From: Daniel Höpfl <daniel@hoepfl.de>
Date: Wed, 02 Feb 2011 13:25:27 GMT
Subject: Re: [Bug] Permissions of temp file created in git's sha1_file.c correct?
Message-ID: <0bd60a95adcc1cda50d6ff0e8d335216@kloeckner.org>
In-Reply-To: <m3vd128uz7.fsf@localhost.localdomain>

On Wed, 02 Feb 2011 04:47:20 -0800 (PST), Jakub Narebski <jnareb@gmail.com>
wrote:

> That is why there is `core.fileMode` config variable, see git-config(1)
> manpage:

I saw (and liked) that option but I do not want to use the executable bit
on my "FAT-compatible" file system for other things anyways.

> Didin't git detect that it was on such filesystem?

I did not create the repository on the device, it was simply copied onto
the volume.
One `git init` later: Git detects that the filesystem knows about the
executable flag and sets filemode = true. (On a real FAT, it would be set
to false, of course).

> Not true.  Whether file can be renamed or deleted is governed by
> permissions in directory that contains given file, not the file itself
> (at least on POSIX).  If I understand things correctly, of course.

OK, then that's a reason to fix my implementation. (Done.)

I still think that a file that is to be written to should not be created
without the right to do so.

> Note that git doesn't store full permissions, in particular read/write
> permissions, only executable bit and "is symlink" thing.

Executable is all I need. Since there was only one bit left in the FAT
directory entries, my file system uses this bit to store the executable
flag. Writeable is derived from/mapped to !readonly. That's all I changed,
compared to a normal FAT.

Thanks for the very fast answer,
   Daniel