[tptacek]

Raymond's "Cathedral" post is a contender for one of the most overrated pieces of technical writing. "The parts that are interesting are not new, and the parts that are new aren't good". It's justifiably infamous for the now-discredited "Linus's law", about "many eyes making all bugs shallow", but large parts of it are also shoplifted from Brooks and Pike.

I think it's remembered fondly mostly because it was an effective bit of advocacy written during a time period people view fondly.

[statictype]

I feel like security bugs are in a class of their own and need different ways of thinking about them.

If you leave aside security bugs, is Linus’ law still invalid?

Any reference to any material on this?

[pvg]

https://ai.googleblog.com/2006/06/extra-extra-read-all-about...

But more importantly, it says 'all bugs' not 'all bugs except for these other bugs which would make the whole thing untrue'

[tptacek]

It's still not true if you leave security bugs out. It's basically never true except for a thin class of superficial bugs --- the bugs you'd intuitively expect to get diagnosed get diagnosed, but nothing else does.