​
Kia ora! Guess what, it's pie-mail time again!
Special hello to new subscribers, it seems there's a few of you this week! Great to have you here!
Well, we're in lockdown again here in New Zealand. Bummer.
I'm actually finding this lockdown the hardest. My job is a lot more 'peopley' now than it was last lockdown, and so, I'm getting a lot more 'zoom fatigue'.
If you're struggling too, I hope this newsletter can serve as a little bit of a distraction!
Hopefully by the time the next one rolls out, the end will be in sight!
🧊 Something cool happened this week 🧊
I'm changing format a little bit this week.
I want to share something exciting (to me).
I've been working at Fergus for almost five months now. Like a lot of companies at our stage of growth, we've got a monolithic product (we call 'core'), plus a few micro-services.
Our core product is, well, big. Ever since I started, there's been a weekly release cycle for this product. It's been a fairly bulky process.
But, as of Wednesday, we've moved to a deploy-on-demand strategy.
It's taken a lot of work to make this happen, but, we've deployed working code to production five times in the last four days.
Huge improvement!
It's a really exciting journey to be on.
I think the ability to deploy frequently is something testers and quality professionals should really be pushing for.
Deploying quickly de-risks our product in a bunch of ways:
- deployments are smaller, meaning if something goes wrong in production, it's easier to identify the culprit
- faster to deploy means faster to roll back too - meaning once a problem has been identified, normal service can be restored quickly
- changes can be put in front of users faster, meaning we get faster feedback. This can be from observability tools, or directly from users. It means code isn't sitting 'dormant' waiting to be deployed.
​
So, how often is your team deploying to production? What's one thing you can do, to help them go faster? Let me know!
✨ Some interesting links ✨
What can we do better in tech?
Storyo interviewed 30 women and gender-diverse people about what needs to improve in the tech industry. Some really powerful and challenging statements in this one - worth a watch.
​
Deciding if something's good enough
Janella Espinas writes about how to get out of the weeds, and deciding when something is "good enough" even if it's not perfect. There's some good tips here.
​
The API First World
Postman produced a comic about their vision for APIs in the future. It's essentially an ad for Postman, but, it's neato.
​
Why GraphQL doesn't have a "SELECT *
Nigel Sampson at it again with some great info about GraphQL. This week he explains why you can't just 'get everything' - it's by design.
​
Shifting accessibility testing to the left
Another great talk from Marie Drake, this time on accessibility testing. Check it out!
🧩 Puzzle time 🧩
Here's a situation we managed to get into last week:
​
Can you see a problem here?
Any ideas how this could have happened?
What might happen if I now pull this code on to my Mac or Windows machine?
🎪 Events coming up 🎪
Events for those of you in New Zealand:
Nope! We're in lockdown!
​
Events for anyone anywhere!
Also not much happening, but here's a couple of online events that might be of interest:
​
Let's Talk Testing Twitter Space (Tues August 24)
I'll be honest, I don't really know what this is. Beth Marshall, Steve Mellor and Mohamed Tarek are starting a Twitter Space to talk testing. What is a Twitter Space? I don't know. I think it's a bit like Clubhouse? But it sounds interesting, so check it out if you can!
​
Free your dev teams with Terraform and AWS Serverless (Thurs August 26)
Mark Tranter from Equal Experts will be sharing some of his learnings about some of the tougher parts of running serverless. Dial in to find out more!
👋 Thanks for reading! 👋
Take care out there all.
If you're locked down and struggling, please feel to reach out via email. I'm more than happy to chat.
You can always reach me on LinkedIn or Twitter too.
Cheers,
James a.k.a. JPie 🥧
​https://jpie.nz​