AI Mode conversation: computer programming meme some practice "considered harmful"

You said: computer programming meme some practice "considered harmful"

computer programming meme some practice "considered harmful"

AI Mode reply for computer programming meme some practice "considered harmful"

The phrase "Considered Harmful" is one of computer science’s oldest and most enduring memes, originating from a famous 1968 letter by Edsger Dijkstra titled "Go To Statement Considered Harmful." Over the decades, programmers have used this exact phrasing to dramatically criticize everything from specific coding practices to everyday technologies.

Here is a breakdown of how the meme is practiced, its real-world impact, and its alternatives.

📜 The Original and Its Spinoffs

Dijkstra argued that the goto statement made code too complex and hard to track. Following his letter, the tech community turned the title into a fill-in-the-blank template to express strong disapproval.

Topic	What the Argument Claims
"Go To Statement Considered Harmful" (1968)	Spaghetti code is impossible to debug; use structured loops instead.
"Considered Harmful Essays Considered Harmful" (1987)	The meme format itself is lazy, dramatic, and shuts down real debate.
"Stored Procedures Considered Harmful"	Putting business logic in the database makes scaling and version control difficult.
"Shared State Considered Harmful"	Multi-threaded apps with shared memory cause unpredictable bugs and race conditions.

🛠️ How the Meme is Practiced Today

In modern software development, writing a "Considered Harmful" essay or post usually follows a specific formula:

The Target: Pick a widely used tool, design pattern, or feature (e.g., Microservices Considered Harmful or Inheritance Considered Harmful).
The Hyperbole: Use an authoritative tone to imply that using this tool is fundamentally dangerous or unprofessional.
The Nuance (Often Missing): While the original paper offered deep academic reasoning, modern uses are often clickbait blog posts or spicy tech-Twitter threads.

⚠️ The Risk of the Meme

While the phrase is a fun piece of hacker history, relying on it too heavily can lead to dogmatism.
In programming, very few things are always bad. For example, goto is still vital for error handling in low-level C programming and Linux kernel development. Declaring a tool "harmful" without context ignores the specific engineering trade-offs required for different projects.

If you are researching this for a specific project or just exploring tech history, I can help you dive deeper. Would you like to:

Look at modern examples of this meme in web development (like React or CSS)?
Understand the technical reason why Dijkstra hated the goto statement?
Draft a satirical or serious post using this template for a specific technology?

AI responses may include mistakes. Learn more

Shared

0 files

Show all

Please write a post titled "Spreadsheets Considered Harmful". It should be light-hearted in tone, but make the point that complex spreadsheets are hard to validate and debug.

AI Mode reply for Please write a post titled "Spreadsheets Considered Harmful". It should be light-hearted in tone, but make the point that complex spreadsheets are hard to validate and debug.

Spreadsheets Considered Harmful

The year is 1978. Dan Bricklin and Bob Frankston create VisiCalc, the world’s first electronic spreadsheet. Business analysts rejoice. Wall Street throws a party. Humanity suddenly has a way to turn rows of chaotic numbers into beautiful, structured grids.

Fast forward to today, and we have created a monster.

What started as a digital ledger has evolved into the accidental backbone of global infrastructure. Entire multi-billion-dollar financial institutions, logistics networks, and scientific research projects are held together by nothing more than digital duct tape, prayers, and a 50MB .xlsx file named Budget_2026_FINAL_v4_DEPRICATED_USE_THIS_ONE.xlsx.

It is time to face the truth: complex spreadsheets are considered harmful.

🕵️‍♂️ The Hidden Code That Nobody Can Read

Here is a radical truth that spreadsheet users rarely admit: a complex spreadsheet is actually a software program.

The only difference between a spreadsheet and a Python script is that Python code is designed to be read by humans. A spreadsheet hides its logic away in a claustrophobic 1-inch formula bar. If you want to audit a cell, you aren't reading clean, linear lines of logic. Instead, you are squinting at an existential riddle:

=IFERROR(VLOOKUP(A2, 'Raw Data'!$B$2:$E$100, 3, FALSE), SUMPRODUCT(--(Month=B$1), Revenue)))

To trace how a single number is calculated, you have to click through twelve different tabs, track down a hidden row that was collapsed in 2022, and blindly trust that the person who wrote the VLOOKUP didn't accidentally shift their index by one column.

🧪 The Disastrous Lack of Tests

When programmers write software, they use tools to make sure they haven't broken anything. They have version control (Git) to see exactly who changed what, code reviews to catch mistakes, and automated testing frameworks to verify that math stays mathematical.

When a spreadsheet user updates a massive model, their "validation process" usually looks like this:

Change a number in cell B14.
Scroll furiously to the bottom of the sheet.
Check if the final number "looks about right."
Save, close, and send it to the entire C-suite.

There are no unit tests in Excel. There is no pytest for Google Sheets. If an accidental keystroke replaces a formula with a static value, that silent killer will sit in row 452, entirely undetected, until an auditor wakes up in a cold sweat three years later.

🦹‍♂️ The "Silent Errata" Hall of Fame

This isn't just a theoretical problem; spreadsheet bugs have caused absolute chaos in the real world:

The Cut-and-Paste Catastrophe: A major investment firm once lost millions because a simple copy-paste error caused a formula to only calculate the total sum of half the rows it was supposed to.
The Genetic Mix-up: Scientists have had to rename human genes because Excel kept automatically converting gene symbols like SEPT4 (Septin 4) and MARCH1 into dates like 04-Sep and 01-Mar.
The Missing Rows: During a major public health tracking initiative, thousands of cases went completely missing simply because the data exceeded the maximum number of rows a standard spreadsheet could handle. The data just... fell off the edge of the earth.

🕊️ A Plea for Sanity

Spreadsheets are brilliant for a quick budget, a simple list, or sketching out an idea. They are the ultimate sandbox. But when that sandbox grows to include database joins, macros, and thousands of lines of hidden logic, it has outgrown its home.

If your spreadsheet requires a 10-page manual just to explain how to open it without crashing your RAM, it’s no longer a spreadsheet. It’s a poorly written application. Let’s do ourselves a favour: check it into Git, port it to a proper database, or rewrite it in a programming language where we can actually see the code.

Your sanity—and your data integrity—will thank you.

If you'd like to share this or adapt it further, let me know:

Who is the target audience? (e.g., software engineers, finance professionals, or general business teams?)
Would you like to add specific technical alternatives like Python, R, or SQL databases to the conclusion?
Should we lean further into the humour or add more real-world horror stories?

AI responses may include mistakes. Learn more

Shared

0 files

Show all

Shared

0 files

Transcribing…