The mind is like a parachute. If it doesn't open, you're meat.

A decade in eight months

29 Nov 2020

It has been a decade in less than 8 months.

The last time I had some time to gather and collect thoughts is shrouded in mental mist. In the beginnings of April, while adjusting to a full lockdown with far reaching personal consequences, I took on a new challenge by casually accepting a new project assignment.

The new team needed a “data lead”, someone described as “knows databases”. It has been about 20 years since I’ve had to earnestly confront normalisation forms or stored procedures or even query optimisation and indexes but I wasn’t really worried (hence the “casual” acceptance).

I was right not to worry… about databases. Eight months later and all we use is a key-value store.

It has been a decade in less than 8 months.

My team has delivered the impossible: a working, web scale system in a landscape of ever changing requirements and priorities that serves tens of millions of clients daily, all in under 4 months.

My team is able to pivot within hours and delivers features within a day or two. My team has had to pivot and re-prioritize every other week for the past 4 months - need trumps planning every freaking time.

My team has done this working 100% remotely. With cats and dogs, kids and spouses present and overburdened wifi.

My team is doing voodoo and they do it soooo well.

I am immenselly proud to be part of this team.

Go and do the voodoo

It has been a decade in less than 8 months.

The list of lessons hammered under relentless pressure and unprecedented visibility is not fully formed yet. And it is already long.

There are so many things to talk about, I need a sabbatical just to put the list together.

Actually having a list of discrete subjects to talk about is the hardest thing. Breaking apart the melange of compromises, decisions, solutions discarded, solutions chosen and lucky guesses all done in a astonishingly short time.

Looking back at incidents in a manner that isolates them from each other, the sequence they occured and the environment they occured in invites errors. Misconception and misunderstanding.

The usual reductive way I follow when deciding what to focus on when crafting an experience report is not going to work this time. There is not a single point, something easily punted into a single sentence aphorism (“Automate all the things!”, “People before technology!”, “Repository first!”) that can do justice to the past eight months.

I’ve read about big formative projects that gave birth to ground-breaking approaches on how teams create software. This feels like one of those projects, even though there is nothing ground-breaking in the tools and methods my team uses.

I catch myself going through all that material, presentations and notes and the memories of conference encounters. This time I look for the uncertainties, the references to compromise, the “it depends”, the hesitation before delivering a list of rules, the caveats.

It would be a lie to say that people don’t talk about compromises and trade-offs. More like that the warnings and conditionals are discarded with the passage of time leaving a shiny, polished success recipe, a myth.

I need a new way of telling stories about (software) systems and the teams of people that build them.

blog comments powered by Disqus