Frustration a necessary part of coding?
Friday March 7, 2008 by Derek Young
I had a few frustrating days this week. I decided to reformat my hard drive using LVM). (I started by trying to reformat my root partition with LVM, then gave up on that). I did this mainly to try Xen, which I promptly learned does not support the binary Nvidia drivers, requiring an uninstall.
I spent the day today trying to figure out why some of my unit tests were failing half the time, spewing build failure emails to the team. (I eventually figured it out after countless test runs). Then I fixed a bug, accidentally implementing a feature that turned out to be unnecessary.
I used to feel like days like these were wasted days. Like I had lost all this time trying to get something to work while I could have been more productive doing real work. I’ve learned to stop feeling this way. Figuring something out yourself can be the only way to truly learn something cold. Reading about it is nice but there’s no substitute for debugging, experimenting, searching the forums, searching the bug databases for a fix, stepping through source, or throwing out a failed implementation and starting again. Lack of frustration really means you’re really not pushing your boundaries enough.
Correct use of ConcurrentHashMap Is Google actually producing?
Thats an interesting thought. I felt like I managed to do only 2 hours of “real work” this week. There is so much stuff going on that seems unnecessary.
Maybe just accepting that the unnecessary stuff is just as important than actual “productive” coding helps to avoid the frustration.
— Jörn Zaefferer Mar 8, 06:55 AM #
Nice post – Failures are the pillars of success!
— Wizard Mar 8, 10:13 AM #