Re: [PATCH] git-difftool: allow skipping file by typing 'n' at prompt

[Sitaram Chamarty <sitaramc@gmail.com>]

On Wed, Oct 5, 2011 at 12:58 AM, Junio C Hamano <gitster@pobox.com> wrote:
> Phil Hord <phil.hord@gmail.com> writes:
>
>> On Tue, Oct 4, 2011 at 11:25 AM, Junio C Hamano <gitster@pobox.com> wrote:
>>
>>> I think I've seen this done as: "do this? [Y/n]" elsewhere.
>>>
>>> Not telling you what to do, but trying to feel what others may think.
>>
>> I think so, too.  The [y]/n syntax is not clear enough for me to
>> confidently know what the default value will be.
>
> One downside of "do this [Y,n,m,o,p,q]? " is that it limits us to
> lowercase responses, which means we cannot assign 'q' for quitting from
> the innermost nested context and assign 'Q' for quitting from the whole
> interactive loop (e.g. "git add -p").
>
>    "do this [y,n,m,o,p,q] (default=y)? "

Does this even make a difference in this case?  I was going to send
out a new patch using [Y/n] instead of my original [y]/n.  There's
only one loop in this thing, and till now people have been presumably
hitting Ctrl-C to get out of it.  I see no real need to make that more
elegant; all I set out to do is add one teeny weeny bit of
functionality to a prompt that -- other than giving you a chance to
hit that Ctrl-C -- was not actually doing anything useful at all.

>
> may have been a better choice in hindsight.
>
> No matter what we end up doing, let's try to be consistent.

The only other part of git where I have ever used a prompt is 'git add
-p'.  Consistency with *that* prompt, to me, would mean colors.  And
help text.  And I'm not sure what else, really, since I only used it
superficially.

Isn't that overkill for this case?

I'll wait a few hours for any further comments then send out a patch
that is the same as my original one except it uses [Y/n] instead of
[y]/n.