A small note on being open about mental health
I strongly believe in bringing my whole self to work where appropriate and creating an environment where my team feel empowered to do the same.
Part and parcel of that is being honest to the team when I’m not at my best. This happened about two weeks ago - it was important to me that my team understood why I was slightly more distracted or on-edge, rather than letting them mistakenly attribute it to something that they had done.
I sent a short email (which, in my opinion, is all it takes) to update them of the situation and that was that.
It doesn’t have to be a huge ceremony or a confessional but it takes effort to normalize this in the workplace, and I am perpetually glad for the emotional intelligence of the team that I am lucky to lead.
I’ve been spending a bit of time doing stuff with Democracy Club open-source projects over the last fortnight.
They run a crowd-sourced site of candidates for elections in the UK, each pulled from a Statement of Persons Nominated and filled out by data-wombling, their lovely term for the act of digging out publibly-available data about candidates e.g. their campaign Twitter.
I did a bit of wombling to warm myself up for the 2nd May locals this year, although I am a fair way behind on the edits leaderboard!
Given the polling day of an election in the UK, when should the Statement of Persons Nominated (SoPN) should be published?
If you’re interested in joining a lovely community, pop into our slack and say hi!
Tidying up our exit process
Shift has some tasks to take care of when a developer leaves Unruly such as credential rotation and revocation.
Petrut, one of our SREs, ran a session where we plotted each of our tasks against the two axes of ‘Effort’ and ‘Pain’:
- Low-effort, low-pain tasks are our target
- High-effort, low-pain tasks are tedious but well-understood - a good candidate for automation and optimisation
- Low-effort, high-pain tasks need more thought applied to their process - how can we remove chicken-and-egg problems?
- High-effort, high-pain tasks deserve their own session as there’s probably a lot of complexity and possible race-conditions.
Through the liberal application of Python, Terraform, and AWS Secrets Manager, we have now automated almost the entire process which frees us up to make iterative improvements every time we need to rotate credentials.
Even though it’s scripted, it’s still a manual process so the next improvement steps will be providing a production-ready rotation lambda for all of our secrets to make our rotations no-touch.
Applying principles from one area to another
Sarah made the great point that we utilise semantics in our commmit messages to be able to reflect on the amount and nature of code we write - why do we not apply the same techniques to our meetings, so we can reflect upon what we’re spending our time on?
She is demoing a proposal this week, and has written about it on her own blog so I’ll not go into too much detail.
It’s a reminder to myself how much we can gain by looking at techniques we apply in some areas and see how they fit in others. Of course, we need to be wary of “when all you have is a hammer, everything looks like a nail” but it’s an interesting dig nonetheless.
What do the concepts of ‘scope’ and ‘type’ mean for meetings? Are they even separable? Does it mean defining our own topology of meetings?
I’m excited to try and answer these questions.
This weekend I finally visited the London Transport Museum - far from being just about trains, it has a fascinating history of London’s transport network from horse and cart to buses to trains to the modern-day tube. I particularly enjoyed the history of its branding and the Johnston typeface.
I baked some banana bread which turned out nice and moist but sunk in the middle - I suspect I might have overbeaten the batter.