I was reading avidly quite strong viewpoints on what to do with code already running in production that you think doesn’t execute ever (aka dead code) (ref: How do you safely delete a piece of code that looks like it’s never entered). As usual the intelligent and often witty posts are upvoted and brought to the top. One of the comments on the responses literally made my day [1]. The answer on which the comment exists itself is brilliant and even quotes Chesterton’s fence. Chesterton’s fence is a great tool for a thinker. You can apply it whenever you think something pre-existing is stupid, foolish or the people who put it in place didn’t know any better. As another comment elucidates, experiments are modern science’s way of dealing with purely ‘thought’ based approach. As a strange example, I invite you to ponder over the fact that telecommunications in India are governed by Indian Telegraph Act of 1883! As a series of codifications to aid in decision making during disputes, I’d say that has stood the test of time (and used by brilliant minds to maintain order as well ingenious rogues to stretch the limits and exploit loopholes), which is not quite dissimilar to software that is a codification of business needs and speeds up the overall pace of business.

---

My first manager, in her welcoming address to us new joinees mentioned a story that was passed on to her by her manager who used to work at a large mfr of tractors. The story goes that this said manager was addressing a similar fresh batch of recruits, and told them never stop questioning the process and encouraged use of tools like 5 Whys to get to the root of why each step is the way it is etc. In the course of many such batches, one wet behind the ears new joinee saw a process that appeared wasteful to him, whereby a 5 inch by 3 inch hole was drilled on a particular piece of metal used in the front fender and then couple of steps before QC of the fender, a sheet was then welded back again to cover the hole. As he ran back to the narrator with the good news of having identified waste, the narrator simply asked him to probe further. Most of the folks working on the line were immersed in their daily tasks, hence didn’t know (or to some extent didn’t care) about the purpose of the hole. Eventually, he traced down an old timer who clarified that hole was to hold the logo of their joint venture partner, and the logos were welded in place before the QC — and the joint venture had ended 15 years earlier.

The story ended there for us to absorb the need to question everything while we could, before we were immersed in the grind. Stepping out of that meeting, I’d have imagined the change to fix the assembly line and fix the hole would be immediate. And probably that’s what even my manager would’ve wanted us to walk out of that meeting with. Yet I am sure the reality would’ve been much more difficult — hold up the production to pause the line (impacting forecasts, milestones, commitments), dismantle the tooling and find use for them, cancel the scrap contract for the 5x3 and the sourcing contract for the 5x3, talk to the assembly staff about transfers to other steps, updating the induction, training, quality, process manuals and getting approvals from various stakeholders — a communication and persuasion challenge, especially in a profitable unit :)

---

[1] I leave you to revel in the beauty of the comment by user A.I. Breveleri that made my day on the above stackoverflow post (VMS is an OS from late ’70s) —

The VMS OS kernel mode clock decoding method SYS$ASCTIM contains a line of code that helps correct for the sidereal drift of the vernal equinox. None of DEC’s unit tests exercises this code. It ran once — correctly — on 2000-02-28 at 23:59:59:999. It will not run again until year 2400. Please do not delete it.