βˆ’2000 lines of code

The story from Folklore.org tells an episode of the development process inside Apple in the early years of the company.



In early 1982, the Lisa software team was working hard to get the software out over the next six months. Some managers decided that it would be nice to measure the performance of each individual engineer in the number of lines of code written per week. The managers created a form that the engineers were required to fill out every Friday. In this form, one of the fields was assigned to the number of lines of code per week.



Bill Atkinson, author of Quickdraw, one of the lead UI designers and by far the most important developer of the Lisa, considered line count as a stupid measure of productivity. Bill saw it as his goal to write programs as small and fast as possible. On the contrary, the indicator in the number of lines of code encouraged to write sloppy, bloated and crooked code.



It was during this time that Bill was working on optimizing the Quickdraw computational snippets. Using a simpler and more general algorithm, he completely rewrote the region engine [the Quickdraw key structure, which compactly represents a fragment of the screen in memory - approx. per.]. After several improvements, regional productivity has grown nearly sixfold. As a side effect, Bill's work dropped the program size by 2,000 lines.



Bill was just finishing tweaking the optimization when it first came time to fill out a new manager form. When the developer came to the field with the number of lines of code per week, he thought for a second, and then wrote down: -2000.



I don’t know how the managers reacted to this. One thing I know for sure: after a couple of weeks, Bill was no longer asked to fill out the form, and he happily ignored it.



All Articles