git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Felipe Contreras <felipe.contreras@gmail.com>
To: Charles Bailey <charles@hashpling.org>,
	Felipe Contreras <felipe.contreras@gmail.com>
Cc: David Aguilar <davvid@gmail.com>, git@vger.kernel.org
Subject: Re: [PATCH 2/2] mergetool: run prompt only if guessed tool
Date: Tue, 22 Apr 2014 02:56:22 -0500	[thread overview]
Message-ID: <535620a6e8274_3d2216372ec9d@nysa.notmuch> (raw)
In-Reply-To: <20140422073008.GA11584@hashpling.org>

Charles Bailey wrote:
> On Tue, Apr 22, 2014 at 01:53:46AM -0500, Felipe Contreras wrote:
> > Charles Bailey wrote:
> > > On Tue, Apr 22, 2014 at 01:24:09AM -0500, Felipe Contreras wrote:
> > > > 
> > > > This is what I get when a tool is not working:
> > > > 
> > > >   Documentation/config.txt seems unchanged.
> > > >   Was the merge successful? [y/n]
> > > 
> > > Does this happen now even with merge tools for which we do trust the
> > > exit code? If so, my original concern is addressed.
> > 
> > Which tools are those?
> 
> I didn't remember off hand, but checking the mergetools directory
> suggests that kdiff3 is one (there's no call to check_unchanged). I
> stopped checking after I found one.

So:

	% cat > ~/bin/kdiff3 <<-\EOF
	#!/bin/sh
	false
	EOF
	% chmod +x ~/bin/kdiff3
  % git -c merge.tool=kdiff3 mergetool
  merge of Documentation/config.txt failed
  Continue merging other unresolved paths (y/n) ?

> > You don't see anything wrong with asking the user *every single time* he runs
> > `git mergetool`, even though he *already told us* which tool to use?
> >  
> > If so, I'm pretty sure everybody else disagrees with you.
> 
> I think that you may have misunderstood me. As I said, I've no
> particular objections to changing the default (subject a few concerns
> that may or may not still apply).
> 
> Having said that, the fact that the user has configured the merge tool
> doesn't mean that he necessarily doesn't want to see the prompt.

Not necessarily, but in 99% of the cases it does. And for the remaining there's
always mergetool.prompt = true.

> In a part of my reply which you snipped, I said that it's sometimes the
> particular file that's due to be resolved that might prompt a user to want to
> skip launching the tool.

That's a possibility, however, in almost all the situations I've wanted to stop
a merge, it's *after* I've seen the actual conflicts in the file.

Anyway, I've revisited the code, and it's only now that I've realized that this:

  Hit return to start merge resolution tool (kdiff3):

Is not actually asking me for the tool I want to use; the value in parenthesis
is not the default, and I can't type in another tool.

So the purpose of the prompt is very different from what I was thinking, yet I
still think the value of such prompt is marginal at best.

> Either way, I think we shouldn't unconditionally override an explicitly
> set mergetool.prompt and if we are (effectively) changing the default we
> should probably update the documentation to say so as well. 

An explicitly set mergetool.prompt = true would override the default. See the
patch.

I looked, the documentation doesn't mention any default. We could add it, but I
don't think it's necesarily part of this patch.

-- 
Felipe Contreras

  reply	other threads:[~2014-04-22  8:06 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-04-21  0:17 [PATCH 0/2] Simple default fixes for v2.0 Felipe Contreras
2014-04-21  0:17 ` [PATCH 1/2] merge: enable defaulttoupstream by default Felipe Contreras
2014-04-22 19:53   ` Junio C Hamano
2014-04-21  0:17 ` [PATCH 2/2] mergetool: run prompt only if guessed tool Felipe Contreras
2014-04-22  4:59   ` David Aguilar
2014-04-22  6:01     ` Charles Bailey
2014-04-22  6:24       ` Felipe Contreras
2014-04-22  6:55         ` Charles Bailey
2014-04-22  6:53           ` Felipe Contreras
2014-04-22  7:30             ` Charles Bailey
2014-04-22  7:56               ` Felipe Contreras [this message]
2014-04-23  7:56                 ` Charles Bailey
2014-04-23 17:07                   ` Junio C Hamano
2014-04-24 18:30                     ` Junio C Hamano
2014-04-22 17:19     ` Junio C Hamano
2014-04-23  7:58       ` David Aguilar
2014-04-23 16:52         ` Junio C Hamano
2014-04-22 17:32     ` Junio C Hamano
2014-04-22 19:55       ` Junio C Hamano

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=535620a6e8274_3d2216372ec9d@nysa.notmuch \
    --to=felipe.contreras@gmail.com \
    --cc=charles@hashpling.org \
    --cc=davvid@gmail.com \
    --cc=git@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).