So basically, we have low level neutron radiation coming at us at all times from space. Mostly from our own sun, some other external sources too. It takes a whole lot of concrete or lead or water to stop that completely, so anything that makes it through our atmosphere is harmlessly passing through all of us.

But since things like computer RAM and other electronic storage have gotten so much smaller, this radiation is now capable of energizing or discharging individual bits — 1s or 0s — in that storage. Imagine you’re in the hospital for a back operation and the robot arm is approaching a 1 bit that tells it to stop… but that 1 flips to a 0 because the sun sneezed and now your spine is in two fun-sized pieces.

This is all mostly moot today, though. ECC-enabled RAM (memory with protections against bit flips) is the norm and this is a pretty well-understood problem.

RetroArch is super popular and available across many systems, with a bunch of open source frontends for it. I have it on a Raspberry Pi, a Mac, an OG Oculus Quest, playing everything from MAME to PSX.

About a decade ago at a job in Philly, we’d hunt down the spOt Burger cart (that’s how they capitalized it). Tiny little trailer/cart only big enough for one person to stand in, and this guy would park it somewhere new around center city/university city area every day. My memory is a little hazy so I might have some details wrong, but every day he’d grind a blend of ribeye and filet fresh to make the burgers in his cart, cooked around a medium, and served them on a brioche bun with pickled red cabbage and some other fixings. He got the fat content just perfect with the steak blend, and the toppings were unexpected but incredible together.

I haven’t been back in awhile but I heard he was opening a brick and mortar restaurant because his cart was so successful. Hope it’s true!

It’s so fluffy! But if you must print unattended*, get you a spaghetti detector cam! Your printer will stop printing within seconds to a couple minutes of something going terribly wrong.

*This still doesn’t make unattended printing safe, just slightly less wasteful.

A handful of years back, JC Penney made a huge deal about stopping this practice in their stores, where everything is on “sale” all the time. Sales plummeted even though the actual product prices stayed the same. They immediately reversed course.

Hard to blame them. Human brains are weird.

How to manage your mail comes down to what type of person you are. There’s a lot of great advice here for “Type A” people who don’t find it burdensome to follow a regimen, however simple, and keep things tagged/foldered/scheduled appropriately.

Type Bs might try that, have it work for a week, fall behind, and naturally let the process die. I’m that person.

What works for me is only caring about two kinds of messages: unread ones and starred ones. If I read a message and there’s something I need to do because of it, I click star. Instead of using my Inbox as-is, I make my main view a filter that only shows unreads at the top, and starred messages (newest to oldest) below. Messages I read but don’t star immediately disappear. Messages I unstar immediately disappear. Nothing is deleted because I rely heavily on search to give me a refresher about certain topics that came up anywhere from the day before to three years ago.

I’ve never been an Apple Mail user— my personal and work email accounts are both Gmail, and Mimestream on Mac supports the above workflow really nicely. It’s a native gmail client that uses the gmail API directly, so things like tagging and snoozing work when you need them, and the search isn’t trying to search gigs of messages on your local hard disk.

100% agree with adding a coating agent after printing, but if you’re looking to minimize small holes and fissures, consistent line width is super important. High speeds can make thinner lines than when it slows down at starts/stops and corners.

Printing PETG slower than PLA is already a common recommendation, but unless your printer supports input shaping and linear (or pressure) advance, I’d go as slow as you can bear. As a perk, if 245 is the optimal temp for your usual speed, going slower will make 240 work better.

This is tough to diagnose without seeing how it was sliced. Could you take a screenshot of the model in your slicer, using the view that shows what it looks like after being sliced?

Cloud architect here— I’m sure someone’s probably already brought it up, but I’m curious if any cloud native services have been considered to take the place of what I’m sure are wildly expensive server machines. E.g. serve frontends from cloudfront, host the read-side API on Lambda@Edge so you can aggressively and regionally cache API responses, anything other than an SQL for the database — model it in DynamoDB for dirt cheap wicked speed, or Neptune for a graph database that’s more expensive but more featureful. Drop sync jobs for federated connections into SQS, have a lambda process that too, and it will scale as horizontally as you need to clear the queue in reasonable time.

It’s not quite as simple to develop and deploy as docker containers you can throw anywhere, but the massive scale you can achieve with that for fractions of the cost of servers or fargate with that much RAM is pretty great.

Or maybe you already tried/modeled this and discovered it’s terrible for you use case, in which case ignore me ;-)