From: Junio C Hamano Date: Wed, 27 May 2015 19:48:45 GMT Subject: Re: [PATCH 3/5] verify_lock(): report errors via a strbuf Message-ID: In-Reply-To: <1432337697-29161-4-git-send-email-mhagger@alum.mit.edu> Michael Haggerty writes: > Instead of writing error messages directly to stderr, write them to a > "strbuf *err". In lock_ref_sha1_basic(), arrange for these errors to > be returned to its caller. I had to scratch my head and view long outside the context before realizing that the caller lock_ref_sha1_basic() already arranges with its caller that errors from it are passed via the strbuf, and this change is just turning verify_lock(), a callee from it, follows that already established pattern. Looks sensible, but the last sentence was misleading at least to me. The caller, lock_ref_sha1_basic(), uses this error reporting convention with all the other callees, and reports its error this way to its callers. perhaps?