Re: review "git range-diff -h" output

From: Duy Nguyen <pclouds@gmail.com>
Date: Dec 12, 2018, 20:09 UTC
Message-ID: <CACsJy8C4LkdB2E7qQEwgPniP5cVwNK9f2jgbbB_14o2LiBQL5w@mail.gmail.com>
In-Reply-To: <CAGZ79kZwS4J_bOd+2iTbVTJqMW_+P5Lzt1Zs92JBtfkBnj3juw@mail.gmail.com>

On Wed, Dec 12, 2018 at 9:05 PM Stefan Beller <sbeller@google.com> wrote:

Show 15 quoted lines

>
> On Wed, Dec 12, 2018 at 11:47 AM Duy Nguyen <pclouds@gmail.com> wrote:
> >
> > I'm not going to bother you with patches (yet) but I could use a few
> > eyeballs to check the help usage (and perhaps the option grouping) for
> > commands that take diff options.
>
> The grouping looks good from a cursory view, but ...
>
> > Forget about hiding irrelevant options for this particular command for
> > now. This is something we can do later.
>
> .. this seems relevant: range-diff is a diff of diffs,
> so for that particular command, I'd want to know if the option
> (e.g. --stat) is applied to the inner or outer diff.

Yeah I'll probably should send "git diff --no-index -h" instead.

> Now back to your topic: Are you planning to send patches that
> autogenerate these helps based on our source code?
> Should I pay attention to some specific part of the output?

All this comes from a new 'struct option' array in diff.c, very much like how we handle options everywhere else. So yeah sort of auto generation, but all the text has to be typed in manually first.

-- 
Duy