Hello! My talk from Deconstruct this year (“Build impossible programs”) is up. It’s about my experience building a Ruby profiler. This is the second talk I’ve given about building a profiler – the first one (Building a Ruby profiler) was more of a tech deep dive. This one is a squishier talk about myths I believed about doing ambitious work and how a lot of those myths turn out not to be true.
There’s a transcript on Deconstruct’s site. They’re also gradually putting up the other talks from Deconstruct 2018, which were generally excellent.
As usual these days I drew the slides by hand. It’s way easier/faster, and it’s more fun.
zine side note
One extremely awesome thing that happened at Deconstruct was that Gary agreed to print 2300 zines to give away to folks at the conference. They all got taken home which was really nice to see :)