Hi reader! Gosh it’s been warm this week, and the pollen (in London, at least) has been going haywire.

• I started taking antihistamines, which help but now there’s a whole host of other annoyances to contend with.
• Anyway in news entirely unrelated to antihistamines resulting in unpredictable sleep, I’m writing this from Heathrow airport, sipping a breakfast of a Bloody Mary. The last real all-nighter I pulled was in 2019 to get to the same airport lounge, to catch the exact same flight. How weird.
• It’s all good though, I got a lot of housework done and even drafted some future blog-posts.

On Wednesday, the FT’s Japanese culture club invited in Urasenke who ran a tea ceremony.

• It was Chado chakai, I believe, or the more casual version of the event. After a short history lesson about the arrival of tea in Japan, our host started talking about the cultural undertones.
• Our whole room sat in silence, as we watched a demonstration of the ceremony itself. It didn’t have a great deal many moving parts, but that turns out to be entirely the point. Every action in the ceremony itself has deep connotations & people spend decades honing the art.
• Each ceremony is intentionally a once-in-a-lifetime, folks invite their friends, families & colleagues to share the experience and it’s as much a shared artistic experience, like a concert, as an occasion where folks drink tea. I’m glad I got to share this one.

Recently I’ve been reading Software Architecture Metrics and I’m really enjoying it, here are a few (of several) highlights:

• Fitness functions come up quite a bit. The concept is simple, but very cool: Like their namesake in machine learning - they’re an objective measure which help [quickly] validate when desired behaviour veers off-course. Think a unit-test (eg in ArchUnit) which validates that your codebase follows MVC.
• One chapter talks about measurement of software quality, more holistically, and how it can be a useful suite to prioritise. Firstly, we’re forced to ensure we’ve thought about how to measure security, performance, functional-suitability, etc. Secondly, the data helps us see when our system architecture doesn’t reflect the priorities of the underlying product.
• Another chapter talks about the DORA metrics in a pragmatic, practical way. It includes common pitfalls & obstacles (such as the common big one - change failure tickets!) and some suggested alternatives & mitigations, closing with suggested ways of visualising each metric over time!