**** BEGIN LOGGING AT Fri Apr 09 02:59:56 2010 Apr 09 03:00:11 Right now I was just reviewing the changes. Apr 09 11:11:48 * Djdas is away: Away Apr 09 14:56:49 holtmann: those cleanups pushed, did you get a chance to test? Apr 09 15:13:03 denkenz: Do you have comments on the submitted stk code? Apr 09 15:13:15 I haven't looked at it yet Apr 09 15:13:23 too busy with other stuff Apr 09 15:14:23 denkenz: Not yet. Just pushed the CRC separation patch. Untested. So hope it still works ;) Apr 09 15:17:36 That's fine. I'm still writing the test cases. I met one problem: In some section, test cases for ber-tlv A and B are both used. Need I collect those cases for B and move them to test suite for B? Apr 09 15:18:31 Probably easiest just to implement the test cases in the same order as they are in the spec Apr 09 15:18:46 However, without the details I don't know for sure Apr 09 15:18:56 For now do what you think is best and if we need to change it, we will Apr 09 15:22:33 denkenz: An example is: Secion 27.22.4.17.1.4.2, which is a secion related to perform card apdu, also includes the cases of power on card. Apr 09 15:23:23 And most of secion includes cases related to terminal response and envelope Apr 09 15:24:26 So I think we may write the test strictly following the spec. Apr 09 15:25:19 Since you don't have terminal response figured out yet, write the relevant proactive command parsers Apr 09 15:25:29 We eventually need to extend all the tests with terminal responses Apr 09 15:25:48 parsers + tests even Apr 09 15:26:28 So these terminal response would be integrated with current cases? Apr 09 15:27:13 We'll see, it might be we use an additional test for the terminal responses, re-using the data structures from the proactive commands Apr 09 15:27:42 e.g. test_parse_proactive_comand_111, test_encode_response_111 Apr 09 15:28:09 Main problem is that the GLib testing framework is a piece-of-shit Apr 09 15:28:28 So you have to bend over backwards to get multiple test datasets Apr 09 15:28:32 ? Apr 09 15:29:13 Just try it, I don't have a final answer for you Apr 09 15:29:25 My feeling is a separate function for encoding terminal responses :) Apr 09 15:29:58 ok, thanks. I will think about it. But firstly make most of proactive command tests work:0 Apr 09 15:30:57 exactly, don't worry about terminal responses until we get that far ;) Apr 09 15:31:18 sure;) Apr 09 15:50:50 crc patch looks good and still works Apr 09 15:51:57 denkenz: Nice. **** ENDING LOGGING AT Sat Apr 10 02:59:57 2010