.
- Unpowered flash media suffers bitrot due to the MOSFET control gate charges fading.
- Hard drive lubricants go solid causing stiction.
- LTO tape drives are too expensive, and often the cartridges have capacity over and above what's needed. The smaller cartridges for older drives are no longer manufactured.
Bluray seemed to be the only option with decent capacity and no moving parts to fail… but it seems the data retention on these things is rubbish.

2026-07-07 06:18:04 UTC

- reply

We should start calling him Pauper Sadim*
(* Midas spelt backwards.)

2026-07-07 02:42:33 UTC

- reply

Well, you're playing appropriately vintage music on it. Maybe not the same vintage as the player, but definitely vintage…

2026-07-07 01:15:28 UTC

- reply

Just reading the transcript of the conversation with their "bot"…
Not sure if maybe asking for ABC Radio National by its callsign 4QG would help? (4QR for ABC Radio Brisbane) … although I note they haven't actually used those call-signs on-air in quite some time.

2026-07-07 01:08:44 UTC

- reply

Yeah, no idea this way… 4BH listeners were complaining about issues with Google "smart" speakers too. So it may not be isolated to the ABC, and apparently Google's people were investigating.
I very much doubt Google "proxies" the stream via their servers, so it'll be an update they've pushed out to the devices that has broken compatibility with ABCs streaming servers in this case.
Surprising in many ways, because streaming audio is something the Internet community has had working since the late 90s.

2026-07-06 22:33:34 UTC

- reply

I could see the graphic, I was just letting you know that the hyperlink you provided with it was broken.

2026-07-06 20:15:14 UTC

- reply

White text on white background is not readable.
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/874/865/059/712/856/original/ffcf47ceb08c60b1.png

2026-07-03 20:35:52 UTC

- reply

Well, the plan was to run some AX.25.. we've put that aside for now.
The equipment though: mostly we're using old Kantronics KPC3s. These connect to a "RFID terminal" that basically is a PocketBeagle with a MAX3232 level shifter; talks to a RFID reader one side, and a TNC the other. (One of the units, we've outfitted with a NinoTNC instead.)
Software stack, we're using a APRS-based protocol built atop `aioax25`. The higher-level protocol uses NTP for time sync (base64-encoding the NTP payloads works surprisingly well even over APRS), CBOR for encoding configuration settings and APRS messaging for the transport layer.

2026-07-03 19:11:18 UTC

- reply

We're mostly through the first leg, and riders have started on the second leg of the journey. No real photos at this time because the lighting is crap for photography.
The sun has not yet risen.

2026-07-03 14:01:07 UTC

- reply

Midnight has arrived… and there's much neighing in the distance.

2026-07-03 01:12:15 UTC

Well, we're on our way back up to #ImbilQLD for another #EmergencyCommsExercise. We'll be stationed at #StirlingsCrossing to assist with the 2026 #TomQuilty Gold Cup.
https://www.tomquilty.com.au/
We will be running FM voice and AX.25 packet on this event.

2026-06-29 22:49:35 UTC

- reply

Ahh, makes more sense.
Still, being SaaS means they can play the same ball-trap capitalism games that Atlassian are currently if they choose. I think that backup/restore functionality will be crucial to keeping them honest.
It'll need to support pushing back-ups to some place off-site that *you* control, and in a format you can parse and convert with reasonable fidelity to something else if needed. Otherwise it'll be like The Who… "meet the new boss, just like the old boss!"

2026-06-29 22:27:30 UTC

- reply

Ahh, it's a MacOS X application?

2026-06-29 22:14:05 UTC

- reply

Is there a link to either of those? I'd use a search engine but both of those are dictionary words and don't seem good candidates as search keywords for useful results.

