[From nobody Mon Aug 29 15:22:52 2005 X-From-Line: marc.abramson@free.fr Mon Aug 29 12:41:05 2005 Return-path: <marc.abramson@free.fr> Envelope-to: domi@localhost Delivery-date: Mon, 29 Aug 2005 12:41:05 +0200 Received: from localhost ([127.0.0.1]) by gandalf with esmtp (Exim 4.52) id 1E9h4Y-0004jD-Ni for domi@localhost; Mon, 29 Aug 2005 12:41:04 +0200 Delivered-To: online.fr-domi.dumont@free.fr Received: from imap.free.fr [212.27.48.2] by localhost with IMAP (fetchmail-6.2.5) for domi@localhost (single-drop); Mon, 29 Aug 2005 12:41:02 +0200 (CEST) Received: (qmail 2064 invoked from network); 29 Aug 2005 10:40:16 -0000 Received: from postfix3-2.free.fr (213.228.0.169) by mrelay3-2.free.fr with SMTP; 29 Aug 2005 10:40:16 -0000 Received: from imp3-q.free.fr (imp3-g19.free.fr [212.27.42.3]) by postfix3-2.free.fr (Postfix) with ESMTP id B47ACC019; Mon, 29 Aug 2005 12:40:15 +0200 (CEST) Received: by imp3-q.free.fr (Postfix, from userid 33) id 9A4DE4729F; Mon, 29 Aug 2005 12:40:15 +0200 (MEST) Received: from gam75-1-81-57-22-129.fbx.proxad.net (gam75-1-81-57-22-129.fbx.proxad.net [81.57.22.129]) by imp4.free.fr (Horde) with HTTP for <marc.abramson@free.fr>; Mon, 29 Aug 2005 12:40:15 +0200 Message-ID: <20050829124015.rl9ij5u3eu8gokk8@imp4.free.fr> Date: Mon, 29 Aug 2005 12:40:15 +0200 From: marc.abramson@free.fr To: vdr@linuxtv.org Cc: Dominique Dumont <domi.dumont@free.fr> References: <87u0h99mn6.fsf@gandalf.hd.free.fr> In-Reply-To: <87u0h99mn6.fsf@gandalf.hd.free.fr> User-Agent: Internet Messaging Program (IMP) H3 (4.0.2) X-SA-Exim-Connect-IP: 127.0.0.1 X-SA-Exim-Mail-From: marc.abramson@free.fr Subject: Re: [vdr] Querying CAM doesn't work X-Spam-Checker-Version: SpamAssassin 3.0.4 (2005-06-05) on gandalf.hd.free.fr X-Spam-Level: X-Spam-Status: No, score=-2.6 required=5.0 tests=BAYES_00,NO_REAL_NAME autolearn=no version=3.0.4 X-SA-Exim-Version: 4.2 (built Thu, 03 Mar 2005 10:44:12 +0100) X-SA-Exim-Scanned: Yes (on gandalf) Lines: 90 Xref: gandalf.hd.free.fr mail.misc:9024 MIME-Version: 1.0 Hello Klaus. you are 100% right. The EN50221 spec does clearly define a QUERY procedure. However, this query is not correctly handle by a lot of module, because for a lot of different CA_smart_card, the only way to be 100% sure if the smart card is able or not to descramble something is to request a real descrambling. But then, if it works, it means that the smart card will also have started to descrambling, and also (depending of what was in the encryption message), to use token or the card, or to start a preview... From all the CA_system I know , there is only one CA system for which the smartcard itself allow a query that will not start "consumming" anything, but only answer the query. another reason is that doing a query to the card is most of the time silly. For example, imagine that the card has not received the monthly subscription.. so, the answer to the query will be 'descrambling not possible' and the host will then NOT request a descrambling... but then, if the card receive the subscription in the next 10 s after the query, the card would be able to descrypt. So, using correctly the query means also doing a query periodicaly . anyway, I do agree that, even if the query is not working well on the card side, a module shall not be disturbed by a query, and shall answer... I can not speak for the SCM module , since I don't how they react. Concerning the ASTON module , all the ASTON V2.0 module (the one with a green led) are fully supporting the query... even if some of them are alway answering by "ok: descrambling possible" , even if the smartcard does not contains the subscription. For the aston module, the answer ok means "descrambling MAY be possible".. and not "descrambling is possible". for some Aston module, depending of the CA_system, the module react in a better way.. It check if there is smart card in the module, and, if yes, it check if the smart card does correspond to the operator targeted by the stream. and, it answer "OK" if this 2 conditions are ok, or "descramling not possible" in all other case. And, for one of the last module done at Aston, with one CA_system, (I'm sorry, but I can not give name, because we have signed some NDA with all CA_system), the smart card is able to do a query, and then the module does answer correctly in all cases. For ALL OTHER MODULES, (MATRIC, MAGIC, POWERCAM, REALITY, PREDATOR), this module are ILLEGAL , since they are copying patended system. The guys which are the author of this things are also only interesting to make money, and to to give a real service, with a support. So, you can not expect this kind of module to be compliant with the EN50221 norm. According to EN 50221 an application should be able to query whether a given CAM can decrypt an actual broadcast by sending it a CA_PMT with a ca_pmt_cmd_id of "query" instead of "ok_descrambling". However, if I change the line capmt[length++] = CPCI_OK_DESCRAMBLING; in VDR/ci.c to capmt[length++] = CPCI_QUERY; I never get a ca_pmt_reply as specified in EN 50221, section 8.4.3.5. Does anybody have an idea why that is so? Is there a bug in VDR's CAM handling? Could it be that the CAMs I have just don't behave correctly? The reason why I'd like this to work is because if somebody has more than one device with CAMs that supports a certain encryption method, but only one of them has an actual smart card in it and can thus actually decrypt, it is important that VDR selects the right device. However, it can't do so if the CAM doesn't react on a query. Klaus ]