# RE: Git-daemon messing up permissions for gitweb

2 messages from 2006-06-09 to 2006-06-09. Participants: Post, Mark K, Jakub Narebski.
Thread: https://gitlist.dev/t/4447

## Post, Mark K, 2006-06-09 19:11

Subject: RE: Git-daemon messing up permissions for gitweb
Message-ID: <5A14AF34CFF8AD44A44891F7C9FF410507957855@usahm236.amer.corp.eds.com>
URL: https://gitlist.dev/e/5A14AF34CFF8AD44A44891F7C9FF410507957855%40usahm236.amer.corp.eds.com

```
I just triple-checked, and both those things are already in place.  The
default umask is 0022, and all the config files have:
[core]
        repositoryformatversion = 0
        filemode = true
        sharedrepository = true

I see that the case of the value is different from what you typed:
SharedRepository 
Is that significant (as almost everything is)?


Mark Post

-----Original Message-----
From: Linus Torvalds [mailto:torvalds@osdl.org] 
Sent: Friday, June 09, 2006 3:06 PM
To: Post, Mark K
Cc: git@vger.kernel.org
Subject: Re: Git-daemon messing up permissions for gitweb



On Fri, 9 Jun 2006, Post, Mark K wrote:
>
> As far as I can tell, the problem is happening because these files are
> being written out with file permissions of 640, and since Apache is
> running as user wwwrun, it can't read them:

You can either make sure that people have something like

	umask 0022

in their bashrc (or that it's the default umask), so that they do things

world-readably by default.

Or add
	[core]
		SharedRepository = true

to the repository config file.

		Linus

```

## Jakub Narebski, 2006-06-09 19:27

Subject: RE: Git-daemon messing up permissions for gitweb
Message-ID: <e6ci1u$939$1@sea.gmane.org>
URL: https://gitlist.dev/e/e6ci1u%24939%241%40sea.gmane.org
In-Reply-To: <5A14AF34CFF8AD44A44891F7C9FF410507957855@usahm236.amer.corp.eds.com>

```
Post, Mark K wrote:

> [core]
>         repositoryformatversion = 0
>         filemode = true
>         sharedrepository = true
> 
> I see that the case of the value is different from what you typed:
>         SharedRepository = true
> Is that significant (as almost everything is)?

No, the keys are case-insensitive.

-- 
Jakub Narebski
Warsaw, Poland

```
