“pelzflorian (Florian Pelz)” <pelzflor...@pelzflorian.de> writes:
> Liam Hupfer <l...@hpfr.net> writes: > >> The commit message’s purpose isn’t to enable people to avoid reading the >> diff; that’s an impossible goal. It’s to contextualize the diff. > > No, I used to and plan to again read `git log` and all commit messages, > and would not read the diff unless interesting. The information on what > files changed (maybe files I care about particularly?) and approximately > how they changed is something I did want to have an overview of. Here I use git-log’s path filtering to exclude files I don’t care about or vice versa. This is faster than manually reading through handwritten file listings to find files I care about. ‘:!*/bioconductor.scm,:!*/bioinformatics.scm,:!*/statistics.scm,:!*/cran.scm’ will make things more digestible if you aren’t an R user… > No, I would not like this, although I currently do not anymore follow > all the commits, so I would not disapprove either. Yeah, reading the entire log doesn’t scale to large projects like the kernel or package distributions. Like drinking from a firehose 😄. > I don’t care about duplicates of what changed, when it was in the commit > message already, though. I don’t think ChangeLog style requires this. Thanks for sharing your use case! It’s good to understand what people like about the current approach. —Liam
