8: Android Perf at Instagram with Mona

In this episode, Mihaela and Pascal chat with Mona from Instagram New York. We talk about Mona's journey from working on web in California to scroll performance at Instagram and how performance overall is approached from measuring, tackling regressions and establishing a company-wide performance culture. In addition to this, we go through some of our listener questions about team structure, time and release management and pick this up later in our interview where we discuss differences in our overall approach to engineering and QA between Facebook and Instagram. Topics discussed: deepfloat: https://code.fb.com/ai-research/floating-point-math/ Litho: https://fblitho.com/ Flipper Notifications: https://fbflipper.com/docs/writing-a-plugin.html#notifications Profilo: https://facebookincubator.github.io/profilo/ Time stamps: |0:00 |Intro | |0:31 |DeepFloat | |1:15 |Flipper News | |3:50 |Q: Team structures at FB | |4:56 |Q: Time management in teams | |7:06 |Q: FB Release Management | |7:33 |Q: Our views on the structure | |9:10 |Guest Intro: Mona | |9:58 |Mona's Journey from Product to Infra | |13:32 |Switching teams at FB | |15:01 |Performance at IG | |15:24 |Measuring performance | |17:27 |Dealing with regressions | |19:07 |Evolution of scroll perf tracking at IG | |20:03 |Good/bad perf practises | |21:36 |Frameworks helping with perf | |22:48 |Engineering differences between FB and IG | |24:01 |QA at IG | |27:08 |App size growth | |28:47 |Maintaining a high perf bar | |33:51 |What would you do if it wasn't perf? | |34:56 |Puppies and oat milk | |36:59 |Outro | |38:49 |Bloopers |