davidbanham

Volunteer fire brigade. Great way to connect with local community. Also a pathway into experiences and skills you’ll never get anywhere else.

Kids school. Everyone is local, engage with the other parents/etc.

Local dog park. Get into a routine of regularly heading up at a similar time and you’ll get familiar with the others that do the same.

The biggest difference to my mind is that at end of a day you have something concrete to look at and say “I done that”.

In software or business we spend so much time on intangibles. Often worthwhile, but just not viscerally satisfying. “I am mentally fatigued because I refactored the whole controller layer of this app. It turned out the proposed redesign wasn’t worth pursuing so we reverted the commit, but the knowledge gained from this spike was very valuable!”

Hits very different to “All my muscles hurt. I dug that hole. That one right there.”

The line between being fulfilled by manual labour versus being crushed by it is flexibility. If you’re choosing when to work and what to work on, you’ll probably love it. When you’re working through an injury on the hottest day of the year and the aircon on the truck is broken and all the fire in the area was put out days ago anyway… you’ll wish you were refactoring.

Yeah a bunch. Satellite monitoring of lightning strikes has been happening for years.

We also have hot spot detection from himawari and other satellites.

Also very common to have aircraft do FLIR camera runs over the bush on high danger days.

A few startups/researchers are looking at distributed sensor arrays with microphones and thermometers to detect ignitions earlier. That’s mostly in plantation forestry at this stage.

There’s also some work being done to enable the thermal/linescan flights from drones rather than aircraft. Idea being to bring the cost down and thus coverage up.

For clear felling in a rainforest, sure. For clear felling in a timber plantation, probably? For selective logging outside particularly sensitive areas, nah.

People have shoehorned diesels and wankel rotaries into BMW airhead frames. It’s possible but not for the faint hearted.

With a small frame like yours you’d probably be better served by an EV conversion if you want to go down that path.

The exhaust gas recirculation valve on a modern diesel engine does exactly this. It diverts a portion of the exhaust product back into the intake manifold of the engine to achieve more complete combustion and lower emissions.

Yes, it does result in a buildup of goo in the intake. No, the engine does not run well in that state. That’s what maintenance is for. Clean out the goo every 50,000km or so and it’s fine.

Modern diesels are very reliable machines, just like modern petrol engines. They’re certainly not the robust rocks they used to be, though. High pressure common rail fuel systems, precision injectors and multi stage turbos are standard issue these days and they all require regular maintenance, clean air and clean fuel to operate properly.

The rotary in the article is a 2 stroke design, by my understanding.

Fuel injected 2 strokes have been around for years now. In saws, the Stihl 500i.

Scaled 2 stroke engines, both with and without fuel injection, have also been around some time. The KTM 300 XC is a very modern example with 300cc displacement, so the same general category as the one in the article. Modern outboards on boats go even bigger.

Yes there’s a push on at the moment to retire the fleet of small 2 stroke engines in older garden power tools. Emissions controls have been tightening up on new tools over the years. Electric tools are rapidly displacing them with lower maintenance requirements and less noise.

There’s a lot more to 2 stroke engine development than just saws and lawnmowers from the 80s, though.

A commodity chainsaw (Stihl ms660) returns 1.4kg per kw from 2 stroke fuel. Before any performance tuning takes place.

That’s at about a quarter of the size of this. I bet there’s efficiencies to be had if it were scaled up.

I’m just not sure this is the game changing breakthrough the article’s tone suggests.

I love this idea of “wherever your service ends up, it will reach out and sort out its DNS and networking so that the world can find it at the usual address”

Moving things from one platform to another is often simple enough on the deployment front. Reconfiguring load balancers, DNS, etc is not rocket science but it’s a hassle. In this version of the universe, I can just fling a service from my home pc to a data center and nothing really needs to know? That’s awesome.

Obvious practical limitations around latency, load balancing multiple services, etc. Hamd wave over that for the moment. This flavour of idea is exciting!

I walk my kids to and from school most days. Then I find a local patch of bush to hike or run, far enough to tire out my cattle dog. We hit the dog park if I’m feeling lazy.

I volunteer as a firefighter, and get plenty of exercise if I get called out on a job. I get some movement in during training.

I find gyms don’t work for me. They just feel a bit pointless. Immediate term, I need to move so that my dog won’t bug me when I’m trying to work. Medium term, I want to be fit enough to perform on the fireground. Long term sure I want to live long and well, but I can’t rely on that for motivation. Too abstract.

The proposed resolutions from Google have been along the lines of changes to the proxy application code. The resistance has been that this is hard because of the way responsibility is divided between the thin layer that provides the proxy web service and the ‘go’ command itself, which actually fetches the packages.

Would it be simple to solve this with an additional layer of the same proxy? Currently, end users request a package from proxy.golang.org as per the default value of the GO_MOD_PROXY env var. Google runs many of these to handle the traffic, let’s say 1000. They all maintain a mirror of all packages that have been recently requested (note that I expect there’s more nuance here around shared lists of required packages, etc, but the structure should hold true)

The result is that every one of the 1000 proxy instances requests the source data from git.sr.ht every day.

Google could set GO_MOD_PROXY on the existing instances to internalproxy.golang.org. They could then run 100, or maybe 10 of these internal instances. This would drop the traffic to hit.sr.ht by one or two orders of magnitude.

I suspect it would require minimal if any change to application code. This might be accomplished entirely within the remit of a sysadmin (SRE?).

Any holes in my reasoning?

They're all over the place here in Australia. Personally I don't think I've encountered anywhere that's switched back from them. It's also very common to see traditional plumbed urinals converted over to waterless.

I believe the fix for that is to chuck a bucket of water through it once in a while.

