Should Developers Ever Use Excel?

People often think of developers as code-writing engineers. Probably due to the fact we usually sit in front of code (or things that look like code).

However, in reality, a developer's toolkit consists of much more: design patterns, frameworks, databases, various coding languages and lots of other fun stuff, based on the experience, personal preferences and skillset of each one.

A few months ago we had a discussion (some will call it an argument) about a task that my colleague was assigned to, that had a repetitive nature. As a part of a database migration, they had to perform a lot of similar requests to the database, with placeholders they imported from an external CSV file (i.e. Find X, set its field Y to be Z). Because the queries had the same structure, they decided to use Excel in order to generate them.

I was terrified to hear that. A developer, master of code, uses the common folk instrument while on duty.

We immediately explained very strictly that this is unprofessional and that developers simply don't do that. Then we took a breath, calmed down and elaborated:

1. Our forte is coding; This kind of repetitive task can be done using code. Hence, code is the best way to execute it.
2. In the case of database queries, there might be optimisations that code implementation allows, where duplicated hard-coded queries would skip.
3. In case this task (or a similar one) will be repeated at some point in the future, writing a code will allow adjusting it and re-use the relevant logic.

But then a thought came into my mind: maybe our judgement is clouded by our egos; Maybe, indeed, if all we have is a hammer, everything looks like a nail. What if this is just a single-time task and it was really a lot simpler for these guys to get it done this way?

Personally I think that basic data handling is a crucial part of every R&D team member's toolbox. Data is the driving force of all software products. Even though it's not necessarily in the scope of every development position, we all touch it - either directly or by proxy - and therefore it's crucial not only to understand what we're doing, but also to sanity-check ourselves.

Additionally, Excel can only handle limited amounts of data before it starts jamming. On the other hand, even the most naive Python code wouldn't mind handling datasets regardless of their size (either horizontal or vertical).

If one practices the power of code, using it becomes a second nature. That, in turn, allows applying it more easily and accurately when circumstances dictate.