[jiveturkey]

what ive found works is to /always/ pair a jr with a sr engineer to write any ddoc. 2nd, always assign a specific sr reviewer. only after that review, release the hounds. others that have interest or particular insight can reflect on deficiencies (or, rarely, strengths) without the dread feeling of having to deeply understand the context or underlying dependencies.

same reason you don’t just throw a code review out to “everyone”. everyone = no one