From: Jonathan Nieder <jrnieder@gmail.com>
To: Johannes Sixt <j6t@kdbg.org>
Cc: Junio C Hamano <gitster@pobox.com>,
Ben Walton <bwalton@artsci.utoronto.ca>,
David Roundy <roundyd@physics.oregonstate.edu>,
GIT List <git@vger.kernel.org>
Subject: Re: [PATCH 3/8] Teach git var about GIT_EDITOR
Date: Fri, 30 Oct 2009 17:47:38 -0500 [thread overview]
Message-ID: <20091030224737.GA16565@progeny.tock> (raw)
In-Reply-To: <4AEB51C6.7060204@kdbg.org>
Johannes Sixt wrote:
> Jonathan Nieder schrieb:
>> From: Johannes Sixt <j.sixt@viscovery.net>
[...]
> Thanks for cleaning up behind me. I don't mind if you take
> authorship, but if you do keep my name, please make it:
>
> From: Johannes Sixt <j6t@kdbg.org>
Thanks for the catch.
>>+ error(
>> "No editor specified in GIT_EDITOR, core.editor, VISUAL,\n"
>> "or EDITOR. Tried to fall back to vi but terminal is dumb.\n"
>> "Please set one of these variables to an appropriate\n"
>> "editor or run again with options that will not cause an\n"
>> "editor to be invoked (e.g., -m or -F for git commit).");
>>+ return NULL;
>>+ }
>
> I somehow dislike that this huge error message is in git_editor().
Makes sense.
I am a bit uncomfortable with this error in general. It makes some
sense to refuse to use $VISUAL and fall back to $EDITOR if TERM=dumb,
since without this the distinction between VISUAL and EDITOR is not
very useful. But wouldn’t that check be equally useful if GIT_EDITOR
or core.editor is set to vi? Ideally, vi itself would check TERM and
error out, and git would only need to report and handle the exit.
Unfortunately, at least vim is happy to assume a terminal supports
ANSI sequences even if TERM=dumb (e.g., when running from a text
editor like Acme). Unless VISUAL, GIT_EDITOR, and core.editor are
unset, nobody gets the benefit of this check.
Should git error out rather than run $VISUAL when TERM=dumb? How
about $GIT_EDITOR?
The advice about options to avoid invoking an editor is not very
helpful except with 'git commit', so probably only 'git commit' should
print that message.
> The return value, NULL, should be indication enough for the callers
> to handle the situation suitable.
Good idea.
> In particular, launch_editor()
> wants to write this big warning, but 'git var -l' can avoid the
> error message and write only a short notice:
>
> GIT_EDITOR=terminal is dumb, but VISUAL and EDITOR unset
Maybe 'git var -l' should omit GIT_EDITOR in this situation.
Thanks for the comments,
Jonathan
next prev parent reply other threads:[~2009-10-30 22:37 UTC|newest]
Thread overview: 65+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-10-28 15:21 packaging vs default pager Ben Walton
2009-10-28 17:55 ` Junio C Hamano
2009-10-29 7:32 ` [PATCH 0/2] " Jonathan Nieder
2009-10-29 7:45 ` [PATCH 1/2] Provide a build time default-pager setting Jonathan Nieder
2009-10-29 7:50 ` [PATCH/RFC 2/2] Provide a build time default-editor setting Jonathan Nieder
2009-10-29 10:36 ` David Roundy
2009-10-29 11:50 ` Johannes Sixt
2009-10-29 20:40 ` Junio C Hamano
2009-10-29 20:57 ` Johannes Sixt
2009-10-29 22:12 ` Junio C Hamano
2009-10-30 2:21 ` David Roundy
2009-10-29 20:43 ` Junio C Hamano
2009-10-30 10:16 ` [PATCH v2 0/8] Default pager and editor Jonathan Nieder
2009-10-30 10:20 ` [PATCH 1/8] launch_editor: Longer error message when TERM=dumb Jonathan Nieder
2009-10-30 10:25 ` [PATCH 2/8] Handle more shell metacharacters in editor names Jonathan Nieder
2009-10-30 10:26 ` [PATCH 3/8] Teach git var about GIT_EDITOR Jonathan Nieder