2026-06-28 09:12:05 UTC

Well, you picked a fine time to leave me boot disk…
Along with your comrade, blowing my OS away…
My pair of downed servers, 12TB of replicas…
now my Ceph cluster won't heal…
…you picked a fine time to leave me, boot disk!
#HomeServer #Ceph
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/826/959/021/148/169/original/50586d0f729a7610.png

2026-06-26 03:10:39 UTC

- reply

The other day, a 1TB SSD turned up (Western Digital Green SN3000). I tried booting from NVMe but didn't notice much of a speed boost… however for storage of my personal data, NVMe is by far the safest place to keep it: if something destroys the compute module, my data is entombed on the eMMC unless I manage to successfully de-solder it and tack wires to the balls to interface to it.
It's also a better place for swap space. The NVMe should have better endurance and performance than the eMMC. I'll keep Debian on the eMMC (where the 64GB size is less of a constraint), `/home` and swap on NVMe.
I've set up a 180GB partition formatted BTRFS, and have Gentoo building inside that.
So far this machine this machine is taking Gentoo in its stride. I intend to get a bootloader going so I can comfortably dual-boot (maybe U-boot, GRUB or rEFInd?) but my journey into the ARM world is going fine so far.
#Gentoo #RaspberryPi #CM5

2026-06-26 01:28:39 UTC

- reply

Found the font… after a lot of searching made painful by a certain company choosing to name its word processing product "Word™"…
https://charlieharvey.org.uk/page/font_of_the_month_sans_bullshit_sans
Sadly its a 2015-era list of buzzwords it's based on:
https://web.archive.org/web/20150208034847/http://ngenworks.com/blog/250-buzzwords-we-love-to-hate/
We need to do a 2026-era port of this.
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/813/812/082/184/325/original/23ad4c129018aa8b.png

2026-06-25 22:58:06 UTC

- reply

Ligatures, it determines how to render groups of glyphs.
I did hear of one font that took various management buzzwords and replaced them with the word "bullshit" for example.

2026-06-20 03:01:27 UTC

- reply

William Dampier was one of the first…
https://en.wikipedia.org/wiki/William_Dampier
He was part of a pretty exclusive club regarding European exploration of Australia until the Batavia crashed into the Abrolhos Islands off WA in 1629.
https://en.wikipedia.org/wiki/Batavia_(1628_ship)
Two from that trip wound up the first permanent European residents.

2026-06-19 10:22:54 UTC

Anyone who thinks Australia ever had "one" culture, is ignorant.
Prior to European exploration, there would have been hundreds of indigenous "nations" spread across the continent in something more like continental Europe.
Some of those in the far north also traded with seafaring traders from South East Asia.
When Europeans *did* finally rock up, they brought with them a motley crew made up of English, Irish, Scottish, Welsh and a smattering of other nationalities as well.
When exploration began through the centre, they took with them Middle-Eastern people who brought camel trains to the outback.
The gold rush brought people from northern Asia to our gold fields.
Multiculturalism has always been here. It won't go away just because some red-headded clown "doesn't like it". #auspol

2026-06-18 06:47:30 UTC

Silly question of the #NetBSD folk… I know some of the BSD kernels (notably #FreeBSD, and once upon a time, #OpenBSD) were capable of Linux syscall emulation, meaning they could run Linux applications.
Is this still the case, and if so, does it extend to containerisation like #Docker? Platforms of interest would be `x86-64` and `aarch64`.
https://kb.shells.com/tutorials/NetBSD/Docker/ seems to suggest it was at least at one point, a thing, but the link to the Docker download site is a 404.

2026-06-17 22:24:10 UTC

I suppose it's time I shut down the server, did a full back-up and migrated the database. #MastoAdmin

2026-06-17 11:09:55 UTC

- reply

Putting this thing through its paces… I installed `mpv` and tried playing a full-HD WebM video ("Ego: The Michael Gudinski Story"), seemed to handle that with flying colours, and the sound is good.
One thing I'll have to do though, is implement a kernel driver for the battery status… a hacky Python script polling /dev/i2c will _not_ do.

2026-06-17 07:43:16 UTC

- reply

And now I'm posting this from the machine in question. Still got to install more packages, but I have a web browser now.

