Contents of /trunk/Drivers/ccid/README

USB CCID IFD Handler
====================

  This package provides the source code for a generic USB CCID
(Chip/Smart Card Interface Devices) driver. See [1] for the USB CCID
specifications from the USB working group.


Author:
=======

- Ludovic Rousseau <ludovic.rousseau@free.fr>


Supported CCID readers:
=======================

(in alphabetical order)
- Gemplus GemPC 433 SL [2]
- Gemplus GemPC Key [3]
- Gemplus GemPC Twin [4]
- LTC31 from C3PO [8]
- OmniKey CardMan 3121 [5]
- SCM Micro SCR 331 [6]
- SCM Micro SCR 335 [7]
- SCM Micro SPR 532 [9]


Unsupported CCID readers:
=========================

None known.


Supported operating systems:
============================

- GNU/Linux (libusb 0.1.7)
- MacOS X/Darwin (libusb 0.1.8beta, CVS snapshot. See "Known problems")
  to libusb)


Known problems:
===============

All platforms
    T=1 and TPDU
    """"""""""""
      I do not yet included support of T=1 cards when the reader is in
      TPDU mode.
      If you have a T=1 card I recommand you to use a reader working in
      APDU mode. T=1 and TPDU will be supported in a future version of
      the driver.

    readers enumeration
    """""""""""""""""""
      You may have wrong reader identification if you connect two or
      more reader at the exact same time.

      The daemon pcscd and the CCID driver do not list connected readers
      in the same order. The driver has no way to know which reader has
      been detected as a new reader by pcscd.
      
      So if you plug two readers at the same time (for example if you
      start pcscd with two readers connected or if you plug a USB hub
      with two readers on the hub) you may have this (with two readers
      foo and bar):
      - pcscd scans the USB bus and finds a new CCID reader foo and
        calls the driver
      - the CCID driver scan the USB bus and finds a new different
        reader bar
      - pcscd continues scanning the USB bus and finds the second reader
        bar and calls the driver
      - the CCID driver scan the USB bus and finds an unused reader foo.

        Now the reader foo will be called bar by pcscd and bar will be
        called foo. This may not be a problem if the two readers have
        the same name. But it is strange to insert a card in the
        physical reader foo and see pcscd detect the card insertion in
        the reader he calls bar.
        
        The solution is to use IFDHCreateChannelByName() instead of
        IFDHCreateChannel() and specify the USB device to use. But pcscd
        does not implement IFDHCreateChannelByName() yet. Maybe in the
        future.
        
        Another solution is to plug readers one after the other and not
        at the same time.

MacOS X/Darwin
    hotplug
    """""""
      Due to libusb work in progress the driver may crash on a reader
      insertion after a few insert/remove actions.
      So if your reader is always connected you should not have
      problems.
      A new version of libusb should fix this bug

*BSD
    libusb
    """"""
      The libusb does not yet support USB interrupt endpoints. You will
      notice that the driver will stop working as soon as you insert a
      card in the reader.
      A new version of libusb should fix this bug.


Licence:
========

  This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 2 of the License, or (at your
option) any later version.

  This program is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
General Public License for more details.

  You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software Foundation,
Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA


History:
========

0.3.2 - 4 November 2003, Ludovic Rousseau
    - src/commands.c: correct a stupid bug that occurs with an APDU with
      2 bytes response.
    - Info.plist: add SPR 532 in list of supported readers
    - parse.c: do not exit if the InterfaceClass is 0xFF (proprietary).
      It is the case with old readers manufactured before the final
      release of the CCID specs.
    - move LTC31 reader from unsupported to supported reader list. It
      was my f ault since in used odd INS byte in my test applet and odd
      INS bytes are forbidden by ISO 7816-4 ch. 5.4.2 Instruction byte.
      Thanks to Josep Mon� s Teixidor for pointing the problem.
    - src/commands.c: comment out the automatic GET RESPONSE part. I
      don't think it should be in the driver. Maybe in pcscd instead?