2009-10-30 20:51 ` Johannes Sixt
2009-10-30 22:47 ` Jonathan Nieder [this message]
2009-10-30 22:43 ` Junio C Hamano
2009-10-31 0:01 ` Jonathan Nieder
2009-10-30 10:29 ` [PATCH 4/8] Teach git var about GIT_PAGER Jonathan Nieder
2009-10-30 10:32 ` [PATCH 5/8] add -i, send-email, svn, p4, etc: use "git var GIT_EDITOR" Jonathan Nieder
2009-10-30 10:33 ` [PATCH 6/8] am -i, git-svn: use "git var GIT_PAGER" Jonathan Nieder
2009-10-30 10:35 ` [PATCH 7/8] Provide a build time default-editor setting Jonathan Nieder
2009-10-30 13:17 ` Jonathan Nieder
2009-10-30 10:39 ` [PATCH 8/8] Provide a build time default-pager setting Jonathan Nieder
2009-10-30 22:59 ` Junio C Hamano
2009-10-30 10:49 ` [PATCH/RFC 9/8] Teach git var to run the editor Jonathan Nieder
2009-10-31 1:20 ` [PATCH v3 0/8] Default pager and editor Jonathan Nieder
2009-10-31 1:24 ` [PATCH 1/8] Handle more shell metacharacters in editor names Jonathan Nieder
2009-10-31 1:30 ` [PATCH 2/8] Do not use VISUAL editor on dumb terminals Jonathan Nieder
2009-10-31 7:46 ` [PATCH v2 " Jonathan Nieder
2009-10-31 1:39 ` [PATCH v2 3/8] Teach git var about GIT_EDITOR Jonathan Nieder
2009-10-31 2:01 ` Junio C Hamano
2009-10-31 2:23 ` Jonathan Nieder
2009-10-31 2:34 ` Junio C Hamano
2009-10-31 4:00 ` Jonathan Nieder
2009-10-31 4:04 ` [PATCH v3] " Jonathan Nieder
2009-10-31 4:53 ` Jonathan Nieder
2009-10-31 7:56 ` [PATCH v4] " Jonathan Nieder
2009-11-01 4:29 ` Junio C Hamano
2009-10-31 19:40 ` [PATCH v2 3/8] " Johannes Sixt
2009-10-31 1:41 ` [PATCH 4/8] Teach git var about GIT_PAGER Jonathan Nieder
2009-10-31 1:42 ` [PATCH 5/8] add -i, send-email, svn, p4, etc: use "git var GIT_EDITOR" Jonathan Nieder
2009-10-31 1:43 ` [PATCH 6/8] am -i, git-svn: use "git var GIT_PAGER" Jonathan Nieder
2009-10-31 1:44 ` [PATCH 7/8] Provide a build time default-editor setting Jonathan Nieder
2009-10-31 2:09 ` Junio C Hamano
2009-10-31 3:26 ` Jonathan Nieder
2009-10-31 19:51 ` Junio C Hamano
2009-10-31 21:21 ` Jonathan Nieder
2009-11-01 4:29 ` Junio C Hamano
2009-10-31 1:45 ` [PATCH 8/8] Provide a build time default-pager setting Jonathan Nieder
2009-11-11 23:51 ` [PATCH v4 0/9] Default pager and editor Jonathan Nieder
2009-11-11 23:52 ` [PATCH 1/9] Handle more shell metacharacters in editor names Jonathan Nieder
2009-11-11 23:56 ` [PATCH 2/9] Do not use VISUAL editor on dumb terminals Jonathan Nieder
2009-11-11 23:57 ` [PATCH 3/9] Suppress warnings from "git var -l" Jonathan Nieder
2009-11-12 0:01 ` [PATCH 4/9] Teach git var about GIT_EDITOR Jonathan Nieder
2009-11-12 0:02 ` [PATCH 5/9] Teach git var about GIT_PAGER Jonathan Nieder
2009-11-12 0:02 ` [PATCH 6/9] add -i, send-email, svn, p4, etc: use "git var GIT_EDITOR" Jonathan Nieder
2009-11-12 0:03 ` [PATCH 7/9] am -i, git-svn: use "git var GIT_PAGER" Jonathan Nieder
2009-11-12 0:03 ` [PATCH 8/9] Provide a build time default-editor setting Jonathan Nieder
2009-11-12 0:04 ` [PATCH 9/9] Provide a build time default-pager setting Jonathan Nieder
2009-11-15 9:04 ` [PATCH v4 0/9] Default pager and editor Junio C Hamano
2009-10-29 16:42 ` [PATCH 0/2] Default Pager and Editor at build-time Ben Walton
2009-10-29 16:42 ` [PATCH 1/2] Provide a build time default-pager setting Ben Walton
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=20091030224737.GA16565@progeny.tock \
--to=jrnieder@gmail.com \
--cc=bwalton@artsci.utoronto.ca \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=j6t@kdbg.org \
--cc=roundyd@physics.oregonstate.edu \
/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).