2026-06-17 04:27:42 UTC

Damn, I knew these things were small, but I didn't imagine *THIS* small! #RaspberryPi #CM5
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/763/555/318/716/301/original/e0bf8a6c05aaff8a.jpg

2026-06-16 00:03:53 UTC

- reply

Ohh gee thanks `curl`… (yes I have rotated that token and fixed the `-v` now).
Looks like there's a big foot-gun using `curl -v` (for debugging) and `--user` (to authenticate), Forgejo doesn't recognise secrets that are Base64-encoded.
Maybe the answer here is I generate `base64( username + ':' + token)` myself and store _that_; use it with `--header "Authorization: Basic ${ secrets.MY_ACCESS_TOKEN }` … that should avoid this problem.
(CC @npub1fh7…ee83)
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/756/855/655/227/922/original/778cd6e0b2bae96d.png

2026-06-15 23:43:20 UTC

Yeah, I am trying to figure out CI workflows, how did you tell? #Codeberg
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/756/774/832/038/968/original/7e5a3f51748d85af.png

2026-06-15 22:04:19 UTC

- reply

Yep, some of our Yahoo addresses predate the Millennium bug.
I acquired mine in 1999.

2026-06-15 03:18:10 UTC

- reply

> I was blown away by the response to it, but I only know of one person who had a goat milling their own. Whilst the mill did a great job (and I explained how to convert your own too), a CNC mill is not the sort of thing that everyone has access too.
Wait a sec, he got a goat to mill plastic? I know those buggers will eat practically anything, but is this wise? Did he consult a vet first?

2026-06-09 03:50:53 UTC

- reply

Would be a terrible shame if the members of that regime were denied entry into Australia for 2032…?

2026-06-01 21:46:17 UTC

- reply

```
stuartl@vk4msl-sb /usr/src/linux-stable $ git pull upstream
remote: Enumerating objects: 49462, done.
remote: Counting objects: 100% (24486/24486), done.
remote: Compressing objects: 100% (3805/3805), done.
remote: Total 49462 (delta 22761), reused 22251 (delta 20681), pack-reused 24976 (from 1)
Receiving objects: 100% (49462/49462), 18.42 MiB | 1.12 MiB/s, done.
Resolving deltas: 100% (41401/41401), completed with 5527 local objects.
fatal: bad object refs/tags/v5.10.256
error: git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git did not send all necessary objects
stuartl@vk4msl-sb /usr/src/linux-stable $ git tag -d v5.10.256
error: tag 'v5.10.256' not found.
stuartl@vk4msl-sb /usr/src/linux-stable $ git remote add google https://kernel.googlesource.com/pub/scm/linux/kernel/git/stable/linux.git
stuartl@vk4msl-sb /usr/src/linux-stable $ git fetch --tags google
remote: Counting objects: 49481, done
remote: Finding sources: 100% (49477/49477)
remote: Total 49477 (delta 39209), reused 49477 (delta 39209)
Receiving objects: 100% (49477/49477), 17.67 MiB | 2.29 MiB/s, done.
Resolving deltas: 100% (39209/39209), completed with 5537 local objects.
fatal: bad object refs/tags/v5.10.256
error: https://kernel.googlesource.com/pub/scm/linux/kernel/git/stable/linux.git did not send all necessary objects
```
*sigh*

2026-06-01 12:22:19 UTC

- reply

Nope, XZ compression is too slow with the size of kernel 6.0 and the watchdog timeout. Switch to gzip which is less good at compression but faster, and it boots before the watchdog barks.
```
Debian GNU/Linux 13 tsimx28 ttyAMA0
tsimx28 login: root
Linux tsimx28 6.0.0-vrt-ts7670+ #2 Mon Jun 1 22:17:57 AEST 2026 armv5tejl
The programs included with the Debian GNU/Linux system are free software;
the exact distribution terms for each program are described in the
individual files in /usr/share/doc/*/copyright.
Debian GNU/Linux comes with ABSOLUTELY NO WARRANTY, to the extent
permitted by applicable law.
root@tsimx28:~# uname -a
Linux tsimx28 6.0.0-vrt-ts7670+ #2 Mon Jun 1 22:17:57 AEST 2026 armv5tejl GNU/Linux
```

