"If a thing's worth doing, it's worth doing badly." I thought this was mine for years, but Chesterton beat me to it. Generally, if something is worthwhile, but

"If a thing's worth doing, it's worth doing badly." I thought this was mine for years, but Chesterton beat me to it. Generally, if something is worthwhile, but not critical, it's worth spending the 20% of the time to get it 80% of the way there, and doing five such slipshod jobs, rather than one perfect one. This is contrary to every reasonable software engineering practice, which is probably because it works better for things which aren't software, like user testing or business planning.