# Should --pretty=format:%b respect i18n.logoutputencoding just like normal "git log"? 2 messages from 2012-11-19 to 2012-11-19. Participants: Nguyen Thai Ngoc Duy, Pat Notz. Thread: https://gitlist.dev/t/32146 ## Nguyen Thai Ngoc Duy, 2012-11-19 13:08 Subject: Should --pretty=format:%b respect i18n.logoutputencoding just like normal "git log"? Message-ID: URL: https://gitlist.dev/e/CACsJy8Ag7wZCKeTM-fW7MUjWs%3D7xq7ciwhfx4kroJupwvs0dXg%40mail.gmail.com ``` Changes around this code in history is a bit unclear. User format learns to get log encoding from a field in 177b29d (pretty.c: teach format_commit_message() to reencode the output - 2010-11-02), but this field is only set for --fixup and --squash (in a few commits later). This makes "git log --pretty=format:%b" always ignore the output encoding config key. I don't think %b output should be different from normal log output, which does respect output encoding. Pat, any reasons not to do it? -- Duy ``` ## Pat Notz, 2012-11-19 17:42 Subject: Re: Should --pretty=format:%b respect i18n.logoutputencoding just like normal "git log"? Message-ID: URL: https://gitlist.dev/e/CADvGaErNPyHEP029-M-UWtDUaH0xJstXUc%2B9FGuy-7YEUM0zpA%40mail.gmail.com In-Reply-To: ``` On Mon, Nov 19, 2012 at 6:08 AM, Nguyen Thai Ngoc Duy wrote: > Changes around this code in history is a bit unclear. User format > learns to get log encoding from a field in 177b29d (pretty.c: teach > format_commit_message() to reencode the output - 2010-11-02), but this > field is only set for --fixup and --squash (in a few commits later). > This makes "git log --pretty=format:%b" always ignore the output > encoding config key. I don't think %b output should be different from > normal log output, which does respect output encoding. Pat, any > reasons not to do it? Nope, that looks like it was an oversight on my part. Nice find. ~ Pat > -- > Duy ```