I'm very happy to announce that today, Feb. 4, 2017, I crossed another item off my 35 by 35 list (and yearly goals) by sending my first outdoor V8-grade boulder problem, Right Exit to Fontainezoo, alternatively called W Goes to Fontainzoo. It represents over a year of effort and attempts, a very Dave MacLeod approach to climbing. Since I first attempted it, I've worked on my climbing roughly 3 days a week on average, including a period with a torn pulley.
Ultimately, I succeeded on this problem ("solved" it, if you will) with a reductionist approach quite similar to the root cause analysis of software failures I've done: in the event of a failure, ask why until you arrive an issue you can address directly. Start at the beginning of the climb, try your hardest, fall off. Why did you fail? Because my foot slipped. Why? Because my center of gravity was too far to the right. Why? Because I was letting my core sag. Pick yourself up, try again, conscious of keeping your core tight.
I performed that process for almost every move of the climb, which was two grades harder than my previous hardest ascent when I started. By reducing the climb to a series of fundamental, actionable improvements, I was able to make progress, despite initially feeling as though the end result was out of reach. Even today, a year later and much stronger, I subtly adjusted my footwork and my angle of attack on holds. And today, all the pieces came together, and it was beautiful.
I'm struck by the similarity to building software and processes. Deliver a release to customers, find out it's failed for some case. Why? Because... Why? Because... Why? Because... Implement the changes at both the software level (so it works next release) and at the process or team level (so that the problem doesn't happen again. No wonder so many climbers work in software.