On 27 April 2018 at 20:28, Andreas Heiduk <asheiduk@xxxxxxxxx> wrote: > Am 27.04.2018 um 19:18 schrieb Martin Ågren: >> On 27 April 2018 at 19:04, Andreas Heiduk <asheiduk@xxxxxxxxx> wrote: >>> The two '<path>' parameters are not optional but the option >>> '--no-index' is. Also move the `--options` part to the same >>> place where the other variants show them. >> >> Should this commit message be updated after the changes you did to >> address Junio's comment? This text suggests you want to place --no-index >> in [] (and you did in v1) but you do not do that below. >> >>> All three items are already correct in the synopsis. >> >> Same here, now you actually do change things there. >> >>> Signed-off-by: Andreas Heiduk <asheiduk@xxxxxxxxx> >>> Reviewed-by: Martin Ågren <martin.agren@xxxxxxxxx> >> >> Strictly speaking, my Reviewed-by was on another patch. I do find this > > Sorry, I've added that trailer after reading "The diff LGTM.", then > applied Junio's changes and forgot to remove the trailer. > >> one better though thanks to Junio's suggestion (except the mismatch with >> the commit message). > > I'll fix that with this: > > doc: align 'diff --no-index' in text with synopsis s/with/and/ since they both change? It's not that the first changes to match the second, but they actually both change to match each other (and to be correct, obviously). > Make the two '<path>' parameters in DESCRIPTION mandatory and > move the `--options` part to the same place where the other > variants show them. And finally make `--no-index` in SYNOPSIS > as mandatory as in DESCRIPTION. Great! Junio had some good reasoning about how --no-index is sometimes optional, but not always. Not sure if it's worth spelling that out. (Although one could argue that it already did trip us up once. :-)) Eric's point about "--options" vs "options" seemed right to me. If you address that, note that this message says "--options".