0.3.1 - 23 September 2003, Ludovic Rouseau
    - add --enable-multi-thread (enabled by default) for thread safe
      support an APDU multiplexing. You will need pcsc-lite-1.2.0-rc3 or
      above to use this feature.
    - add --enable-libusb=PATH option is your libusb is not installed in
      /usr or /usr/local
    - honor DESTDIR in install rules (closes [ #300110 ]). Thanks to
      Ville Skytt� for the patch.
    - src/ccid.c: do not switch the GemPC Key and GemPC Twin in APDU
      mode since it also swicth in EMV mode and may not work with non
      EMV cards
    - src/ccid_serial.c: complete reimplementation of the Twin serial
      protocol using a finite state automata (code much simpler)

0.3.0 - 10 September 2003, Ludovic Rousseau
    - support of GemPC Twin connected to a serial port. Thanks to Niki
      W. Waibel for a working prototype.
    - support of auto voltage at power up if the reader support it
      instead of forcing a 5V in all cases.
    - support of APDU mode instead of just TPDU if the reader support
      it. Thanks to Jean-Luc Giraud for the idea and inspiration I got
      from his "concurrent" driver.
    - support of "time request" from the card.
    - parse: new indentation for more readability of supported features.
    - switch the GemPC Key and GemPC Twin in APDU mode since they
      support it but do not announce it in the dwFeatures.
    - new build process using autoconf/automake.

0.2.0 - 26 August 2003, Ludovic Rousseau
    - Works under MacOS X
    - Info.plist: use an <array></array> for the alias enumeration
    - Makefile rework for *BSD and MacOS X

0.1.0 - 13 August 2003, Ludovic Rousseau
    - First public release


Bibliography:
=============

[1] http://www.usb.org/developers/devclass_docs/ccid_classspec_1_00a.pdf
[2] http://www.gemplus.com/products/gempc433/
[3] http://www.gemplus.com/products/gempckey/
[4] http://www.gemplus.com/products/gempctwin/
[5] http://www.omnikey.com/en/produkt_details.php3?produkt=1&variante=47
[6] http://www.scmmicro.com/security/SCR331.html
[7] http://www.scmmicro.com/security/SCR335.html
[8] http://www.c3po.es/pcsc.htm
[9] http://www.scmmicro.com/security/SPR532.html

