few questions regarding org.freesmartphone.GSM.SIM

[Thomas Seiler]

> Can you please come up with a method/signal API draft for that so
> that we have a base for discussion?

<!-- GENERIC SIM ACCESS -->

    <method name="CallSIMFunction">
      <annotation name="org.freedesktop.DBus.GLib.Async" value="fso_gsm_sim" />
      <doc:doc>
         <doc:description>
           (OPTIONAL) Call a function on the SIM card and return the
result, or an error if the operation failed or is not allowed.
           Not all modems support Generic SIM Access and those who do
might restrict access to some instruction classes for security
reasons.
         </doc:description>
         <doc:inote>
           This can map to the GSM 07.07 command +CSIM (Generic SIM
Access), see 3GPP TS 07.07 Chapter 8.17
           It might also map to the GSM 07.07 command +CRSM
(Restricted SIM Access), see 3GPP TS 07.07 Chapter 8.18
         </doc:inote>
      </doc:doc>
      <arg type="s" name="command" direction="in">
         <doc:doc>
            <doc:summary>The command to send to the SIM encoded as a
ASCII HEX string.</doc:summary>
         </doc:doc>
      </arg>
      <arg type="s" name="response" direction="out">
         <doc:doc>
            <doc:summary>The response message from the SIM encoded as
a ASCII HEX string.</doc:summary>
         </doc:doc>
      </arg>
    </method>

I think this is all that is needed to hook up wpa-supplicant and
SIM-card, and keeping it this generic would allow for other
applications using the SIM
(i.e. crypto filesystem or libpam). The application using the

The TI Calypso is actually denying access to the GSM command class
(A0, http://lists.openmoko.org/pipermail/gsmd-devel/2007-March/000009.html)
However, the USIM commands are working fine (class 00).

wpa-supplicant uses by default the USIM challenge and uses GSM only as
a fallback. This means a wpa-supplicant EAP-SIM authentication would
only work
if the SIM is a USIM. But as most SIMs with a data flat are actually
USIM's (to benefit from the higher UMTS data rates), so this should
not be such a big limitation.

Comments welcome, as usual. And feel free to beat me to an actual
implementation : )

Cheers,
Thomas

-- 
Excercise 17:
If the human brain was simple enough for us to understand we'd be so
simple we couldn't understand.
Prove this by induction.