2026-06-01 12:10:36 UTC

- reply

```
HTLLCLLC
U-Boot 2014.10-00007-g8c6fb433d54 (Jun 01 2026 - 14:08:13)
CPU: Freescale i.MX28 rev1.2 at 454 MHz
BOOT: SSP SD/MMC #0, 3V3
Watchdog enabled
I2C: ready
SPI: ready
DRAM: 128 MiB
MMC: MXS MMC: 0, MXS MMC: 1
MMC0: Command 8 timeout (status 0xf02c4020)
** Bad device mmc 0 **
Using default environment
In: serial
Out: serial
Err: serial
Net: FEC0 [PRIME]
Press Ctrl+C to abort autoboot in 5 second(s)
Booting from the SD Card ...
** File not found /boot.ub **
4815616 bytes read in 1980 ms (2.3 MiB/s)
21195 bytes read in 100 ms (206.1 KiB/s)
Kernel image @ 0x42000000 [ 0x000000 - 0x497b00 ]
## Flattened Device Tree blob at 41000000
Booting using the fdt blob at 0x41000000
Loading Device Tree to 47b4e000, end 47b562ca ... OK
Starting kernel ...
Uncompressing Linux... >hang<
```
Ookay, kernel 6.0 will need more work!

2026-05-26 23:09:11 UTC

- reply

The word "Mac" does not appear anywhere in your post.

2026-05-26 23:06:25 UTC

- reply

I see no such option named "Automatically detect voice activity to create My notes".

2026-05-26 22:50:52 UTC

- reply

Can't see that option in Zoom 7.0.0.1666-r1 (Gentoo ebuild.)

2026-05-21 23:21:59 UTC

Protip for the ABC… white text on light grey background is really hard to read.
White text on white background is impossible to read.
I've been around the Internet and I can't find another news article like it.
https://www.abc.net.au/news/2026-05-22/ghost-communities-of-robinvale-where-people-dont-exist-in-census/106572538
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/615/133/150/243/422/original/16aae81ef5bf18ce.png
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/615/133/157/773/298/original/164ad5dd42642558.png

2026-05-20 10:51:32 UTC

- reply

I've got some music by them.
"Chirpy Chirpy Cheap Cheap" and "Soley Soley".
Scottish 70s group if I'm not mistaken.

2026-05-18 21:17:26 UTC

- reply

I think [@Mer__edith](https://mastodon.world/@Mer__edith ) might know (or the people behind @npub1rkw…5q62).
One possible alternate videoconferencing platform (I realise is not Signal) is Jitsi. Haven't tried it with a largish group, but I did try it 1-1 and no sign-up was required.
You set up a meeting, then got given a link you could pass to the other party.
https://meet.jit.si/
That said, I'm interested in whatever others suggest. Video conferencing is hard to self-host because the bandwidth and CPU requirements means VPS hosting gets expensive and home Internet connections are rarely fast enough.

2026-05-18 21:00:12 UTC

- reply

Yes, and a tool named "rtl_*fm*" also sounds like it's for receiving Frequency Modulation signals, not Single Side-Band.
So it's a poorly named tool. I'd have called it `rtl_rx` or `rtl_demod`.

2026-05-18 20:53:37 UTC

- reply

Yes, that is one way… my question was whether someone had written such a program, because the idea that I'm the first person in the world to want to do this seemed odd. Why re-invent the wheel?
Turns out, `rtl_fm` exists (as per @npub1ucv…5a8g 's post elsewhere in this thread). Not a 100% solution, but 90% of the way there.

