An evolution: Shitty software back in the days and in 2018

Shitty software has improved a lot during the past years.

Not long ago, developers hacked some shit and just shipped it. Can you imagine something so unprofessional?
Fortunately, thanks to new methodologies and 16 additional layers of abstraction we can now proudly say: all problems have been solved. We now hack some shit, write shitty fake unit tests, push it through a CI pipeline (running with https://github.com/auchenberg/volkswagen), put it in a container, upload it to a cloud, throw it into a cluster and then ship it.

Shit is hidden so deeply now, nobody can smell it anymore. That’s pretty cool.

An if there’s still some error, it’s pretty easy to find the root cause. Just setup end-to-end testing, cluster wide performance measurements, big data log analytics, another 12 agents injecting code into your applications and before you can say „Does it scale?“, you hit the shit and find your error. Awesome!

Got to love this simplicity.