All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jeff King <peff@peff.net>
To: Junio C Hamano <gitster@pobox.com>
Cc: Devin Lehmacher <lehmacdj@gmail.com>, git@vger.kernel.org
Subject: Re: [GSoC][PATCH 2/3] credential-cache.c: Make git use XDG_CACHE_HOME for credentials
Date: Mon, 13 Mar 2017 15:03:49 -0400	[thread overview]
Message-ID: <20170313190349.wbg7nfmrxvvzc5zv@sigill.intra.peff.net> (raw)
In-Reply-To: <xmqqtw6xrquw.fsf@gitster.mtv.corp.google.com>

On Mon, Mar 13, 2017 at 11:09:11AM -0700, Junio C Hamano wrote:

> > +	if (home_socket)
> > +		if (file_exists(home_socket))
> > +			return home_socket;
> > +		else
> > +			free(home_socket);
> > +
> > +	return xdg_cache_home("credential/socket");
> > +}
> 
> I somehow feel that the order of precedence should be the other way
> around, though.  
> 
> If somebody really wants to use the xdg location and has a socket
> already there, using that existing socket would be the right thing
> to do.  However, when neither ~/.git-credential-cache/socket nor
> ~/.cache/git/socket exists, why should we prefer the latter over the
> former?

How would they get a socket in the xdg location if we never create one?

I think the point of this commit is that we should generally prefer the
xdg locations to ones that were simply made up by Git.

As an aside, I also wondered if stale socket files might cause us to
keep picking the "old" location forever. But it looks like
credential-cache--daemon does use a tempfile handler to make sure it
cleans up the socket afterwards (and we are checking for the actual
socket, not just the presence of the outer directory, which is good).

-Peff

  reply	other threads:[~2017-03-13 19:03 UTC|newest]

Thread overview: 54+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-03-13 17:22 [GSoC][PATCH 0/3] Move ~/.git-credential-cache to ~/.cache/git Devin Lehmacher
2017-03-13 17:22 ` [GSoC][PATCH 1/3] path.c: Add xdg_cache_home to get paths under XDG_CACHE_HOME Devin Lehmacher
2017-03-13 18:48   ` Jeff King
2017-03-13 19:44     ` Devin Lehmacher
2017-03-13 17:22 ` [GSoC][PATCH 2/3] credential-cache.c: Make git use XDG_CACHE_HOME for credentials Devin Lehmacher
2017-03-13 18:09   ` Junio C Hamano
2017-03-13 19:03     ` Jeff King [this message]
2017-03-13 19:05       ` Junio C Hamano
2017-03-13 19:24         ` Devin Lehmacher
2017-03-13 19:22     ` Devin Lehmacher
2017-03-13 18:11   ` Junio C Hamano
2017-03-13 19:01   ` Jeff King
2017-03-13 17:22 ` [GSoC][PATCH 3/3] Update documentation to reflect new socket location Devin Lehmacher
2017-03-13 18:04   ` Junio C Hamano
2017-03-13 20:42   ` [GSoC][PATCH v2 2/2] credential-cache: use XDG_CACHE_HOME for socket Devin Lehmacher
2017-03-13 20:43   ` [GSoC][PATCH v2 1/2] path.c: add xdg_cache_home Devin Lehmacher
2017-03-13 20:43     ` [GSoC][PATCH v2 2/2] credential-cache: use XDG_CACHE_HOME for socket Devin Lehmacher
2017-03-13 21:52       ` Junio C Hamano
2017-03-14  0:32         ` [GSoC][PATCH/RFC v3 0/3] Fix commit messages, check if socket is socket Devin Lehmacher
2017-03-14  0:32           ` [GSoC][PATCH/RFC v3 1/3] path.c: add xdg_cache_home Devin Lehmacher
2017-03-14  0:32           ` [GSoC][PATCH/RFC v3 2/3] credential-cache: use XDG_CACHE_HOME for socket Devin Lehmacher
2017-03-14  1:50             ` Junio C Hamano
2017-03-14  0:32           ` [GSoC][PATCH/RFC v3 3/3] credential-cache: only use user_socket if a socket Devin Lehmacher
2017-03-14  0:40             ` Devin Lehmacher
2017-03-14  0:44               ` Brandon Williams
2017-03-14  1:30                 ` Devin Lehmacher
2017-03-14 17:03                   ` Brandon Williams
2017-03-14  1:52             ` Junio C Hamano
2017-03-14  6:10               ` Junio C Hamano
2017-03-14  6:17                 ` Devin Lehmacher
2017-03-16  5:18             ` [GSoC][PATCH v4 0/4] Moving credential-cache socket to xdg path Devin Lehmacher
2017-03-16  5:18               ` [GSoC][PATCH v4 1/4] path.c: add xdg_cache_home Devin Lehmacher
2017-03-16  5:18               ` [GSoC][PATCH v4 2/4] dir: add directory_exists Devin Lehmacher
2017-03-16 16:09                 ` Junio C Hamano
2017-03-16 18:26                   ` Junio C Hamano
2017-03-16  5:18               ` [GSoC][PATCH v4 3/4] credential-cache: use XDG_CACHE_HOME for socket Devin Lehmacher
2017-03-16 16:20                 ` Junio C Hamano
2017-03-16 18:02                   ` Jeff King
2017-03-16  5:18               ` [GSoC][PATCH v4 4/4] credential-cache: add tests for XDG functionality Devin Lehmacher
2017-03-16 16:29                 ` Junio C Hamano
2017-03-16 17:58                   ` Jeff King
2017-03-17  2:53                 ` [GSoC][PATCH v5 1/3] path.c: add xdg_cache_home Devin Lehmacher
2017-03-17  2:53                   ` [GSoC][PATCH v5 2/3] credential-cache: use XDG_CACHE_HOME for socket Devin Lehmacher
2017-03-17 16:12                     ` Ramsay Jones
2017-03-17 19:31                       ` Devin Lehmacher
2017-03-17  2:53                   ` [GSoC][PATCH v5 3/3] credential-cache: add tests for XDG functionality Devin Lehmacher
2017-03-17  5:38                     ` Junio C Hamano
2017-03-17 12:36                     ` [GSoC][PATCH v6 1/3] path.c: add xdg_cache_home Devin Lehmacher
2017-03-17 12:36                       ` [GSoC][PATCH v6 2/3] credential-cache: use XDG_CACHE_HOME for socket Devin Lehmacher
2017-03-17 13:42                         ` Jeff King
2017-03-17 12:36                       ` [GSoC][PATCH v6 3/3] credential-cache: add tests for XDG functionality Devin Lehmacher
2017-03-17 13:38                         ` Jeff King
2017-03-17 13:42                       ` [GSoC][PATCH v6 1/3] path.c: add xdg_cache_home Jeff King
2017-03-13 21:39     ` [GSoC][PATCH v2 1/2] " 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=20170313190349.wbg7nfmrxvvzc5zv@sigill.intra.peff.net \
    --to=peff@peff.net \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=lehmacdj@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.