I suspect much of it has to do with the number of unexpected scenarios that widely deployed free software encounters. It doesn’t cost anything and any random pe

I suspect much of it has to do with the number of unexpected scenarios that widely deployed free software encounters. It doesn’t cost anything and any random person within an organization can figure out how to get it running and try it out (and submit patches/big reports/etc). Whereas most commercial software sees relatively few different deployment scenarios since they were built and bought to a smaller set of things.