Analysis Paralysis: Over-thinking and Knowing Too Much to Just CODE - Scott Hanselman

Scott Hanselman says experience can turn simple coding tasks into analysis paralysis. He describes getting stuck on side projects because he keeps thinking through too many possible failures, especially around architecture and edge cases, instead of just implementing the feature.

He says YAGNI helps in theory, but pairing with another person helps more in practice. Working with another brain brings focus, fresh perspective, and enough momentum to finish work instead of spinning on possibilities.

Reading notes

#

• Experience can make a programmer slower because more possible problems come to mind.
• A simple task like writing a file can expand into worries about permissions, locking, concurrency, filesystems, temp names, randomness, and documentation.
• Hanselman says this is not age-related senility but overthinking plus experience.
• His side projects stalled when he kept starting features and then getting overwhelmed.
• He describes the problem as architectural paranoia that grows from seeing many possible bugs.
• YAGNI is presented as the standard answer, along with the idea that perfect is the enemy of good.
• He says he often needs another brain to complement his own.
• Pair programming is the method that works best for him.
• Pairing with Greg and Jan gave him fresh perspective and focus to get features done.
• He says he also brings that same value when others pair with him.
• Because he works remotely, his pairing happens through Skype, Lync, Join.me, and screen sharing.
• He and his pair work from Trello boards, take turns on the screen, discuss, code, commit to Git, sync, and switch.
• Sessions can last up to three hours, but after that both people get worn out.
• The post ends by asking whether readers get paralyzed in their coding life and how they deal with it.