Excel and LibreCalc Are Not Databases, But Are Good Tools for Viewing Data

It’s a popular meme that Excel is not a database. And I mean, they aren’t wrong.

But spreadsheets are a good way for looking at tabular data and manipulating data when you need to quickly, manually edit many rows of data at once. I woud argue there aren’t many good web apps for manipulating a spreadsheet’s worth of data with quick filters, sorts, and searches quite as easy as your typical spreadsheet app.

Spreadsheets have their place for manipulating individual rows quickly, doing data entry and searches of smaller datasets. I often find myself using spreadsheets as a way to read and edit dumps from databases like duckdb and db2. Often I will pull data out of a database using SQL, then format it up, purty it up with some R code before sending it to a client, or looking at in Excel. While RStudio can “preview” datasets, the interface is clumsy and slow, and in most cases for all but the quickest views, its best to dump to an Excel file or at least a CSV file and opening up in a spreadsheet.

Would I store a database in Excel? Hell no, that would be slow and error prone. But I will input and output data using Excel, and send data in that format to clients, as few of them know what an Apache Parquet file is, and while most can open a CSV, nothing beats a beautifully formatted Excel file output from OpenXLSX in R.