$Id$

 vim:ts=20
1	USB CCID IFD Handler
2	====================
3
4	This package provides the source code for a generic USB CCID
5	(Chip/Smart Card Interface Devices) driver. See [1] for the USB CCID
6	specifications from the USB working group.
7
8
9	Author:
10	=======
11
12	- Ludovic Rousseau <ludovic.rousseau@free.fr>
13
14
15	Supported CCID readers:
16	=======================
17
18	(in alphabetical order)
19	- Gemplus GemPC 433 SL [2]
20	- Gemplus GemPC Key [3]
21	- Gemplus GemPC Twin [4]
22	- LTC31 from C3PO [8]
23	- OmniKey CardMan 3121 [5]
24	- SCM Micro SCR 331 [6]
25	- SCM Micro SCR 335 [7]
26	- SCM Micro SPR 532 [9]
27
28
29	Unsupported CCID readers:
30	=========================
31
32	None known.
33
34
35	Supported operating systems:
36	============================
37
38	- GNU/Linux (libusb 0.1.7)
39	- MacOS X/Darwin (libusb 0.1.8beta, CVS snapshot. See "Known problems")
40	to libusb)
41
42
43	Known problems:
44	===============
45
46	All platforms
47	T=1 and TPDU
48	""""""""""""
49	I do not yet included support of T=1 cards when the reader is in
50	TPDU mode.
51	If you have a T=1 card I recommand you to use a reader working in
52	APDU mode. T=1 and TPDU will be supported in a future version of
53	the driver.
54
55	readers enumeration
56	"""""""""""""""""""
57	You may have wrong reader identification if you connect two or
58	more reader at the exact same time.
59
60	The daemon pcscd and the CCID driver do not list connected readers
61	in the same order. The driver has no way to know which reader has
62	been detected as a new reader by pcscd.
63
64	So if you plug two readers at the same time (for example if you
65	start pcscd with two readers connected or if you plug a USB hub
66	with two readers on the hub) you may have this (with two readers
67	foo and bar):
68	- pcscd scans the USB bus and finds a new CCID reader foo and
69	calls the driver
70	- the CCID driver scan the USB bus and finds a new different
71	reader bar
72	- pcscd continues scanning the USB bus and finds the second reader
73	bar and calls the driver
74	- the CCID driver scan the USB bus and finds an unused reader foo.
75
76	Now the reader foo will be called bar by pcscd and bar will be
77	called foo. This may not be a problem if the two readers have
78	the same name. But it is strange to insert a card in the
79	physical reader foo and see pcscd detect the card insertion in
80	the reader he calls bar.
81
82	The solution is to use IFDHCreateChannelByName() instead of
83	IFDHCreateChannel() and specify the USB device to use. But pcscd
84	does not implement IFDHCreateChannelByName() yet. Maybe in the
85	future.
86
87	Another solution is to plug readers one after the other and not
88	at the same time.
89
90	MacOS X/Darwin
91	hotplug
92	"""""""
93	Due to libusb work in progress the driver may crash on a reader
94	insertion after a few insert/remove actions.
95	So if your reader is always connected you should not have
96	problems.
97	A new version of libusb should fix this bug
98
99	*BSD
100	libusb
101	""""""
102	The libusb does not yet support USB interrupt endpoints. You will
103	notice that the driver will stop working as soon as you insert a
104	card in the reader.
105	A new version of libusb should fix this bug.
106
107
108	Licence:
109	========
110
111	This program is free software; you can redistribute it and/or modify
112	it under the terms of the GNU General Public License as published by the
113	Free Software Foundation; either version 2 of the License, or (at your
114	option) any later version.
115
116	This program is distributed in the hope that it will be useful, but
117	WITHOUT ANY WARRANTY; without even the implied warranty of
118	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
119	General Public License for more details.
120
121	You should have received a copy of the GNU General Public License
122	along with this program; if not, write to the Free Software Foundation,
123	Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
124
125
126	History:
127	========
128
129	0.3.2 - 4 November 2003, Ludovic Rousseau
130	- src/commands.c: correct a stupid bug that occurs with an APDU with
131	2 bytes response.
132	- Info.plist: add SPR 532 in list of supported readers
133	- parse.c: do not exit if the InterfaceClass is 0xFF (proprietary).
134	It is the case with old readers manufactured before the final
135	release of the CCID specs.
136	- move LTC31 reader from unsupported to supported reader list. It
137	was my f ault since in used odd INS byte in my test applet and odd
138	INS bytes are forbidden by ISO 7816-4 ch. 5.4.2 Instruction byte.
139	Thanks to Josep Mon� s Teixidor for pointing the problem.
140	- src/commands.c: comment out the automatic GET RESPONSE part. I
141	don't think it should be in the driver. Maybe in pcscd instead?
142
143	0.3.1 - 23 September 2003, Ludovic Rouseau
144	- add --enable-multi-thread (enabled by default) for thread safe
145	support an APDU multiplexing. You will need pcsc-lite-1.2.0-rc3 or
146	above to use this feature.
147	- add --enable-libusb=PATH option is your libusb is not installed in
148	/usr or /usr/local
149	- honor DESTDIR in install rules (closes [ #300110 ]). Thanks to
150	Ville Skytt� for the patch.
151	- src/ccid.c: do not switch the GemPC Key and GemPC Twin in APDU
152	mode since it also swicth in EMV mode and may not work with non
153	EMV cards
154	- src/ccid_serial.c: complete reimplementation of the Twin serial
155	protocol using a finite state automata (code much simpler)
156
157	0.3.0 - 10 September 2003, Ludovic Rousseau
158	- support of GemPC Twin connected to a serial port. Thanks to Niki
159	W. Waibel for a working prototype.
160	- support of auto voltage at power up if the reader support it
161	instead of forcing a 5V in all cases.
162	- support of APDU mode instead of just TPDU if the reader support
163	it. Thanks to Jean-Luc Giraud for the idea and inspiration I got
164	from his "concurrent" driver.
165	- support of "time request" from the card.
166	- parse: new indentation for more readability of supported features.
167	- switch the GemPC Key and GemPC Twin in APDU mode since they
168	support it but do not announce it in the dwFeatures.
169	- new build process using autoconf/automake.
170
171	0.2.0 - 26 August 2003, Ludovic Rousseau
172	- Works under MacOS X
173	- Info.plist: use an <array></array> for the alias enumeration
174	- Makefile rework for *BSD and MacOS X
175
176	0.1.0 - 13 August 2003, Ludovic Rousseau
177	- First public release
178
179
180	Bibliography:
181	=============
182
183	[1] http://www.usb.org/developers/devclass_docs/ccid_classspec_1_00a.pdf
184	[2] http://www.gemplus.com/products/gempc433/
185	[3] http://www.gemplus.com/products/gempckey/
186	[4] http://www.gemplus.com/products/gempctwin/
187	[5] http://www.omnikey.com/en/produkt_details.php3?produkt=1&variante=47
188	[6] http://www.scmmicro.com/security/SCR331.html
189	[7] http://www.scmmicro.com/security/SCR335.html
190	[8] http://www.c3po.es/pcsc.htm
191	[9] http://www.scmmicro.com/security/SPR532.html
192
193	$Id$
194
195	vim:ts=20
Name	Value
svn:eol-style	native
svn:keywords	Author Date Id Revision