scd: PC/SC pinpad input improvement

NIIBE Yutaka gniibe at fsij.org
Fri Aug 30 10:50:01 CEST 2013


On 2013-08-30 at 09:40 +0200, Werner Koch wrote:
> Thus for Windows it is up to the Windows driver for the reader (which
> seems to be either the generic CCID driver or a customized SCM
> version).  If that is only a problem on Windows we may treat this as a
> problem of the Windows driver (maybe there is even a Registry entry
> tochange that).  OTOH, it shouldn't harm if we add a configuration
> option to allow the user to change it.

I think that this is a kind of corner case.  Card reader should just
work with no configuration.  When it properly emits time extension
reply message, there is no need to specify such a low-level timeout,
in the first place.

When we need to specify, I think that it would be easier for a user to
use system wide setting to specify pinpad timeout (if available).
That's because I don't think it is application specific.

It would be card reader specific.  The method of input PIN would be
different.

Suppose that there is another application than GnuPG, which uses the
card reader (well, it would use different kinds of cards).  In this
case, a user naturally wants to specify pinpad timeout, system widely.

If we will find there's no way to change the default pinpad timeout,
then, we need to support the way to specify it by the application.

When an application needs timeout to cancel authentication, it is
different story.  In this case, it is an application which specifies
timeout.  It is only application which knows this kind of context.
-- 





More information about the Gnupg-devel mailing list