Git development
 help / color / mirror / Atom feed
From: Patrick Steinhardt <ps@pks.im>
To: Junio C Hamano <gitster@pobox.com>
Cc: "Phillip Wood" <phillip.wood123@gmail.com>,
	"SZEDER Gábor" <szeder.dev@gmail.com>,
	git@vger.kernel.org,
	"brian m. carlson" <sandals@crustytoothpaste.net>,
	"Elijah Newren" <newren@gmail.com>,
	"Derrick Stolee" <stolee@gmail.com>,
	"Phillip Wood" <phillip.wood@dunelm.org.uk>
Subject: Re: [PATCH RFC v2 2/2] Move libgit.a sources into separate "lib/" directory
Date: Thu, 2 Jul 2026 07:21:16 +0200	[thread overview]
Message-ID: <akX1TMoRr87Id8Ss@pks.im> (raw)
In-Reply-To: <xmqq1pdmrcp6.fsf@gitster.g>

On Wed, Jul 01, 2026 at 07:45:09AM -0700, Junio C Hamano wrote:
> Phillip Wood <phillip.wood123@gmail.com> writes:
> 
> > As I said last time this came up, I don't really buy the discoverability 
> > argument because there are just as many files to trawl through to find 
> > what you're looking through and now there is an extra directory to 
> > check. I think the solution to that is to recommend folks use "git grep" 
> > or ctags etc. not moving code to a new directory.
> 
> Hear, hear.  Also it would be great if we can trick some talented
> technical writer into writing the "map" of the source so that by
> reading this one or two pager, any new person with reasonable
> competence will know how things are partitioned into pieces and how
> these pieces fit together.  I wonder how good LLMs are these days?
> ;-)

This isn't about discoverability of the library files though, I
specifically want to improve discoverability of all the other files that
we have in our root directory. So yes, I fully agree that this change
does not help to make that one file that is part of our library easier
to find.

> > I do however think putting all the library code in a subdirectory makes 
> > it easier to say things like "please try to avoid new uses of 
> > 'the_repository' and prefer 'error()' over 'die()' in library code" 
> > because all the library code is in the same directory. I think that is a 
> > much stronger selling point.
> 
> Yes.  "library code (things outside the subdirectories) should not
> use X" would work just fine, though.

That rule doesn't quite work:

  - We have several C files that are not library files and that are in
    the top-level directory. For example "scalar.c" or "shell.c".

  - We have several C files that are part of the library and that are in
    a subdirectory. For example "compat/", "refs/", "odb/".

So having this properly cleaned up would help to have clear indicators
what component a given file belongs to.

> > Another cost is remembering things have moved - the other day I spent 
> > too long wondering why "git show origin/seen:wt-status.c" wasn't working 
> > until I ran "git log origin/seen" and realized it had move to 
> > lib/wt-status.c.
> 
> Yes, this has bit me multiple dozen times, as the tip of 'seen' is
> contaminated with this rename, already.  It is a huge pain.

Yeah, this one I don't have any arguments against besides a very
hand-wavy "it'll get better over time" :)

Thanks!

Patrick

  reply	other threads:[~2026-07-02  5:21 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-16 13:24 [PATCH RFC 0/2] Move libgit.a sources into separate "lib/" directory Patrick Steinhardt
2026-04-16 13:24 ` [PATCH RFC 1/2] t/helper: prepare "test-example-tap.c" for introduction of "lib/" Patrick Steinhardt
2026-04-16 13:24 ` [PATCH RFC 2/2] Move libgit.a sources into separate "lib/" directory Patrick Steinhardt
2026-04-17 17:08   ` Elijah Newren
2026-04-17 19:18   ` brian m. carlson
2026-04-17 21:18     ` Junio C Hamano
2026-04-17 21:51       ` brian m. carlson
2026-04-20  6:41         ` Patrick Steinhardt
2026-04-19 14:11 ` [PATCH RFC 0/2] " Phillip Wood
2026-04-20  6:41   ` Patrick Steinhardt
2026-04-20 12:03 ` Derrick Stolee
2026-04-21  5:55   ` Patrick Steinhardt
2026-04-21 14:13     ` Derrick Stolee
2026-04-22  6:39       ` Patrick Steinhardt
2026-06-22 10:38 ` [PATCH RFC v2 " Patrick Steinhardt
2026-06-22 10:38   ` [PATCH RFC v2 1/2] t/helper: prepare "test-example-tap.c" for introduction of "lib/" Patrick Steinhardt
2026-06-22 10:38   ` [PATCH RFC v2 2/2] Move libgit.a sources into separate "lib/" directory Patrick Steinhardt
2026-06-22 13:08     ` Junio C Hamano
2026-06-24 10:32       ` Patrick Steinhardt
2026-06-24 11:23       ` Oswald Buddenhagen
2026-06-26 16:01     ` Johannes Schindelin
2026-06-26 18:50       ` Junio C Hamano
2026-07-01  6:54       ` Patrick Steinhardt
2026-06-27  6:40     ` SZEDER Gábor
2026-07-01  6:55       ` Patrick Steinhardt
2026-07-01 13:26         ` Phillip Wood
2026-07-01 14:45           ` Junio C Hamano
2026-07-02  5:21             ` Patrick Steinhardt [this message]
2026-07-01  6:59 ` [PATCH RFC v3 0/2] " Patrick Steinhardt
2026-07-01  6:59   ` [PATCH RFC v3 1/2] t/helper: prepare "test-example-tap.c" for introduction of "lib/" Patrick Steinhardt
2026-07-01  6:59   ` [PATCH RFC v3 2/2] Move libgit.a sources into separate "lib/" directory Patrick Steinhardt

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=akX1TMoRr87Id8Ss@pks.im \
    --to=ps@pks.im \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=newren@gmail.com \
    --cc=phillip.wood123@gmail.com \
    --cc=phillip.wood@dunelm.org.uk \
    --cc=sandals@crustytoothpaste.net \
    --cc=stolee@gmail.com \
    --cc=szeder.dev@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