**** BEGIN LOGGING AT Thu Mar 25 02:59:57 2010
Mar 25 05:49:11 <zhenhua> denkenz: still there?
**** BEGIN LOGGING AT Thu Mar 25 17:16:14 2010
Mar 25 17:16:28 <holtmann> If I reject primary and secondary NBNS options, then it works fine.
Mar 25 17:16:33 <kristenc> and so we keep sending another request, but it's never good enough for the modem.
Mar 25 17:17:48 <kristenc> could the list be dead?  I still don't see the patch there.  Maybe mail it to me directly?
Mar 25 17:19:45 <holtmann> I saw it.
Mar 25 17:19:57 <holtmann> It is on the list.
Mar 25 17:20:03 <holtmann> Problem is on your end ;)
Mar 25 17:20:12 <kristenc> or with linux.intel.com...
Mar 25 17:20:26 <kristenc> which is where I'm subscribed.
Mar 25 17:20:44 * kristenc mails self to test
Mar 25 17:21:37 <holtmann> Send another one to your linux.intel.com address.
Mar 25 17:22:03 <kristenc> weird - got the one you sent me directly right away.
Mar 25 17:22:17 <kristenc> and my test mail.
Mar 25 17:22:21 <kristenc> not sure what's up with that.
Mar 25 17:24:05 <kristenc> seems really straight forward.  Can you capture reads and writes with strace and mail them to me?  I want to see what the modem is sending you.
Mar 25 19:36:26 <holtmann> kristenc: I could, but I prefer that we add some proper debug to key points in the code first.
Mar 25 19:37:03 <holtmann> Especially for the LCP stage, we should have some debug support.
Mar 25 20:53:43 <denkenz> I hate gprs
Mar 25 20:56:33 <akiniemi> random review: commit 9390dbaf31f2bd61ff566eeaac6644dce1fd13b4 seems doc contradicts implementation? s/uint16/string?
Mar 25 20:59:08 <denkenz> Yeah the doc is wrong, we switched to strings a while ago and I just copied the error from network one
Mar 25 21:01:59 <denkenz> fixed
Mar 25 21:03:03 <akiniemi> By the way, I like the new tech names ;)
Mar 25 21:04:29 <denkenz> good, holtmann persuaded me they were better
Mar 25 21:05:01 <akiniemi> Personally, I could've gone with just 2g, 3g, 4g
Mar 25 21:05:13 <akiniemi> but I suppose we can expect some level of telco understanding here
Mar 25 21:06:13 <holtmann> akiniemi: Problem is that we would have needed 2.5g and 3.5g as well.
Mar 25 21:06:16 <denkenz> these might come into play once we have  cdma or dual mode modems
Mar 25 21:10:05 <akiniemi> And LTE is almost a household name nowadays anyway
Mar 25 21:43:36 <kristenc> holtmann: ping
Mar 25 21:46:05 <kristenc> nm, I sent you email.
Mar 25 22:11:04 <denkenz> akiniemi: still up?
Mar 25 22:36:33 <holtmann> kristenc: I am here.
Mar 25 22:42:42 <kristenc> holtmann: I took a look at the strace output -- it looks to me like we are doing what we are supposed to be, but obviously there must be something else it's not doing.
Mar 25 22:43:07 <kristenc> is it possible for you to connect via network manager/pppd and do an strace to see the traffic when pppd is talking to the modem?
Mar 25 22:43:18 <kristenc> it would help me to be able to see a working connection.
Mar 25 22:43:50 <kristenc> when I strace network manager, I have to tell strace to follow the processes and spit out each process into a separate file.
Mar 25 22:44:00 <kristenc> then pick the one that has the pppd process in it and send it to me.
Mar 25 22:45:27 <kristenc> I use strace -x -e trace=read,write -f -ff -o nm NetworkManager --no-daemon
Mar 25 22:45:54 <kristenc> then grep all the nm.* files for pppd.
Mar 25 22:46:28 <holtmann> kristenc: I have a working connection. Only when I accept the NBNS options instead of rejecting them this goes wrong.
Mar 25 22:46:40 <holtmann> Meaning when I am accepting all options it fails.
Mar 25 22:47:11 <holtmann> And the NetworkManager part is funny. Are you serious? I don't have that installed on my system ;)
Mar 25 22:47:12 <kristenc> holtmann: right, what I am hoping is that pppd accepts the NBNS option and we can see that
Mar 25 22:48:18 <kristenc> unfortunately, I can't duplicate this on my own hardware, so I have to rely on you to test this out.  We can always just reject the nbns options for now until I get more hardware.
Mar 26 02:08:04 <zgli> :denkenz, which modem implement the radio-setting feature? 
Mar 26 02:12:07 <denkenz> isi only
Mar 26 02:12:27 <zgli> ok , thanks
Mar 26 02:48:25 <zhenhua1> denkenz: i am not sure whether we need two flags. maybe i think too complicated.
Mar 26 02:48:53 <zhenhua1> one is for blocking for each command, another is for blocking for the whole command line.
Mar 26 02:49:15 <zhenhua1> server should not eat next command line before current whole line is fully parsed. right?
Mar 26 02:54:18 <denkenz> of course not
Mar 26 02:54:26 <denkenz> 1 command outstanding at a time
Mar 26 02:54:39 <denkenz> doesn't matter if it is on a multi-command line or single-command line
Mar 26 02:55:15 <zhenhua1> what if a new data comes in, will it trigger server_feed?
Mar 26 02:56:24 <zhenhua1> image an ATD need a long time to back, and a new line comes in, will it replace current line?
Mar 26 02:56:40 <zhenhua1> correct me if i am wrong
**** ENDING LOGGING AT Fri Mar 26 02:59:57 2010