[PATCH 1/3] Let the PAM client send its PID -- 2nd version

[Sumit Bose]

Hi,

this is the new version of the cli_pid patch with the missing size
added. There are two related patches.

0002 removes the unused client locale. It was added at a time where we
talked about generating all messages the client should display on the
server side.

Based on the previous discussion 0003 makes cli_pid a mandatory item in
the pam protocol and increments the protocol version to 3. IMO it is not
necessary to push it, but I wouldn't mind if you prefer it this way.

bye,
Sumit

[Simo Sorce]

On Mon, 2009-09-14 at 10:09 +0200, Sumit Bose wrote:
> Hi,
>
> this is the new version of the cli_pid patch with the missing size
> added. There are two related patches.
>
> 0002 removes the unused client locale. It was added at a time where we
> talked about generating all messages the client should display on the
> server side.

0001 and 0002 look good to me.
> Based on the previous discussion 0003 makes cli_pid a mandatory item
> in
> the pam protocol and increments the protocol version to 3. IMO it is
> not
> necessary to push it, but I wouldn't mind if you prefer it this way.

I don't think it is strictly necessary, but the extra check in the v3
parse function is worth to have. and version numbers are cheap anyway,
so I'd go for it.

ACK.

[Simo Sorce]

On Mon, 2009-09-14 at 11:36 -0400, Simo Sorce wrote:
> On Mon, 2009-09-14 at 10:09 +0200, Sumit Bose wrote:
> > Hi,
> >
> > this is the new version of the cli_pid patch with the missing size
> > added. There are two related patches.
> >
> > 0002 removes the unused client locale. It was added at a time where
>
> we
> > talked about generating all messages the client should display on
>
> the
> > server side.
>
> 0001 and 0002 look good to me.
> > Based on the previous discussion 0003 makes cli_pid a mandatory item
> > in
> > the pam protocol and increments the protocol version to 3. IMO it is
> > not
> > necessary to push it, but I wouldn't mind if you prefer it this way.
>
> I don't think it is strictly necessary, but the extra check in the v3
> parse function is worth to have. and version numbers are cheap anyway,
> so I'd go for it.
>
> ACK.

and puched

Simo.