'git format-patch <ref> --summary' seems to be turning off the stat

'git format-patch <ref> --summary' seems to be turning off the stat

[Kaartic Sivaraam]

The documentation for "--summary" option in the format-patch Doc states, 
--summary Output a condensed summary of extended header information such 
as creations, renames and mode changes. It doesn't state anything about 
turning of the stat. Why does the stat get turned off when "--summary" 
option is used along with "git format-patch" ? Note: I do see that 
"--summary" is a diff-option but does that mean we should't be printing 
stat information in the patch when the user didn't mention "--no-stat"? 
Yeah, "git format-patch <ref> --summary --stat" does bring back the 
stat. --- Kaartic

Re: 'git format-patch <ref> --summary' seems to be turning off the stat

[Junio C Hamano]

Kaartic Sivaraam <kaartic.sivaraam@gmail.com> writes:

> Note: I do see that "--summary" is a diff-option but does that mean we
> should't be printing stat information in the patch when the user
> didn't mention "--no-stat"? Yeah, "git format-patch <ref> --summary
> --stat" does bring back the stat. --- Kaartic

The old design decision shared across "log" family of commands [*1*]
wrt to these "how and what are shown" options is that they started
without any default set of options and the user gave all options (if
they want both summary and stat, they gave them from the command
line); over time they gained a set of built-in default options that
take effect only when the user gives no option.

So, yes, giving "--summary" will turn off the built-in ones that are
"--summary --stat" (IIRC).


[Footnote]

*1* "show" turns "-p" on by default, but with "show --stat" you can
    get "--stat" without the patch.  That comes from the same design.

Re: 'git format-patch <ref> --summary' seems to be turning off the stat

[Kaartic Sivaraam]

On Tuesday 19 December 2017 12:07 AM, Junio C Hamano wrote:
> Kaartic Sivaraam <kaartic.sivaraam@gmail.com> writes:
> 
>> Note: I do see that "--summary" is a diff-option but does that mean we
>> should't be printing stat information in the patch when the user
>> didn't mention "--no-stat"? Yeah, "git format-patch <ref> --summary
>> --stat" does bring back the stat. --- Kaartic
> 
> The old design decision shared across "log" family of commands [*1*]
> wrt to these "how and what are shown" options is that they started
> without any default set of options and the user gave all options (if
> they want both summary and stat, they gave them from the command
> line); over time they gained a set of built-in default options that
> take effect only when the user gives no option.
> 
> So, yes, giving "--summary" will turn off the built-in ones that are
> "--summary --stat" (IIRC).
>

Is it worth documenting that built-in defaults are turned off when user 
gives other options? Not sure where, though?


-- 
Kaartic