Re: git-pasky file mode handling

From: Petr Baudis <pasky@ucw.cz>
Date: Apr 16, 2005, 14:56 UTC
Message-ID: <20050416145650.GB19099@pasky.ji.cz>
In-Reply-To: <20050416104559.A12943@flint.arm.linux.org.uk>

Dear diary, on Sat, Apr 16, 2005 at 11:45:59AM CEST, I got a letter where Russell King <rmk@arm.linux.org.uk> told me that...

> Hi,

Hello,

Show 10 quoted lines

> It seems that there's something weird going on with the file mode
> handling.  Firstly, some files in the git-pasky repository have mode
> 0664 while others have 0644.
> 
> Having pulled from git-pasky a number of times, with Petr's being the
> "tracked" repository, I now find that when I do an update-cache --refresh,
> it complains that the files need updating, despite show-diff showing no
> differences.  Investigating, this appears to be because the file modes
> are wrong for a number of the files.  All my files do not have group
> write.

this is was a problem with git apply, which did not apply mode changes correctly until recently. If you have no local changes,

	checkout-cache -f -a

should fix this. Hopefully.

> I notice in the changelog what appears to be a dependence on the umask.
> If this is so, please note that git appears to track the file modes,
> and any dependence upon the umask is likely to screw with this tracking.

I personally don't think I like the mode tracking at all. Some people (Linus?) may want to have group +w. Other people (me) have their default group as 'users', and I definitively don't want everyone to be able to write to the files. :-)

I think we should track only whether the file is executable or not. Linus?

-- 
				Petr "Pasky" Baudis
Stuff: http://pasky.or.cz/
C++: an octopus made by nailing extra legs onto a dog. -- Steve Taylor