4 ms·

> You use structure, design away the complexity, implement constraints, error checking and tests. I've never thought of Excel as an ideal tool for any of these

• ACow_Adonis 6y ago
  depending on your point of view, you don't do them in the "same" way (I.e with separate unit tests or compiler enforced safety), and explicit difs are hard/impossible. And ideal is dependent on the task (it's ideal for me to get a working interactive graphic + visualisation to a user behind a corporate firewall via email in the same afternoon I get the request, I wouldn't actually choose to do 'programming' in Excel, but I consider a budget small and simple and programming is probably overkill) what you do is much closer to old-school low level programming. define the relationships between your tables and variables well, set up explicit corresponding arrays of 1s and 0s that are themselves error checks on the underlying structure/ contents of your tables, calculate things in two different spots/ways and verify equivalence holds, and use simple red/green conditional formatting to draw attention to when things fall outside of expected state, etc.