2026-05-18 19:52:43 UTC

- reply

No, because `slowrxd` is running on the RPi4, if I did what you're suggesting there, I'd be sending the I/Q data off elsewhere to be decoded, then sending the decoded audio *back* to the RPi4 to demodulate the SSTV transmission.
Utterly pointless.
This is a headless station. There's no phone. There's no laptop. There's no desktop. There's no monitor. Just a radio, a RPi4, and an Internet connection. It hears a SSTV signal, it decodes it, it slaps it up on the Internet via the [@vk4msl_sstv](https://mastodon.longlandclan.id.au/@vk4msl_sstv ) bot and throws the image up on a website.
Putting a second machine in only moves the problem, because the second machine will be headless like the first.

2026-05-18 06:01:02 UTC

- reply

Not sure how that helps… this is a piece of hardware I do not have, it's not a piece of software I can download and install.

2026-05-18 05:02:47 UTC

Silly SDR question for people.
At the moment [@vk4msl_sstv](https://mastodon.longlandclan.id.au/@vk4msl_sstv ) is down because I had to grab the radio I use for it to run a number of WICEN events… e.g. it was the base packet station radio at Kilkivan.
Using a full-blooded 100W rig like a FT-897D is a bit overkill too.
I have a headless Raspberry Pi 4 that I use for that SSTV bot. Headless meaning not only does it not have a monitor, it also doesn't run X or Wayland. That rules out anything GUI-based like GQRX, and no, I'm not going to run VNC/RDP/NX/Spice/<insert GUI remote desktop protocol here>.
Is there a SDR receiver that can drive a RTL-SDR dongle and decode SSB audio for piping to another application (either via Pipewire or TCP sockets) and be controlled via a CLI? I have an upverter to shift HF into VHF, I just need something to decode the I/Q data and control the VCO in the RTL-SDR.
#AmateurRadio #SDR

2026-05-14 21:37:11 UTC

- reply

How on earth do they email someone who's not using Office365?
Pretty sure Postfix doesn't support MAPI, nor does Gmail.

2026-05-03 21:43:32 UTC

- reply

…and it's a `file:///` link.
You might as well say it's a print-out kept on your dressing table.
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/512/824/387/763/900/original/4ed2b549398f8b90.png

2026-05-03 21:17:14 UTC

- reply

I can't see a `mobile` directory below `/var`.

2026-05-03 21:06:15 UTC

- reply

File not found.

2026-04-29 08:27:06 UTC

Odd thing this afternoon… get up the stairs, dump my bags on the table, cast aside the mask and earmuffs… looking around I notice a benchtop supply sitting on top of the fridge is off.
Odd, don't recall turning that off… the radio attached to it yes, but not the supply itself. There was a PocketBeagle SBC and a Kantronics KPC3 TNC running off it.
Sharing the power point is the aforementioned fridge. Overhead lights work.
Hmm, do we have a black-out? I turn the lights on above the rangehood, this is on a separate circuit (GP2) directly coupled to mains. Nope, GP2 is fine, we have lights there. GP1 and lights are on back-up power, but GP2 is raw mains.
I reach over and try another under-cupboard light, one sharing the circuit (GP1) with the fridge. Nope, dead as a doornail.
Something on GP1 upset the breaker… not quite sure what. Judging from the load traces, sometime around 16:10. The only things turned on there (on GP1) were the server rack, a few small things in the study, the fridge, and the bench PSU, and a small LiFePO₄ charger.
The good news is the fridge was only off for an hour or two.

2026-04-29 08:20:27 UTC

- reply

Re-ran the build… and `portage` in its wisdom decided to re-install all the packages from pre-built binaries, *including* `bison` before retrying `iproute2`.
Of course in doing that, it overwrote my fixed `bison` for a broken one in the binary packages… and `iproute2`'s build blew up again with a segfault.
Well, pretty sure it's `bison` … small chance it's `make` but `make` seems to work fine for other packages, so I'm going with `bison` being the culprit.
So re-build `bison`, *manually* build `iproute2` before it has a chance to break `bison` again… then re-try the build, and we should finally see the finish this week-end.

2026-04-28 11:03:49 UTC

No idea why tonight… but 21:03… and I'm wide awake for no reason whatsoever.
It sucks because I need to be on a bus at 06:40 tomorrow. Ugh.

