Sprint 3 retro
Fred Zimmerman
algorithmic book publishing. Getting too much message spam here, pls. use contact form at NimbleBooks.com if you want to chat about work.
What went well:
* I achieved the core objectives, i.e. capturing my key metrics and dogfooding a way to inspect them.
* I proved the ability to check database value for bixbyuserid and update accordingly.
* I planned on it being a short sprint and stuck to it.
What went badly/could have gone better:
* I lost three days on struggling with various architectures for new user check.
* A breaking bug in the Bixby IDE #4001 appeared on day n-1 of the sprint and prevented me from finishing up the sprint strong and is still not resolved. This reduced the scope.
Issues identified in previous retros and still open:
* User feedback loop will not be available until I begin capturing altbrain requests and addressing them. This is waiting on Samsung review of capsule version 3.3.10.
* Can't start on marketing funnel until feedback loop is available. See bullet above.
New issues:
* Sprint goals with conceptual uncertainty -- I've never done X before and don't know the recommended path -- tend to soak up a lot of calendar time and cause high stress. Adding well-understood features is faster and more pleasant. However, if I only added features that I already understand, I would never address major user experience blockers.