One thing I’ve learned about writing UI code is to TRIPLE YOUR ESTIMATE. No matter how good you get, the feature will take three times as much work as you initially thought. I’ve found this to be an ironclad rule. I think this is because humans are messy, and human interaction is difficult to predict in enough detail to form complete requirements up front, especially for a new feature or mode of interaction. After your prototype “works”, you’ll spend a VERY long tail finding issues and tweaks that you need to make, especially if other people’s UI work start abutting yours. The work to tune the UI may not happen in a single clump—it may be spread out over weeks or months, and you may have to context switch back periodically to fix newfound issues with something you thought you were already done with. But the work will add up across the schedule, and you’d better have budgeted for it. #FridayDevAdvice #software #UI