Four Days or Five? | ✉️ #69

Illustrated image with "MKDEV Dispatch #69" text, asks "Four Days or Five?" Features a person with glasses, wearing an orange sweater over shoulders. Background has a paper plane pattern, with a gray and orange color scheme.
Last updated: | Published:
Illustrated image with "MKDEV Dispatch #69" text, asks "Four Days or Five?" Features a person with glasses, wearing an orange sweater over shoulders. Background has a paper plane pattern, with a gray and orange color scheme.

Hey! 👋

This week some more about Japan for your pleasure, but this time from me, not Kirill. The other day Tokyo’s city hall gave 160 000 public-sector staff permission to squeeze the classic 40-hour work week into just four days, starting this April. A radical move, if you ask me, for a country where “death from overwork” is a dictionary word.

That's enough about Japan. They did great, but they are simply following the footsteps of most western countries. Many of which already tried four days work week and reported reduced suicide rate burnout. Was it a triumph? In the UK’s 2023 trial, 71% of employees reported lower burnout and stress fell 39%. A year-long US/Canada cohort even clocked a nice 69% burnout drop while revenue ticked up 15%.

Huge success, eh? But the wider market still looks like a zombie stand-up: 36% of professionals told Robert Half last month they already feel fried, which is up from a year ago. So the real question isn’t “four days or five?” but “what’s your Mean Time to Burnout, and who’s watching the dashboard?”

That’s exactly what we discussed in the latest DevOps Accents episode with the folks behind Welbemo, a startup turning wellbeing signals into something as actionable as a failed health-check. They explained why stress metrics belong next to latency SLOs, and how trimming a whole day off the calendar only works if you also automate the 2 a.m. pager noise. Go listen now.​

As for us… Well, whether you’re already flirting with a 32-hour week or just trying to make Friday a “No-Deploy-Day”, our crew has been busy replacing toil with pipelines and measuring the human side of uptime. Hit reply if you too want to pilot shorter weeks without blowing up budgets — we know how.

Because if Tokyo can shake off its 24/7 legacy, maybe the rest of us can too. Ideally before the coffee IV becomes an RFC :)

What We've Shared

  • DevOps Accents #60: Is Wellbeing the Most Ignored Problem in Tech? For episode 60 of DevOps Accents, our guests are Samuele Monasterolo, Lorenzo Grassi and Fabio Ciravegna from Welbemo, an Italian startup for workplace wellbeing assessment.

And on the website there are two new articles this time:

  • ​The Ultimate Guide to Cloud Cost Optimization: Most companies are bleeding money in the cloud without realizing it — but it doesn’t have to be this way. In this article, Kirill Shirinkin breaks down practical, no-nonsense strategies that can cut cloud costs by up to 72% without sacrificing performance. If you're tired of bloated cloud bills and ready to take control, this is the playbook you've been looking for.

  • ​Where Container Images Are Stored: Introduction to Skopeo. This lesson of our Dockerless Course introduces Skopeo as a lightweight alternative to Docker for exploring and manipulating container images, highlighting its ability to inspect and copy images across various storage types without a full container runtime. It demystifies how container images are stored and structured under the hood, setting the stage for a deeper dive into the OCI image specification.

What We've Discovered

  • AWS now allows customers in Europe to pay For their usage in advance: Advance Pay opens up a useful capability: actually enforcing budgets for your divisions or teams. You could just say "here is 100 000 eur for the next 12 months for your data analyics service", and thus "motivate" the team behind it to either fit within this budget or watch the infrastructure go down in flames. Just kidding, of course - don't do that. On the other hand...
  • AWS simplifies Amazon VPC Peering billing: One of the games that no one likes to play is "figure out what exactly causes these AWS data transfer charges". This new change at least will let you see how much exactly the VPC peering traffic costs you - and something is something.
  • ​Common Fate is winding down: We've never used Common Fate's main product, but granted.dev is the tool to use to juggle dozens of AWS accounts and establishing secure shell or browser sessions. It's also an excellent tool to distribute ~/.aws/config files among your colleagues. Pity that Common Fate is shutting down as a business, but at least they leave a fantastic legacy behind - and granted.dev will continue living and evolving under the new governing body.
  • ​Atuin - Making your shell magical:​​ Atuin stores all of your shell history in SQLite and allows you to think it between machines, with end to end encryption. Latest feature - Atuin Scripts - goes one stay further by making your shell scripts re-usable and shareable. We are trying this tool out at mkdev to see if we can innovate even faster, still with zero compromises.
  • The Best Programmers I Know: If you are thinking about the skills you need to grow as engineer, this list is completely excellent. Even just being good at half of the things mentioned in the article makes anyone stand out as a professional.

The 70th mkdev dispatch will arrive on Friday, May 23rd. See you next time!