Re: [RFC PATCH] scripts/get_maintainer.pl: add interactive mode

[Joe Perches]

On Mon, 2010-09-20 at 23:53 +0200, Florian Mickler wrote:
> On Mon, 20 Sep 2010 12:43:08 -0700
> Joe Perches <joe@xxxxxxxxxxx> wrote:
> > #(toggle), A#(author), S#(signed) *(all), ^(none), O(options), Y(approve): 
> I don't think the descriptions on the bottom are that descriptive. Nor
> can they be. There should probably be just a quick primer like:
> "([N]+:select) ([N]-:deselect) ([N]?:toggle infos) (Y:approve) (h:help)"
> and on "h" a long description gets displayed.

We disagree.  There could be a useful "H" help option.

> > Because git history is now not searched by default
> > when there is a named maintainer, there are no
> > commit signers.
> 
> Don't know if this is intuitive. If there is the possibility to have
> them shown but not selected, that would be ideal as it relieves the
> user from pressing extra keys while still having a sane behaviour.

Use a command line option: --git.

All command line options apply to create the initial list of
displayed names.

Or add code to set

	$git-fallback = 1 if $interactive;

> > 
> > If the "G" option is entered, you get:
> > 
> > #(toggle), A#(author), S#(signed) *(all), ^(none), O(options), Y(approve): g
> > --------------------------------------------------------------------------------
> > *  # email/list and role:stats                                        auth sign
> > *  1 Stephen Hemminger <shemminger@xxxxxxxxxxxxxxxxxxxx>                 1    0
> >      maintainer:SKGE, SKY2 10/100...,commit_signer:38/69=55%
> >    2 "David S. Miller" <davem@xxxxxxxxxxxxx>                             7   57
> >      commit_signer:57/69=83%
> >    3 Mike McCormack <mikem@xxxxxxxxxx>                                  16   16
> >      commit_signer:16/69=23%
> >    4 Joe Perches <joe@xxxxxxxxxxx>                                       4    4
> >      commit_signer:4/69=6%
> > *  5 netdev@xxxxxxxxxxxxxxx                                           
> >      open list:SKGE, SKY2 10/100...
> > *  6 linux-kernel@xxxxxxxxxxxxxxx                                     
> >      open list
> 
> I would try to fit reviewed-by in the statistics on the right, because
> that actually says a bit about the ability and willingness to review
> code... 

I think it's not worth it.

Only about 2% of signatures in git history are reviewed-by:

Using S shows the signature type.

Over the last year:
$ git log --since=1-year-ago | grep -i "by:.*@" | \
  cut -f1 -d":" | sort -i | uniq -ci | sort -rn | head -10
  83413     Signed-off-by
   6544     Acked-by
   2022     Reviewed-by
   1691     Reported-by
   1065     Tested-by
    111     Reported-and-tested-by
     83     Suggested-by
     31     Requested-by
     28         Signed-off-by
     26     Fixed-by

> [snip menu parsing code]
> 
> That should probably go in an extra function and be slimmed down, like
> I did in a later version I sent you.

Maybe.  I think it doesn't matter much though.
Menu handling code tends to get long.

> > +sub bool_invert {
> > +    my ($bool_ref) = @_;
> > +
> > +    if ($$bool_ref) {
> > +	$$bool_ref = 0;
> > +    } else {
> > +	$$bool_ref = 1;
> > +    }
>  +}
> 
> That should just be $$bool_ref = !$$bool_ref (and probably not a
> function)

I think it needs to be a function.
I want a 0 or 1, not "" or 1.


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/