The Optimal Line Length for Code: Unraveling the 88-Character Mystery

2025-08-12

This article delves into the age-old debate of optimal line length in coding standards. It traces the history of the 80-character limit back to physical constraints of teletypes and punch cards, then explores the scientific basis for line length limitations rooted in human visual perception. While acknowledging the advantages of wide screens, the author argues that excessively long lines reduce readability. The article ultimately advocates for a 88-character maximum, justifying the choice through considerations of indentation, diff tool compatibility, and overall code clarity.

Read more

R0ML's Ratio: Avoid the Bozo Trap in Enterprise Software Licensing

2025-08-10

This article introduces a clever methodology for evaluating volume purchases: R0ML's Ratio. Using the example of buying thousands of clown noses, it explains how to calculate the ratio: divide the total purchase price by the full retail price of all units. A ratio under 1 indicates a good deal; above 1 means you've been had. This is especially crucial for software and SaaS licensing, where accurately estimating usage is key to avoiding losses from underutilization. The author suggests empowering employees with corporate cards for individual software purchases as a safer alternative.

Read more

Why I Gave Up on GenAI Criticism

2025-06-05

The author, a self-described "thinky programmer," has long been skeptical of generative AI. Drowning in the constant discourse, he attempts to logically frame his concerns, but ultimately fails. The article delves into his negative experiences with genAI, encompassing its aesthetic flaws, productivity issues, ethical concerns, energy consumption, impact on education, and privacy violations. Despite presenting numerous arguments, he admits he can't rigorously refute pro-AI proponents. He ultimately surrenders, recognizing the prohibitive cost and futility of combating the immense influence of generative AI.

Read more