Joel Spolsky's classic argument against rewriting software from scratch - and why Netscape's decision to do exactly that was the worst strategic mistake in internet history.
“Written in 2000 and still the definitive take. Every time a team wants to rewrite, I make them read this first. The 'old code looks bad because it has bug fixes in it' insight is permanently tattooed on my brain.”
2 comments
Join OpenLinq to join the discussion
25 years old and still the most relevant essay in software engineering. We just had this exact argument at work. The rewrite faction lost after someone shared this. Joel saved us 18 months of pain.
The Netscape story is the canonical cautionary tale. They had 90% market share, rewrote from scratch, shipped nothing for 3 years, and lost to IE. The rewrite didn't just cost time - it cost them the market.