--DocE+STaALJfprDB
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Jun 09, 2011 at 07:49:06AM +0900, Yukihiro Matsumoto wrote:
> Hi,
> 
> In message "Re: [ruby-core:36863] Object#trust vs Object#taint"
>     on Thu, 9 Jun 2011 06:37:13 +0900, Aaron Patterson <aaron / tenderlovemaking.com> writes:
> 
> |What is the difference between tainting and untrusting an object?
> 
> Tainting is a mark for data from outside world.  Data from outside
> cannot be trusted.  Untrusting is a mark for data from untrusted code
> (that run under $SAFE level 4).

So who decides to trust / untrust an object?

> |When should library authors taint vs untrust objects?
> |When should application authors taint or untrust objects?
> 
> Basically you don't have to taint or untrust object explicitly.  Ruby
> should mark those flags automatically.  The authors can clear those
> flag, but I recommend strict validation before clearing flags.

When does Ruby set the untrust flag?  I can't seem to find any
documentation on when that will happen.

-- 
Aaron Patterson
http://tenderlovemaking.com/

--DocE+STaALJfprDB
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (Darwin)

iQEcBAEBAgAGBQJN8BNMAAoJEJUxcLy0/6/GbzsIAJuvwzVrxRekeqOVd3Il3j5a
zicr4+I1KqrpeMWUSKmqRBrrEIVUs9OTDW6ePAwxbvQJOFrYh+rkLGoz7PWb5uLb
E10SvjfmXnEh4TQCLhPObiqcIHhTanEWZo617SX/zVdPJ58FTlrnVli/l8TNPl/F
dxK1WWy11Q43jt7zCiuK5ODVfFieg1OVRjHS5fNrCO7/bH5sh6lyQObY0wmwt/oi
zOwOVk+RoARYtdySPSkYu0ChOHAH/2+beiLhZEvuOIs6W68EJC5mHz+J4WgrEFPm
GmymIdL75lG02Qoupyb49pdRLAeez4kfJbPbt9H9k8tvdf1GxbHgjGEkhCdMCPYnr
-----END PGP SIGNATURE-----

--DocE+STaALJfprDB--