My reference for the opposition to the urinals from plumbers came from this Wired article, iirc.

https://www.wired.co.uk/article/the-fight-to-stop-flushing

> they were met with resistance from plumbers fearful of losing their jobs

Same thing happened to waterless urinals. Those are still making headway due to the waterr saving benefits but it’s been a hard fight.

Just before you get too far down the content-editable-is-amazeballs rabbit hole I have to have a quick talk with you.

Sadly, the cross browser support situation is pretty flaky. In practice this means you’ll find a lot of variance in the user input you get depending on which browser it’s coming from. Makes it really difficult to keep styling consistent or parse the content of it.

Sadly this is one of those times we can’t have nice things.

I was pretty ticked that it was down right when I wanted to send an email to notify a client that I’d done something I’d promised by COB. Not like it happens often, though. They’re a good provider.

Figure out why people show up. This may be homogenous across the group, but more likely there are two or three main categories.

Then, ensure you’re making space for them to get the thing they’re showing up for. Show a direct connection to how what they’re doing supports the outcome they’re trying to achieve. Show clearly that without them performing this task, the outcome they’re invested in will be compromised.

This game is all carrot, no stick. The only threat you have is that you might forbid them from continuing to counter their time, and that’s usually gated behind a pretty laborious process.

If someone’s primary reason is something like “I want to protect my community” then you can probably rely on them to keep their nose to the grindstone on less-fun tasks. If their driver is “I like having coffee with other other volunteers in between working” then they’re still an asset, but you probably need to hand them short simple tasks.

Expect attrition in the group and always ensure you’ve got new blood coming in. Be prepared for tasks to get dropped on the floor if someone gets busy or bored. If something absolutely must happen no matter what, that’s a job for a paid member of staff or it needs to be distributed amongst a large group with a sophisticated collaborative structure.

Source: volunteer firefighter that manages an equipment maintenance team and spends way too much of his time doing stuff because he loves it.

This has an email newsletter built in. People can subscribe via the site and you can publish newsletters through their back end.

Block out focus time in your calendar during the work day. Spend that time walking/running/cycling/doing push-ups/whatever. While you’re doing it, think about whatever problems you’re working on.

You’ll come back with way better ideas and solutions than you would have generated at your desk. You’ll have built up enough dopamine that you’ll be able to actually apply them effectively.

My solution to this was to try and solve real people’s problems with tech. Being able to sit down with the person using your thing gives a tangible feedback cycle to your work.

Instead of “I pushed this update and maybe this metric changed 8%? Or maybe something else did it?”

You get towards “Sue saved 2 hours of her normal workday today and got to go home to her family on time”

You’re still using those skills you’ve built, but in a very different way. One that I find creates a lot more dopamine.

It copped a few too many splashes at the beach and never came back to life.

A key criteria for the replacement was "I never want to have to think about whether something will break it" and the Instinct has lived up to that. Have spent the last few years wearing it surfing, snowboarding, through many bush fires and one structure fire where I forgot to take it off before donning. Hasn't skipped a beat.

No, better locks are already invented and will fit that standardised barrel just fine. This is either someone minmaxing the BOM or just not knowing much about what they were specifying.

I loved my pebbles and was very sad when my Time Steel finally died.

I now wear a Garmin Instinct. The UI isn’t as joyful and it’s not quite as pretty. It’s every bit as practical as the pebble and then some. Also you couldn’t kill it with a stick.

Not having an accessible mechanical latch in the front of the glovebox hatch means it’s now a single smooth surface. That’s way cheaper to injection mould than something with a hole in it.

Also, the mechanical latch mechanism needs to be strong enough to withstand hard use over many years. AND it needs to be made in a shape and finish that fits the design of the interior. AND it needs to feel good in the hand. AND it might need a lock in it, either due to expectation or regulation.

Suddenly a simple little solenoid and metre of thin gauge wire starts looking pretty cheap.

I run a pretty-much-fanless Arch Linux desktop. It’s a Lenovo m920q I bought second hand, ex lease on eBay. Super simple, no fuss at all. Comes with remote KVM in the motherboard. (The security implications are totally probably fine)

It’s totally silent in normal use. (Go web dev mostly)

Occasionally the fans will start up to alert me to the fact that I’ve landed on some social scummy website that’s crammed with 20 banner ads on the one page and is probably trying to mine something in the background as well. I shut that browser tab and the noise goes away.

Majestic monolith in Go. Server rendered HTML with tailwind UI on recent products. Materialise css on legacy.

Google cloud platform utilising Postgres Cloud SQL, Cloud Run and Tasks. Email through AWS SES and blob storage on S3.

Hahaha holy shit is GSMIS still running? In all it’s TUI glory?

When I left, the mobile division had its customers split between three different systems; GSMIS, Focus and Arbor. The poor customer service reps would have no idea which one any given user was in when the phone rang. The only way to figure it out was to ask the person for their phone number, then type that number into each backend and see which one returned a result.

And correlated with some pretty horrific fire events:

https://en.m.wikipedia.org/wiki/1974%E2%80%9375_Australian_b...

https://en.m.wikipedia.org/wiki/Black_Christmas_bushfires

I run a platform which has many connected Stripe accounts. Collectively they’ve now transacted over $10M through the platform with very peaky seasonal patterns. I’ve seen many accounts get flagged, I’ve never ever seen one get “nuked”. They may ask for additional verification. Rarely they may place a 30 day hold on some portion of your funds. If these things happen and you feel they’re unjustified, talk to someone about it. It’s a decision that can be changed or reversed like any other. Explain the situation and demonstrate why you aren’t a risk. They’ll listen.

If you’re worried about being flagged, talk to them about your expected increase in volume _now_.

It was the first sentence of the second paragraph of the introduction.