[PATCH] posix: add configuration option to ignore execution mode changes (issue3301)

[Matt Mackall]

On Wed, 2012-03-07 at 11:52 +0100, marco gaddoni wrote:
> On Tue, Mar 6, 2012 at 19:40, Matt Mackall <mpm at selenic.com> wrote:
> 
> > On Tue, 2012-03-06 at 09:04 +0000, A. S. Budden wrote:
> > > On 2 March 2012 19:14, Matt Mackall <mpm at selenic.com> wrote:
> > > > On Fri, 2012-03-02 at 12:20 +0000, A. S. Budden wrote:
> > > >> # HG changeset patch
> > > >> # User A. S. Budden <abudden at gmail.com>
> > > >> # Date 1330690534 0
> > > >> # Node ID 105fae68f78292cebd41e118bc68b24c171fa553
> > > >> # Parent  85db991780b7b86913c3e29a5fe7ef25ab5ee93c
> > > >> posix: add configuration option to ignore execution mode changes
> > (issue3301)
> > > >>
> > > >> Using a posix version of Mercurial on a filesystem (such as FAT)
> > > >> without support for the executable bit can result in Mercurial
> > > >> erroneously reporting/committing executable bit changes.
> > > >
> > > > I'd rather not go this route. If you know of a scenario where we record
> > > > imaginary exec bit changes, report the details so we can work out
> > > > whether we can avoid it.
> > >
> > > Unfortunately, the problem is with Windows applications not knowing
> > > anything about Cygwin permissions and hence setting the executable bit
> >
> > I am not at all interested in your diagnosis or prescription until AFTER
> > I hear about the details of your symptoms.
> >
> >
> i have this problem too.
> the details are:
> the repo is on win7 ntfs filesystem.
> i have a repo committed with the windows version of tortoisehg
>    the files in the repository are marked as not executable
> 
> in cygwin, using the python & mercurial of cygwin
> (trying to avoid the bug that the windows version of python do not
> detect correctly that the terminal is interactive)
> I see all the files as executable.
> 
> $ hg diff Project.sln
> diff --git a/Project.sln b/Project.sln
> old mode 100644
> new mode 100755
> 
> maybe the solution is to use the noexec option
> of mount in cygwin.

I think we'll probably end up disabling exec support in cygwin by
default (and symlinks too).

See also: http://mercurial.selenic.com/bts/issue3301

-- 
Mathematics is the supreme nostalgia of our time.