Git development
 help / color / mirror / Atom feed
From: "D. Ben Knoble" <ben.knoble+github@gmail.com>
To: git@vger.kernel.org
Cc: "D. Ben Knoble" <ben.knoble+github@gmail.com>,
	"brian m . carlson" <sandals@crustytoothpaste.net>,
	Patrick Steinhardt <ps@pks.im>, Taylor Blau <me@ttaylorr.com>,
	Caleb White <cdwhite3@pm.me>, Calvin Wan <calvinwan@google.com>,
	Junio C Hamano <gitster@pobox.com>,
	Elijah Newren <newren@gmail.com>,
	Andrew Berry <andrew@furrypaws.ca>, Jeff King <peff@peff.net>,
	Derrick Stolee <stolee@gmail.com>,
	Phillip Wood <phillip.wood@dunelm.org.uk>
Subject: [PATCH v2] ignore: note info/exclude lives in GIT_COMMON_DIR, not GIT_DIR
Date: Fri,  8 May 2026 10:14:14 -0400	[thread overview]
Message-ID: <d58b6e921d3005c6170fc6c47f175214acb3fa68.1778249267.git.ben.knoble+github@gmail.com> (raw)
In-Reply-To: <e3ee0a11b566dd2cc605447c111ae4620bce0fe6.1777050300.git.ben.knoble+github@gmail.com>

gitignore(5) says that the per-repository ignore file is
$GIT_DIR/info/exclude, but in a worktree that is not the case:

    git rev-parse --git-path info/exclude
    /path/to/main/worktree/.git/info/exclude
    git rev-parse --git-common-dir
    /path/to/main/worktree/.git

We actually use $GIT_COMMON_DIR/info/exclude. Adjust the documentation
to say so.

Signed-off-by: D. Ben Knoble <ben.knoble+github@gmail.com>
---

Notes (benknoble/commits):
    Changes in v2:
    
    Only adjust the documentation.
    
    brian points out that a more general extension would allow using more
    info/ files as "per-worktree," which I don't have the impetus to
    implement myself.
    
    Phillip and Junio asked for a concrete use case:
    
        A colleague is developing a tool for managing the "skill files" of
        various LLM tools (Claude, Windsurf, etc.). The files have
        requirements that make it hard to generically ignore them (e.g.,
        filenames and front-matter have to match), but different tasks
        (corresponding to worktrees) may want different active skills, so it
        is desirable to ignore the files. Think of this like node_modules.
    
        Unfortunately, since per-worktree ignores don't work, the current
        solution is to put a .gitignore file in the corresponding directory
        with the installed skills that ignores itself and the installed
        skills.
    
    Since overall reactions seem fairly negative (or require a more general
    extension, which I think is probably the right course but not simply
    implemented), I've opted to adjust the docs. They originally confused
    me, as I was surprised when my colleague reported that per-worktree
    ignores didn't work (the docs imply they should by use of $GIT_DIR).
    
    Link to v1: <e3ee0a11b566dd2cc605447c111ae4620bce0fe6.1777050300.git.ben.knoble+github@gmail.com>
    
    v1 notes:
    
    Discussed briefly at https://lore.kernel.org/git/CALnO6CCXmA+ATT7CuyWkU6P8qmLCCpMi5Ppr1c78s0heznpVyw@mail.gmail.com/T
    
    This is based on next (4f69b47b94 (Merge branch 'ps/test-set-e-clean'
    into next, 2026-04-23)) but cleanly applies to master (94f057755b (Git
    2.54, 2026-04-19)) and seen (50541634cb (Merge branch
    'js/parseopt-subcommand-autocorrection' into seen, 2026-04-23)).

 Documentation/gitignore.adoc | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/Documentation/gitignore.adoc b/Documentation/gitignore.adoc
index a3d24e5c34..c423b650de 100644
--- a/Documentation/gitignore.adoc
+++ b/Documentation/gitignore.adoc
@@ -7,7 +7,7 @@ gitignore - Specifies intentionally untracked files to ignore
 
 SYNOPSIS
 --------
-$XDG_CONFIG_HOME/git/ignore, $GIT_DIR/info/exclude, .gitignore
+$XDG_CONFIG_HOME/git/ignore, $GIT_COMMON_DIR/info/exclude, .gitignore
 
 DESCRIPTION
 -----------
@@ -34,7 +34,7 @@ precedence, the last matching pattern decides the outcome):
    includes such `.gitignore` files in its repository, containing patterns for
    files generated as part of the project build.
 
- * Patterns read from `$GIT_DIR/info/exclude`.
+ * Patterns read from `$GIT_COMMON_DIR/info/exclude`.
 
  * Patterns read from the file specified by the configuration
    variable `core.excludesFile`.

base-commit: 4f69b47b940100b02630f745a52f9d9850f122b2
-- 
2.54.0.564.ge3ee0a11b5.dirty


  parent reply	other threads:[~2026-05-08 14:15 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-24 17:09 [PATCH] dir: use per-worktree repository ignore patterns upon request D. Ben Knoble
2026-04-24 19:35 ` brian m. carlson
2026-04-24 19:53 ` Phillip Wood
2026-04-25  3:06   ` Junio C Hamano
2026-05-08 14:14 ` D. Ben Knoble [this message]
2026-05-09 14:08   ` [PATCH v2] ignore: note info/exclude lives in GIT_COMMON_DIR, not GIT_DIR brian m. carlson
2026-05-11 10:30   ` Phillip Wood
2026-05-11 19:55     ` D. Ben Knoble
2026-05-12 21:21   ` [PATCH v3] " D. Ben Knoble
2026-05-13 14:02     ` Phillip Wood

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=d58b6e921d3005c6170fc6c47f175214acb3fa68.1778249267.git.ben.knoble+github@gmail.com \
    --to=ben.knoble+github@gmail.com \
    --cc=andrew@furrypaws.ca \
    --cc=calvinwan@google.com \
    --cc=cdwhite3@pm.me \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=me@ttaylorr.com \
    --cc=newren@gmail.com \
    --cc=peff@peff.net \
    --cc=phillip.wood@dunelm.org.uk \
    --cc=ps@pks.im \
    --cc=sandals@crustytoothpaste.net \
    --cc=stolee@gmail.com \
    /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