[virt-tools-list] [PATCH virt-viewer 2/2] Quit application if in kiosk mode and authentication has been cancelled

[Pavel Grunt]

On Mon, 2017-05-22 at 15:50 +0200, Victor Toso wrote:
> On Mon, May 22, 2017 at 10:39:15AM -0300, Eduardo Lima (Etrunko)
> wrote:
> > On 21/05/17 04:23, Xiaodai Wang wrote:
> > > 
> > > ----- Original Message -----
> > > > From: "Victor Toso" <lists at victortoso.com>
> > > > To: "Eduardo Lima (Etrunko)" <etrunko at redhat.com>
> > > > Cc: virt-tools-list at redhat.com
> > > > Sent: Saturday, May 20, 2017 5:26:29 AM
> > > > Subject: Re: [virt-tools-list] [PATCH virt-viewer 2/2] Quit
> > > > application if in kiosk mode and authentication has been
> > > > cancelled
> > > > 
> > > > Hi,
> > > > 
> > > > On Fri, May 19, 2017 at 05:16:53PM -0300, Eduardo Lima
> > > > (Etrunko) wrote:
> > > > > Related: https://bugzilla.redhat.com/show_bug.cgi?id=1446161
> > > > > 
> > > > > Signed-off-by: Eduardo Lima (Etrunko) <etrunko at redhat.com>
> > > > 
> > > > I'm not 100% sure that this would be the desirable behavior.
> > > > If one runs
> > > > the app in -k, isn't it the point that the user would not have
> > > > access to
> > > > the desktop?
> > > > 
> > > > AFAICT, the available options are:
> > > > 
> > > > 1) Click cancel, close virt-viewer (this patch)
> > > > 2) Click cancel, open dialog again (your suggestion on
> > > > bz#1446161)
> > > > 3) Remove cancel button as (2) shows that it is useless
> > > 
> > >        -k, --kiosk
> > >        Start in kiosk mode. In this mode, the application will
> > > start in fullscreen with minimal UI. 
> > >        It will prevent the user from quitting or performing any
> > > interaction outside of usage of the remote desktop session.
> > > 
> > > From the man page of virt-viewer, i think 2) or 3) is better
> > > than 1).
> > 
> > Hiding/blocking cancel button is, IMO, useless in this case.
> 
> Why? I don't understand. Useless would be a cancel button that is
> show
> but can't really cancel anything (as virt-viewer will be kept alive
> in
> kiosk mode waiting for input to do something).
> 
> > We will only add unnecessary complexity to the dialog code for
> > this
> > specific use. I will rather go with option 2 then.
> 
> Yeah, I don't mind keeping code simple.
> 
> > 
> > By the way, it would be nice to have some documentation about what
> > are
> > the goals with this feature.
> 
> My take for this feature is a client environment where a user should
> never be using the local machine, only the remote one. There is no
> explicit use case in the manual but there is hints that suggest that
> (so
> I agree that we could document use cases somewhere, unless it is
> already
> documented and we don't know here!)

For instance a client system which runs `virt-viewer -k` just after
the boot. The user must not touch the client system, must log in. So
cancelling the dialog or putting invalid credentials should bring the
dialog again.

https://bugzilla.redhat.com/show_bug.cgi?id=872723

> 
> Cheers,
>     toso
> _______________________________________________
> virt-tools-list mailing list
> virt-tools-list at redhat.com
> https://www.redhat.com/mailman/listinfo/virt-tools-list