2026-04-28 07:12:50 UTC

- reply

Cue Yothu Yindi — Treaty (the original one not the Filthy Lucre mix)

2026-04-28 03:06:43 UTC

- reply

After one cousin presenting to emergency with gut pains 3 times that was "dismissed" as a stomach ache… except that third visit where it was identified correctly as appendicitis (about to burst!!!)… not sure overworked emergency physicians are a big improvement.
That said, a hospital has accountability, a chatbot doesn't. So still the better bet to go to ER than faff about with a computer.

2026-04-27 08:31:12 UTC

Ahh, the joys of bitrot… Brisbane WICEN has a competitor tracking project that uses RFID readers hooked to AX.25 radio stations to track competitors on a horse ride.
Started back in 2018… and has seen steady improvements since then… but then there was a period in 2023 where all the packet stuff just got left to sit.
Consequence: PocketBeagle SBCs still running Debian 11.
Lost config on my Raspberry Pi 3+ w/ UDRC-II.
Stuff that doesn't play nice with Python 3.13/3.14.
The RPi3 of mine was still running Debian 12, a little newer… but updating that took ages (extracting the SD card for re-flashing is a pain in the arse).
I think today I got back to the point where the project works the way it did back in 2023… but doing so on current libraries, and updated to Debian 13 as the OS.
Spent much of the day re-building stuff then listening to the squawk of Bell-203 modulation as APRS messages got fired between two radios in the kitchen on UHF (439.075MHz). But… I think I got somewhere!

2026-04-27 02:51:37 UTC

- reply

Nearly at the home run of the proper stage 3 tarball.
Of course the build of `iproute2` got blown up by a miscompiled `bison` (`bison` segfaulted processing one of `iproute2`'s source files): I really need to figure out how to get Portage to use `-O2` not `-Os` for that one. Everything else runs fine with `-Os`, and I do want to optimise for size as these machines have at most 256MB RAM to work with.

2026-04-27 00:04:07 UTC

- reply

Heavy as flipping lead though.

2026-04-26 23:36:09 UTC

- reply

My father used to work for Telecom Australia, and he'd bring home a "laptop".
https://en.wikipedia.org/wiki/Toshiba_T3100
Can't recall exactly what model they had… but it was one of the 286 variants, apparently had an 80287 math co-processor as well.

2026-04-26 06:08:05 UTC

- reply

Yeah, "weaponise the penis" sounds like something out of a Skyhooks show.
https://mastodon.longlandclan.id.au/system/media_attachments/files/116/469/508/634/954/761/original/84720414578fd807.mp4

2026-04-26 05:49:39 UTC

- reply

Give it a few weeks… if nothing is done, then draw a penis on it and it'll be fixed overnight.

2026-04-25 23:05:32 UTC

- reply

The way one rolled away, one could be forgiven for thinking you're losing your marbles.

2026-04-25 22:53:11 UTC

- reply

I'd imagine the Italian→English translation would be: "get these infernal things off my ears!"

2026-04-25 09:16:45 UTC

- reply

Looks like it's a "borrowed" word, which would be written in the katakana character set.
I could be wrong though… because it's been 30 years since I looked at that.

2026-04-25 06:36:07 UTC

- reply

I wanted an A4 laser that did duplexing and worked with Linux.
Settled on the Kyocera P5021cdn. Ethernet, colour, duplexing, Postscript level 3. The gotcha being the cartridges are expensive, but it works out to about AU$0.20 a page… so not terrible.
I've had it working in both Windows 10 and CUPS on Linux pretty easily … in fact the latter it just works.

2026-04-25 06:30:56 UTC

- reply

Back around 2001 or so… I was doing some industrial experience with Australian Computer Traders at Bowen Hills.
At the time they were selling Lexmark inkjets for AU$50.
The printer cartridges they took sold for AU$80.

2026-04-25 06:28:07 UTC

- reply

Sleepy Don.

2026-04-25 02:13:17 UTC

- reply

No proprietary software tied to hardware IDs?