[darcs-devel] Any answers for IsiSetup's concerns?

[David Roundy]

On Wed, Dec 12, 2007 at 09:29:06AM -0800, zooko wrote:
> On Dec 12, 2007, at 9:14 AM, David Roundy wrote:
> 
> >>Do you think that this is analogous to two different repositories
> >>containing source code having different requirements for the source
> >>code?
> >
> >I don't think so, I think that having darcs manage permissions would
> >be like having it manage file ownership.  We could do that, but file
> >ownership is meaningless on most other computers.  Similarly the
> >meaning of "g" and "o" are different on each computer, since the group
> >is different, and the set of users with access to the computer are
> >different.
> 
> So, concretely, suppose that there were a darcs patch type for  
> changing permissions, and suppose that a patch of that type could  
> express something like "make this file unreadable by people that  
> aren't the owner".  Actually, let's back up and ask "What kinds of  
> changes to permissions do people need?".  The IsiSetup folks, for  
> example want to manage certain kinds of permissions on files, and  
> apparently git+cogito gives them the ability to do this to their  
> satisfaction.  I would like to know what exactly they need to do.
> 
> I'm not convinced that it is a priori impossible for darcs to do as  
> well as git or SVN or other tools do.

On that page they also complain about git only managing the executable
bit.  I agree that managing the executable bit is a good idea.  But they'd
like more than git gives them, and that's where I disagree (and agree with
Linus).

> I know that I personally often want darcs to stop unsetting the "x"  
> bit.  We could imagine a "permission settings" patch type and patches  
> which say "set executable" and "set not-executable".  I know you've  
> already said that this would make sense.  I had thought, until your  
> recent message, that progress on this front, as well as on the  
> symlink front, was waiting for darcs-2 patch theory to be sorted out.

Certainly, I'm not going to be adding features like this myself until
darcs-2 is out.

However, darcs shouldn't be unsetting your "x" bit. I'm pretty certain that
is a bug!  :(
-- 
David Roundy
Department of Physics
Oregon State University