Please note: This is not authoritative information; if you use it and kittens pop out of your router or there’s some way simpler/better way to do things:

1. don’t blame me for the kitten thing
2. please document it and send me a link so I can learn from you.

The Problem.

IPv6. It’s a thing. Who even wants NAT anyway? Docker’s neat, it lets you run containers and stuff. The docs about enabling IPv6 on Docker make it look so simple. Just assign an IP range, right?

A recent post on Richard Branson’s blog about neurodiversity was a good read.

The world needs a neurodiverse workforce to help try and solve some of the big problems of our time. Many people on the autism spectrum excel in areas such as logic, technology skills, problem-solving, pattern recognition, precision, sustained concentration, analysis and other unique cognitive functions. Yet people on the autism spectrum are often overlooked for jobs that they might be brilliant at. It’s a staggering statistic that 85% of autistic adults are unemployed or under-employed.

[Read More]

This article on the New York Times about the idea of “Languishing” explains a lot about my productivity in the last year or so. I’m finally hammering away at a lot of little medium-effort tasks - and not just because I started being medicated for my ADHD at age 37. 🤔

It wasn’t burnout — we still had energy. It wasn’t depression — we didn’t feel hopeless. We just felt somewhat joyless and aimless. It turns out there’s a name for that: languishing.

[Read More]

Quality post for sure - 512 Pixels on the historic sounds that Macs made when they failed to boot. I miss the whimsy in the personality of the early Macs. 🥰

I have to say the PowerPC Performa death sound with the little badoom-tish at the end was my favourite.

I had an issue in the Proxmox VE manager, where on Safari (all the time) or Firefox, when the screen width was too low, the various UI elements made it so I couldn’t access the buttons in the top left.

I found that all of the necessary things to change were in /usr/share/pve-manager/js/pvemanagerlib.js.

Messing around, I screwed it up and had to learn how to extract a file from a .deb file to get the old one back. Let’s grab the package. Don’t judge me for running an old version, my firewall/router is on VM host, which makes upgrades… probematic. I’ve finally got around to making a temporary thing to use to run as a firewall to do the upgrade now 🤣

I’ve been having this issue for a few weeks, where my IPv6 address kept changing on my laptop. It doesn’t affect me on other devices because they don’t have long-lived connections to things like SSH.

Turns out it’s a bug in pfSense which turned up in a recent update of radvd, where it’s not RFC8106 compilant due to mis-handling RA lifetimes.

There’s supposedly a fix coming in a future version (see here), but for now I need a working network!

I had to reset all my tokens for Github recently because I’m dumb and accidentally pushed one to a code repository.

This broke HACS on my Home Assistant install because the token wasn’t working and rather than gracefully handle that, or make it possible for users to re-enter it, it just dies.

What I found is that you really should remove the HACS integration, restart HA and re-add the integration, which will prompt you to do the “Github Device” setup

Well that was weird, I have been having loads of issues with my MX Master 3 for Mac stuttering and lagging and generally being a horrible experience. I have been using it with a unifying receiver plugged into my OWC Thunderbolt (4) Dock, which is sitting about 60cm away under my laptop stand.

Turns out - that’s enough distance and electrical noise to cause the problem. I moved the receiver closer (using a port on my monitor arm) and it works fine now! I generally hate wireless mice, and I’m not really sold on the weight of this one, but finding a fix for my ~$150AUD mouse being a complete arse is nice.

In another strange turn, Firefox has been prompting people with sponsored relevant links in the address bar and on the “new tab” screen. This is how to disable it. I’m guessing I don’t see these things because I typically run a Pi-hole and the data can’t get through.

1. Click in the address bar and type or paste about:config then hit enter.
2. It’ll probably throw up a warning screen, click through it.
3. Filter the list to find browser.newtabpage.activity-stream.showSponsoredTopSites - you can just type *sponsored* and it’ll show the relevant things. There’s likely to be four entries if you do the wildcard filter, the first two are the settings and the second set are if you want to sync them.

• browser.newtabpage.activity-stream.showSponsored
• browser.newtabpage.activity-stream.showSponsoredTopSites
• services.sync.prefs.sync.browser.newtabpage.activity-stream.showSponsored
• services.sync.prefs.sync.browser.newtabpage.activity-stream.showSponsoredTopSites

I’ve got the browser* ones set to false and the sync* ones set to true, which seems to work. Double-click on them to change the setting.

If you get the LMIC_ERROR_TX_NOT_FEASIBLE error when doing TTNMapper or generally using the arduino-lmic library (my version is 3.3.0) then do this patch in src/lmic/lmic.c. This patch was shown to me by Chris from Accelerando Consulting.

Original code:

u1_t maxFlen = LMICbandplan_maxFrameLen(LMIC.datarate);
if (flen > maxFlen) {
	LMICOS_logEventUint32("frame too long for this bandplan", ((u4_t)dlen << 16) | (flen << 8) | maxFlen);
	return 0;
}

New code adds a check for maxFlen:

u1_t maxFlen = LMICbandplan_maxFrameLen(LMIC.datarate);
if (maxFlen > 0) {
	if (flen > maxFlen) {
		LMICOS_logEventUint32("frame too long for this bandplan", ((u4_t)dlen << 16) | (flen << 8) | maxFlen);
		return 0;
	}
}

There’s something about checking for max frame length that only gets triggered when you’re using the 923Mhz bandplan, and this fixes it. I’m guessing it’s part of the lookup in LMICas923_maxFrameLen in lmic_as923.c which return 0 but I haven’t had time to figure out what’s triggering it.