[prev in list] [next in list] [prev in thread] [next in thread] 

List:       ms-smartcardsdk
Subject:    Re: Microsoft PC/SC implementation for multithread
From:       Eric Perlin <ericperl () MICROSOFT ! COM>
Date:       2001-11-05 17:51:53
[Download RAW message or body]


The scenario that we want to avoid is:
Application presents a PIN, then crashes.
If no reset occurs, the card is opened for use for all the other
applications.

You should only need to handle this reconnect if reset logic at the
beginning of a transaction.
I don't believe a reset would occur while a transaction is in progress
in another context.

Hope this helps.

-----Original Message-----
From: Sergio Ortiz [mailto:blind@CTV.ES] 
Sent: Monday, November 05, 2001 1:10 AM
To: SmartCardSDK@DISCUSS.MICROSOFT.COM
Subject: Re: Microsoft PC/SC implementation for multithread


Hi Eric,

First..thanks for answering.

> I don't see how this can cause unstability.
The problem is that our application must be prepared to handle a card
reset in any part of the program. In case there's more than a thread (
in one or more processes ) I think that the card state should not be
changed because it's not possible to know the effect of a card reset in
the other processes.

A failure in a different process can affect me...so I should include
ALWAYS in design that a card reset can occur during a session in ANY
place. Design gets even more complicated because of this behaviour.

Imagine an scenario in which depending on a card type we should
authenticate the PIN. When a card reset is produced, I should reconnect
the card and reauthenticate the PIN depending on the card type.
But...where is the program execution point where the card reset appears?

I see this behaviour unnecessary. I think that an inconsistent state due
to a failure should be handled and cleaned inside the resource manager.

Perhaps I don't see the possible inconsistent card state scenario. Can
you give us an example? Anyway...I'll handle the card reset :)

Thanks in advance










>
> -----Original Message-----
> From: Sergio Ortiz [mailto:blind@CTV.ES]
> Sent: Thursday, October 25, 2001 8:19 AM
> To: SmartCardSDK@DISCUSS.MICROSOFT.COM
> Subject: Re: Microsoft PC/SC implementation for multithread
>
>
> Hi,
>
> > I developed a similar soft and I created a service to manage the 
> > reader
> and
> > every other process is a client to this service and it works well.
> So...did you have the same problem?
>
> My hypothesis:
> If a SCardConnected process dies or it's terminated without 
> SCardDisconnecting, the card is reset. So, all other processes using 
> the card have to reconnect...
>
> Is this correct?  I think this can cause unstability of the system.
>
> Bye
>

[prev in list] [next in list] [prev in thread] [next in thread]