Welcome to Platform Weekly, the newsletter that takes the best of the cloud native and platform engineering world and delivers it hot and crunchy to your inbox every week 🥐
After an incredible PlatformCon 2022, the team knew we had to step up our game and shine the spotlight on more resources and voices from the community. So whether you plan to become a weekly reader or a regular contributor, we're pumped to have you here!
Let’s get bakin’.
pc: Daniel Bryant, Head of DevRel @ Ambassador Labs
Beating Cognitive (Over)load
Understanding the cognitive load on developers and finding ways to ease it was a major theme at PlatformCon 2022. (Which shouldn't be surprising given it's why we build internal platforms in the first place! 😉)
🧠 Cognitive load is the amount of information processing required by a person to perform a task. It becomes detrimental to learning when it exceeds working memory capacity. With the evolution of DevOps spawning an increasing amount of new tools and frameworks to learn and apply, it's difficult for developers to keep up. Ambassador Labs Head of DevRel Daniel Bryant illustrated the increase in cognitive load over the course of his career in his talk (see the graphic above).
The solution is a platform with golden paths. 🛣️ Spotify's engineering team describes golden paths as an "opinionated and supported path to build something." They help engineers avoid reinventing the wheel, give them fewer decisions to make on the day-to-day, and let them move faster. At the same time, it still offers developers the freedom to do what they need (when they need) without other teams. The reduced cognitive load enables greater productivity.
The cognitive load doesn't just disappear, though. 👻 Syntasso COO Paula Kennedy explains that the cognitive load is transferred to the platform team. So we should think about the platform team experience, also.
pc: Paula Kennedy, COO @ Syntasso
Short on time? We got you 😋
🥐 Infrastructure as code is expanding. Now we have environments as code, security as code, storage as code, and so so much more. Patrick Debois dove into the trends from 50+ concepts "as code" and came up with some interesting conclusions, including that the creation of MLOps and DataOps is blurring the lines between code and data. Check it out.
🥐 The first tweet says it all. I broke down Naphat Sanguansin's (ex-Dropbox) Platform Engineering article about why platform teams should prioritize shipping workflows over shiny new tech.
🥐 If you don't understand CPU limits and requests on Kubernetes, check out Natan Yellin's (CEO @ Robusta.dev) article on the subject. Or check out the recording from the webinar we had yesterday on Platform Engineering community.